Home

Vectorbeam Series

image

Contents

1. i I i 1 i ia i 1 1 H 1 l a T 1 BE 1 i 1 CORRECTIONS tdi 1 MUP WF jJ FOCUS 1 iL MASTER gt i w amp STIG ia CU m 1 STIG STIGMATOR i i ba T a 1 1 1 Y a i i mis DOSE CLOCKS it il CONTROL a N DCP L La 7 1 BSD DETECTOR i 1 i MARK 1 l Al LOCATE VIDEO DEAD VIDEO i SEM amp INTERFACE AMPUFIER LASER f VIDEO 1 HEAD HEIGHT i AMPLIFIER HEIGHT SENSOR DATA amp CONTROL HEIGHT METER LI SENSOR 1 INTERFACE 1 1 1 I i 1 TRANSMISSIO DETECTOR Figure 4 4 18 and 20 bit pattern generator digital and analogue crates Part Number 878275 Vectorbeam Operator Manual 4 3 1 2 1 4 3 1 2 2 4 3 1 2 3 Page 41 18 bit The pattern generator consists of a crate of digital electronics and a crate of analogue electronics The digital part of the pattern generator has a 18 bit output i e the maximum main fieldsize is divided into 262144 exels The area of the mainfield which is covered by the subfields can be selected by the operator in the form of a square which is centred in the mainfield This defines the maximum area which can be used for exposing patterns The number of subfields in the X and Y directions into which this square is divided can be selected by the operator to be between 25 and 64 inclusive maximum 4096 in total Each subfi
2. essesss 257 LC NOMEN Pakem SEEN TET DOTT 258 248 Fielgsize adjustment FOSOIMION iiiuss ooa ren ces erae trei Io era REP Ye rt de yd ota 258 2491 6 bipattefmgemelalol s stet eater te cte e sebo eres 258 2492 e bitpattermgenelratol c iecit p eit i reo nS 259 24 9 9 20 bit pattern generator iide c mo re ERE 259 249 Gran generalo aide orta pa reu o ive ntis bv cub Lia ACA LUN eed annals 259 24 9 1 Detining the gratiis cioe coco erecti eoo EI cease ice ee et erede 260 24 9 2 Displaying grating definition eese 262 24 9 3 Clearing grating definition 262 24 9 4 Selecting the algorithmic grating generation 262 24 9 5 Defining the order of the shapes ssessssseees 262 23 10 ape eroBIOD ases estais dd aca ciSda oc ld qid ead dle dh Trdur dia ea OA RS 263 24 10 1 normal OpIOR iiri ct ete eric eot e aeiia 264 2410 2 Nodiseamd OpliOn crrr aa a 264 24 10 3 None or Noerode Option sea 265 2411 Zao dimension Shapes rar a a A 266 24 12 Algorthimio Moen eisiaa gaidi 266 25 Routine maintenance and servicing sesecsssosscesssocecssssoceesssseeee 207 20 Mu i Oe 27 Recovery from exception conditions 269 2 4 aobcontol Window IaeKs UD srianan tura acier dari EH c ede tard du Ed Td dd 269 27 2 Top Mall OF tbe scream GOSS DIAC ea esti et d ende ve Pert
3. sese 91 11 4 8 Choosing signal parameters sss 92 11 4 4 Choosing scan parameters ssssssessseeeenneren 94 11 5 Mak locaton COMMING sca eosetodcc orania cO SD esa d E liac pou Rc cras ci eet Ox cesi eU a da dua 95 1415 Mark 3dininistEaon TUTIODOTIS au cas 2s arredo sitos kit n deks ad cedi dn id dd dba du ns ead 95 11 7 Diagnostic output of the mark locate function eee 95 TLS PIR VS Cae GIO iT 96 pps MEE SIIIE LIIS enn 96 11 82 BisectioriSegreli iie in rete erri Eier oit iste a daran 96 T83 ARINC SCANCH eU 97 13189 Edge locate Wcl w lo i lc 1 meee etree Pen rereeet reer eer Terre err rerreetr ae Ferree res metre ttre er 98 WAGO Ts GoatseSearcl ue diiit radere A 98 11 92 JI BSSOaICl iei ottiene Has dieeure ito suaaiadsdessahedeneseaggaceasenes 99 TIO Gross locate alOOPIDET ani api rer pb errr enn rere renes tite babies ia rere errr 100 TIO HASLER SCANCH assesses osi te Eo I NESSRN EAS RE SXub sese dana ARa 100 T102 JIneSeatcliose d connecctsstenssacscnanedsatsocusanuies a aag 101 11 11 Actions if the mark locate consistently fails to find the Mark 104 EL GS is d FADE ERR EMOMNO DEUM E EU M DLP TED CM le 12 1 Database SITUE Be eir ieerri ir trek perta Qi Ede En ou nna REM E Rena RS nA aS 105 i22 Database Param Sterns cocido bien dee pte a 105 12 2 4 Headertorall databases a iri ei edant ssi votes ead
4. 0x400d0001 0x400d0002 0x400d0003 0x400d0004 0x400d0005 0x400d0006 Slave Micros Manager Errors Receive from SuP Manager Cmnd queue failed Reboot PG Watchdog timeout on SuP Mgr command send Reboot PG Watchdog timer Start error Reboot PG Watchdog timer Cancel error Reboot PG Send failed SuP Mgr gt Return queue Reboot PG Not allowed to send to this destination Reboot PG Vectorbeam Operator Manual Page 291 0x400d0007 Send failed Sup Manager gt qPatternIP Reboot PG 0x400d0008 No source for Async mailbox interrupt Reboot PG 0x400d0009 Send failed L W EOF interrupt gt qPatternIP Reboot PG 0x400d000A Slave inaccessible from VME bus not booted Possible PG configuration file error or slave not working Reboot PG 0x400d000B Couldn t find Sup Manager task s priority Reboot PG 0x400d000C Couldn t reset Sup Manager task s priority Reboot PG 0x400d000D Can t communicate with slave unknown H W type Possible PG configuration file error Reboot PG 0x400d000E Can t communicate with slave unknown mailbox Possible PG configuration file error Reboot PG 27 6 6 11 CPU Board Errors 0x400e0001 Error in DMA write Reboot PG 0x400e0002 Requested VME DMA xfer over 256 byte boundary Reboot PG 0x400e0003 Can t set up mailboxes Reboot PG 0x400e0004 DMA xfer not on 4 byte boundary Reboot PG 0x400e0005 Can t understand vxWorks version number Reboot PG 27 6 6 12 Pattern Writing errors Part Number 878275 0x400f0
5. Part Number 878275 Vectorbeam Operator Manual Page 257 Figure 24 1 Example of random field placement used for sparse tiling The pattern is an array of crosses shown in red The standard tiled mainfields 5x5 total 25 completely enclose 4 crosses in the lower left corner but only 1 or 2 otherwise and are shown in green The random field placements 3x3 total 9 each contain 4 crosses and are shown blue Random field placement also allows the area enclosed at each position to be varied 24 5 Selectable field correction interpolations 24 6 Optimising throughput The individual contributions of such things as calibration direct write alignment and pattern exposure to the total time for a job should be estimated This will highlight areas where improvements may be made In many cases changing parameters to increase the speed of operation results in a decrease in the pattern fidelity This performance is a fundamental characteristic of the system The goal is to achieve the best compromise for the application and the points listed below should be considered 24 6 4 Increasing current and spot size 24 6 2 Reducing command processing times The total command processing time overhead may be significant for jobfiles containing many commands Some command processing times in particular those due to the operation of Emma may be reduced by the following actions 1 Turn the logging off by typing qset logging off SLO
6. Unless there is an absolute length standard available it is only possible to match the x and y scaling to each other and rely on the accuracy of the laser interferometer for absolute length Therefore for simplicity the method described here only applies an x scale correction and the y scale is left at 1 0000000 This effectively sets the Y interferometer to be the reference for absolute length Also for simplicity only a y rotation correction is used and the x rotation is left at 0 0000000 The signs of the coefficients calculated from the data sets for each rotation reflect the amounts of scale keystone etc in the measured data In order to correct for the amounts shown an equal and opposite coefficient must be entered into Emma and this is already taken into account in the equations in the table Part Number 878275 Vectorbeam Operator Manual Page 141 Stage position Correction calculation Emma distortion MPST coefficient X scaling correction Syo70 Syigo YO Y scaling correction 1 0000000 X rotation correction 0 0000000 Y rotation correction x90 S aso Sm 4 X keystone correction K CE Sd Rig Kon 4 Y keystone correction Ks K o Kun Ko 4 X scaling Linearity Lo T Iu zm Fs L2 correction 1 Y scaling Linearity correction B L Ly99 Ly Ls 4 X bow correction B B o Bip Bios 4 Y bow correction B B o B go a By yo 4
7. Vectorbeam Operator Manual 27 6 8 27 6 8 1 27 6 8 2 Part Number 878275 0x400f0050 0x400f0051 0x400f0052 0x400f0053 0x400f0054 0x400f0055 0x400f0056 0x400f0057 0x400f0058 0x400f0059 EO error messages Ethernet Errors Page 298 Could not access the barcode reader link Could not open serial link to barcode reader Could not close serial link to barcode reader Could not allocate memory for barcode reader command Error reading from barcode reader serial link Barcode reader link timeout Incorrectly formatted reply received Barcode symbology not supported Barcode reader rejected command Could not read barcode All of these are internal so the advice is 1 Check ethernet cabling transceivers bridges etc 2 Reboot subsystem 0x40010001 0x40010002 0x40010003 0x40010004 0x40010005 0x40010006 0x40010007 0x40010008 0x40010009 0x4001000a 0x4001000b 0x1001000C 0x4001000D 0x4001000E 0x4001000F 0x40010010 0x400b0001 0x400b0002 0x400b0003 0x400b0104 0x400b01 41 0x400b0301 0x400b0302 0x400b0401 0x400b0402 0x400b041 1 0x400b0412 0x400b0500 0x400b0501 0x400b0502 0x400b0503 0x400b0504 0x400b0505 0x400b0506 0x400beeee 0x400c0101 0x400c0102 0x400c0103 0x400e0001 0x400e0002 0x400e0003 0x400e0004 0x400e0005 Ox30AA0001 O0x30AA0002 0x40AA0003 Couldn t spawn Ethernet output process Couldn t create Ethernet input queue Couldn t create Ethernet output queue Couldn t set up Ethernet packet in
8. Part Number 878275 Vectorbeam Operator Manual Page 167 following section unless there is a reason for not doing so e g the spot roundness is not important but speed is An automatic focus adjustment is carried out as follows 1 Adjust the gain and backoff as described in the section Video gain and backoff setup 2 If jobcal has not been run since the pattern generator was last rebooted it might be necessary to set the focus and stigmation algorithm step size parameter VB_OPER gt QSET REG 3 17 0 04 And set the focus and stigmation algorithm range parameter VB_OPER gt QSET REG 3 18 0 4 3 Run the automatic focus adjustment specifying the mark type e g FM VB OPER2QCAL FOC FM DACPOS 3 ACC 0 005 FILT 8 POINTS 8 SCANLENGTH 0 5 LINES 1 17 4 2 3 Automatic fine focus and stigmation adjustment An automatic fine focus and stigmation adjustment is carried out as follows 1 Adjust the gain and backoff as described in the section Video gain and backoff setup 2 If jobcal has not been run since the pattern generator was last rebooted it might be necessary to set the focus and stigmation algorithm step size parameter VB_OPER gt QSET REG 3 17 0 04 And set the focus and stigmation algorithm range parameter VB_OPER gt QSET REG 3 18 0 4 3 Run the automatic stigmation adjustment specifying the mark type e g FM VB_OPER gt QCAL STIG FM DACPOS 3 ACC 0 005 FILT 8 POINTS 8 SCANLENGTH 0 5 LINES 1 Th
9. Shift x Mid Align X 20 000 Kv 2 2 6 500 0 0 0 v 6300 0 0 019 0 013 0 021 0 003 0 000 0 000 0 00 0 77 0 09 0 82 Figure 7 12 Display Database Column Part Number 878275 Vectorbeam Operator Manual Page 56 The Display Database Pattern panel Figure 7 13 displays information about the currently selected pattern file Emma Pattern File Resolution 0 0000 nm Size X0 000 Y 0 000 mm Stage Step Size x0 YO exels Number of Fields 0 Cancel Figure 7 13 Display Database Pattern The Display Database Strategy panel Figure 7 14 displays information about the set up of the pattern generator Emma Invert Mode No Inversion VRU 0 Resolution 0 0000 nm SS Settling Ons SF Settling Ons AA Settling Ons Cancel Figure 7 14 Display Database Strategy The Display Database Sensitivity Figure 7 15 displays current sensitivity coefficients affecting the main field subfield and beam error feedback Main Field Bef Field Trap Field Main Pivot Trap Pivot Facusmeight Rotmeight 2 18 Amm 2 18 mm 14 18 Amm 40 80 mm 40 80 mm 0 008100 0 016730 Figure 7 15 Display Database Sensitivity Part Number 878275 Vectorbeam Operator Manual Page 57 The Display Database Calibration Figure 7 16 displays current calibration values affecting the main field subfield and beam error feedback Calibration EHT Calibration height Stage Position Main Field X Y Bef Field X Y Tr
10. e 4 LHS measurement fault No signal No plate Investigate e 5 LHS off LHS has not been calibrated since being switched off or reset Recalibrate e 6 LHS PSU fault PSU faulty investigate e 7 LHS system fault H W faulty investigate e 8 LHS comms fault PG cannot establish comms Faulty PG LHS or wire joining them e 9 LHS stage fault Stage not moved Not LHS fault Height sensor reset Reset the height sensor by pressing the red button on the edge of the PCB located in the top of Cabinet A and repeat QCAL HEIGHT command Check Ensure that there is a holder on the stage VB OPER MVSPFM move to the Focus Mark VB OPER QCAL HEIGHT TAB 7 calibrate the height sensor VB OPER QDISP HEIGHT measure the height Subfield calibration error If the subfield calibration fails with an error message and the residual errors are very large this is because the subfield deflection is not working There are a number of possible causes 1 The subfield calibrations coefficients have got themselves out of range This could happen if there is a glitch in the mark locate position or other disturbance to the tool Therefore try the command VB OPER Qqcal trap init load in order to zero the coefficients Then try running jobcal If the same error occurs during the subfield calibration part of jobcal then enter the qcal trap init load command again before moving on to point 2 below 2 Power off and on the PGD run star
11. 08019328 08019330 08019338 08019340 08019348 08019350 08019358 08019360 Part Number 878275 Page 285 COLUPDATE Unable to update column settings INVFILHDR Invalid file header FILEMPTY Pattern file empty UNABLEOPN Unable to open file SWERR Sub system software error SUBBUSY Sub system Busy INVBLKALN Invalid block alignment MISSPKT Missing ethernet packet during block transfer SLVCPUERR Slave Processor is not initialised for communications NOT PAT FIL Not a pattern file PTNHDRERR Pattern header transfer failed FLDADRERR Pattern field address transfer failed FLDHDRERR Pattern field header transfer failed FADDR Pattern file offset address invalid PTNDATA Pattern data transfer failed PGNRDY Pattern Generator is not ready for exposure PTNNSLC Pattern file is not selected PTNRESERR Pattern resolution does not match calibrated resolution PTNLASTFIRST Last field lt First field PTNBADFIRSTXY First x y badly specified PTNBADLASTXY Last x y badly specified PTNBADRANGEAREA Bad pattern Range Area PTNCANTFIND Cannot find fields in given Range Area CANTREEXP NOPAT Cannot reexpose as no pattern yet exposed CANTREEXP PRTFLD Cannot reexpose as pattern has part fields EXPMAT BAD REP Expose matrix bad repeat value EXPMAT BAD POS Expose matrix stage move outside limits STGMOVBLK Stage Move Block Transfer Error BADSTAGEBLK Stage move bloc
12. 6 inch wafer 25 3422 14941 6 1 2543 422 76 5 90 5 2 1 19 41 electrostatic 16 5 6 Setting up the holder parameters The positions in the tables above are a rough guide to positions which must be entered for each holder in the file holder table com in vb seq To find the positions for each holder 1 Load the holder 2 Remove any datum offset by typing VB OPER mvhm 3 Type VB OPER mvpo x y Part Number 878275 Vectorbeam Operator Manual Page 152 Where x y is the approximate position of FM 4 Switch on SEM mode and using the joystick position the FM mark in the centre of the screen see Figure 16 3 5 Make a note of the position and enter this into the appropriate position in vb seg holder table com under the entry for the particular holder 6 Repeat points 3 to 5 for GT which should be set to any large bright area of metal see Figure 16 4 7 Repeat points 3 to 5 for the marks DP and FFS if they are required For VB5s as the Faraday cup is on the holder the FC position must be set up and entered in holder table com for each holder For VB6s as the Faraday cup is mounted on the superplate the FC position must be defined once only in holder table com in a position where it will be set whenever a holder is initialised On a VB6 in order to see the FC in SEM mode the video gain must be increased and the SEM magnification decreased to minimum Figure 16 3 Photo showing SEM monitor with image o
13. Beam diameter measurement The system can also give scanning electron transmission images of transparent samples see chapter Creating scanning electron image files Part Number 878275 Vectorbeam Operator Manual 15 15 1 15 2 15 2 1 15 2 1 1 15 2 1 2 15 2 1 3 15 2 2 15 2 2 1 15 2 2 2 Page 131 Machine set ups The machine set ups and calibrations in this chapter will need to be performed the first time a particular beam accelerating voltage is used Only relevant set ups or calibrations will need to be repeated if mechanical work has been carried out on the column or stage such as cleaning or height adjustment If no such work has been carried out then the routines may need to be performed at an interval of a few months Temperature control In order to obtain specified performance levels it is vital that the temperatures of the various machine components such as the column and stage are held within specified limits This is done by maintaining the machine environment to the installation site specifications and by the temperature control baths which circulate water around the machine In order to obtain specified performance levels without long stabilisation times after loading substrates onto the stage it is important that the temperatures of the stage and airlock or Brooks handling system are matched Temperature measurement Location of temperature measurement sensors VB5 There are
14. Page 183 Checking deflection field corrections It is useful to check the field corrections after they have been loaded from a database or after fullcal if there is any doubt about their accuracy In order to check the current field corrections the calibration commands are used with noupdate qualifiers Field focus and stigmation 1 Align final aperture 2 Run jobcal 3 VB_OPER gt JOBCAL 4 Adjust the on axis focus and stigmation as described in the Section Automatic focus and stigmation adjustment in order to define the parameters which will be used by the next step for example the accuracy value 5 Use Emma command VB_OPER gt QCAL STIG FM DIST NOALIGN GRID 5 COVER F TRNLNM VB_COVER DAC POS 3 NOUPDATE DIAG Mainfield distortion 1 Align final aperture 2 Run jobcal VB_OPER gt JOBCAL 3 Use Emma command VB_OPER gt QCAL MAIN FM DIST NOALIG GRID 8 COVER F TRNLNM VB_COVER NOUP DATE DIAG Subfield distortion 1 Align final aperture 2 Run jobcal VB_OPER gt JOBCAL 3 Use Emma command VB_OPER gt QCAL TRAP FM DIST NOALIG GRID 8 COVER F TRNLNM VB_COVER NOUP DATE DIAG Beam error feedback distortion Do not use until further notice Zero any existing corrections using VB_OPER gt QCAL BEF DIST INIT LOAD Part Number 878275 Vectorbeam Operator Manual Page 184 18 8 Fine tuning deflection field corrections It can be useful to fine tune the deflection field correcti
15. Part Number 878275 Page 287 UNLOADERROR Substrate unload error CASSUNDEF Cassette undefined NOCASHOLDER No holder in airlock HOLDERTHERE Airlock already occupied HOLDONSTAGE Holder already on stage NOHOLDSTAGE No holder on stage XFERERROR Substrate transfer error HOLDINAL Holder already in airlock HOLDINFP Holder already in front pouch HOLDINCP Holder already in crane pouch NOHOLDINAL No holder in airlock NOHOLDINFP No holder in front pouch NOHOLDINCP No holder in crane pouch HLDRPOSINIT Cannot initialise holder Centre FM DP not defined HOLDERINIT Cannot initialise this holder DVECALLOC dvector allocation failure IVECALLOC ivector allocation failure DMATRIX1 dmatrix allocation failure 1 DMATRIX2 dmatrix allocation failure 2 SINGMATRIX Singular matrix DWVECALLOC DW transform matrix allocation failure LUMATDEC DW LU matrix decomposition error DWBACKSOLN Invalid DW back transform DWNACBACK Inaccurate back transform DWINVPARAMS Invalid number of exp obs points INVPOSID Invalid position identifier HGTMEASERR Height Measurement Error HGTTBLDNLD Height Table Download Failed HGTPRVRDNG Using previous height reading CALNOUPFORCE Not in Absolute mode NOUPDATE DIAG options DWABSMODINV Absolute mode is invalid for this command STNATMODINV Natural mode is invalid for this command DWCURMODINV Current mode Absolute is
16. Part Number 878275 Vectorbeam Operator Manual 11 4 11 4 1 11 4 2 Page 91 features on a bright background They may also be etched topographical features exhibiting no apparent difference in brightness between the actual mark and the background instead they present a change in signal level as the beam is scanned across the edge of the feature There are three main mark locate algorithms implemented at present and these are known as pit edge and cross The pit algorithm will locate rectangular octagonal and circular marks by examining the difference in the detector signals between inside and outside the mark The edge algorithm will locate rectangular octagonal and circular marks by examining the rate of change of the detector signal when scanning across the mark The cross locate algorithm will locate crosses with limbs made up of one or more fingers Definition of parameters for mark location There are separate commands to define the parameters for the mark locate and to start the mark locate function There are also commands for viewing the list of defined marks deleting marks from the list and showing what parameters have been set There are no boot up defaults for the mark locate parameters or run time defaults however all systems will have a selection of marks in the current database The user must specify all the relevant parameters as in general there is simply no way that the system can guess what sort of mark th
17. Pea KOIKE MEM eu cs tinci xta cta taam EO A oa con E 57 1 9 Etna directory SIF olulB uiae ascen nitepa t spas br Y aaea 60 7 4 Will T 60 8 Joly CORIO OU Temm 1 8 1 Li COIT ma S Traoi AeA AEA 61 8 2 Eiliria Comas uie pudarietesseissebndai vibus desde stu se Huaddsse ptu M Yr aadgrse Pudet pus T 61 8 3 Supplied OU TINGE P 61 8 4 Job tile Creati eT O 61 8 5 JOD TINS RANTS ears A era cess E Ata vh ces A E seb ideals cdi edP enis 62 8 6 EuhnnaggebBlls Suas mias tior pa Doa pon Pod rrnee enn SUC ery rrr trees 62 8 7 rc reAeej pU 62 Part number 878275 Vectorbeam Series Vectorbeam Operator Manual Page iii 874 Buting DCL commatid executio itc reete tte qe eese 62 8 72 During Emma command execution te eco te te dene 62 8 7 3 Notes on GrRLE GO and GTRE Y soi ic ctt ptt Eti rens 62 8 8 opua c aaa a N A 63 8 8 1 Recommended tasks to avoid user hang ups ssesssss 63 6 89 2 lhecomlthiendatloni oiii ied a qoo asad 64 8 9 Bowel PET 64 8 9 1 UJsefullogicals cob ppt E Fest RR eR 64 810 CRANE compiled JODES niriana data ek bod x Kader a a ek ira acp da incon 64 9 LOr sb LOB uui cia PME NIMM qp dde 9 1 Corrections Tor absolute SOOHU BO o iacet ite a iind kde toss tu SN DIA CAD Ra dO a FREUE 67 9 1 1 Stage mapping absolute mode ssssssseses 67 9 1 2 Beam error feedback BEF co
18. Pi2 Pe1 Pe2 Logicals Emma defines the logicals VB_VAC_GAUGE_n with the pressure in Torr where nis a value between 0 and 7 Part Number 878275 Vectorbeam Operator Manual Page 37 4 Computer system Three levels of computer systems are normally used with the Vectorbeam system the pattern data preparation computer the operator terminal and the computer controlled subsystems These computers are all connected via Ethernet Figure 4 1 shows the system without the pattern data preparation computer SPER MAIN ETHERNET CONTROL DESK J LJ LJ ALPHA umm SYSTEM CONTRO LOCAL STORAGE ETHERNET BRIDGE IP CRATE EO CRATE ccu LOCAL ETHERNET PATTERN MANIPULATION SYSTEM PLINTH INTERLOCK CONTROL SYSTEM 1 CERI STAGE VME PICS MOTOR DRIVE Figure 4 1 Computer system 4 1 Pattern data preparation computer The pattern data preparation computer is normally a PC running Linux Previously it was an Alpha station running Open VMS The CATS and CAPROX fracturing software runs on this system Output pattern files are downloaded via the ethernet to the operator terminal 4 2 Operator terminal An Alpha workstation from HP Compaq is supplied with the machine as the operator terminal Virtually all control of the Vectorbeam system is exercised via this
19. S ubstrate Figure 9 2 Effect of changing height on fieldsize At stage position A the fieldsize F1 At stage position B the fieldsize F2 which is larger due to the angle of the beam and the larger height On UHR systems there is also a change in field rotation The height corrections are the adjustments made to the focus point and the field scaling and rotation UHR only to compensate for their dependence on the height of the substrate see Figure 9 1 The height corrections are calculated from the difference in the height meter readings on the calibration mark and on the substrate The fine focus setting is adjusted so that the beam remains focused on the surface of the substrate However when the fine focus setting is varied the rotation of the deflection field changes and this effect must also be nulled The BEF sub field and main field deflection calibrations are adjusted so that deflection distances remain accurate The error in the deflection calibrations due to a micron of height change is about 20 nm at the edge of a 1 mm field Calibrating the main field deflection using a QCAL MAIN command resets the reference values for the height corrections The height meter reading and the fine focus setting are taken before calibration of the deflection and saved Therefore it is vital that the focus is correct using the automatic focus and stigmation facility at this height before calibrating the deflection The height corr
20. The OffsetX and the OffsetY parameters are the X and Y distances respectively between the centre of a die and the lower left corner of the lower left mark CDCX CDCY The CDCX and CDCY parameters are the distances between the die marks in the X and Y directions respectively Pattern data preparation See Section Pattern data preparation Part Number 878275 Vectorbeam Operator Manual Page 238 20 5 Job preparation 20 5 1 Layout parameter file set up In addition to the parameters for non aligned exposures described in Section Layout parameter file set up further parameters are required for exposures with direct write alignment and these are described below Create a file which contains all the DCL statements listed in the parameter file in the previous chapter followed also by all those below This is most easily done by copying the template file vo jobs layout_dw_parameters_template com to a suitable file name for the exposure Edit the file and change the parameters to those required Global_alignment Defines the global alignment method 0 for no global alignment 1 for manual global alignment 2 to select wafer mode previous global alignment which is useful when several layouts on the same substrate are exposed Global_PitchInX 15 30 Defines the pitch in mm in the x direction of the global cells see Figure 20 2 Global PitchInY 14 35 Defines the pitch in mm in the y direction of the gl
21. The cooling water is kept at constant temperature in a closed system and is used to stabilise the temperature of e the electron optical lenses e the reference circuits in the electronics rack e the stage suspension e the turbo vacuum pumps On systems with HR final lenses without Brooks handling three circulator temperature control chiller baths are provided for this purpose An extra bath is added with Brooks handling and a further bath is added for the UHR final lens 1 8 1 VB5 1 8 2 VB6 The temperature control system for VB6 HR without Brooks handling is shown in Figure 1 6 Son a n oa Figure 1 6 Temperature control system for VB6 with 10 chuck airlock Bath 1 controls the temperature of the chamber lid the column and the laser The flow meter is mounted on the crane side of the plinth and the three circuits should be balanced to give equal flow of between 0 7 and 1 3 l min Part Number 878275 Vectorbeam Operator Manual 1 9 1 10 1 10 1 Page 13 Bath 2 controls the temperature of the airlock substrate handler side pouches chamber floor and turbo pumps The flow meter is mounted on the front side of the plinth and the three circuits should be balanced to give equal flow between 1 and 1 3 l min Bath 3 controls the temperature of the pattern generator analogue crate The flow meter is mounted on the front side of control electronics rack 2 CER2 Bath 4 controls the temperature of
22. This method should be used with caution on machines with TFE cathodes Emma help Adjustment Save Figure 7 10 Set lens adjustment 7 2 2 Display menu Also available from the menu in the STATUS WINDOW is Display information about a range of parameters that are currently loaded Select with the mouse left button Display from this menu keeping the mouse button pressed down a list of parameters will be displayed select the subject required then release the mouse button After approximately 30 seconds the appropriate panel will appear These panels can all be cancelled by selecting Cancel at the bottom The Display Column panel Figure 7 11 displays the drive levels to various beam forming elements and provides links to the Set panels for adjustment Part Number 878275 Vectorbeam Operator Manual Page 55 Emma help Status Column Settings EHT 0 00 Filament 0 00 Wehnelt 0 00 Gun Aligner Suppressor 0 00 Extractor 0 00 Tilt X 0 000 Tit Y 0 000 Shift X 0 000 Shift Y 0 000 Lens Adjustment Stigmator Focus Fine Focus Stig Axial Stig Diag Video Gain 0 000 Video Level Backoff 0 000 Filter 0 Done Cancel Figure 7 11 Display Column The Display Database Column panel Figure 7 12 displays the drive levels to various beam forming elements and these are part of the current database Gun Voltage Filament Wehnelt Suppressor Extractor Tit x
23. To select the bias type VB_OPER gt QSET DETECTOR TR BIAS The preamplifier has 4 channels of which channels 1 and 4 are used Each channel has four attenuation ranges Channel 1 10X 30X 100X 300X Channel 4 100X 300X 1000X and 3000X The desired attenuation can be selected by selecting the appropriate channel and attenuation VB OPER QSET DETECTOR TR T1 T4 BIAS ATTEN MIN LOW HIGH MAX If itis not possible to select detector T4 using this command this might be due to a configuration error in the OA config vw file In order to be able to select t1 and t4 there should be a line modint 2 316 9 This shows which transmission detectors are available The final value in modint 2 316 x is a bit significant value with T1 1 T2 2 T3 4 T4 8 To indicate more than one detector sum their values The channel and attenuation should be selected to avoid saturation in the amplifier chain The maximum beam currents that can be used are given below Part Number 878275 Vectorbeam Operator Manual Page 129 Attenuation Max beam Max beam current at 50 kV current at 100 kV The tranmission detector signal has its own gain and backoff levels Usually the gain is set up towards 1 0 and the backoff must be adjusted to around the 0 5 level The substrate holders are designed so that a knife edge or transmission sample can be positioned over the detector This allows spot size measurements to be made see Section
24. Y SCAN pre X SCAN ELECTRON OPTICS COLUMN CONTROL CRATE UNIT PGA Cooling COLUMN CONTROL UNIT PSU E o jj MM A E E NIE 92 amp g z z 07 Qon ajv o m zz z zo al z aaa aZe x Qa 9o o piminiz O z3 2 2 adarr O 6 8181s q eal 80 x amp 0 Figure 1 3 Control electronics rack 2 Electronics cabinets The cabinets satisfy the following requirements e They accommodate standard sub racks 19 inch e Airflow is provided to prevent temperature rise in the cabinets e Standard mains distribution with facilities for filtering and mains voltage stabilisation are available the latter is optional Cabinet A Cabinet A contains the On axis controller the laser height meter electronics the image processor IP crate the stage controller crate the stage motor drive controller and the interferometer electronics See Figure 1 4 Cabinet B Cabinet B contains the pattern generator digital crate PGD See Figure 1 4 Part Number 878275 Vectorbeam Operator Manual CABINET E MAAK LOCATE Page 10 CABINET A Scale VME CRATE EL o Ane E x v STAGE CONTROL ST4ace Oc LLISracE Jonne wonne T4 OPE e ORME E ORNWE Pou M gt gt p EZE FAM TRAY IM SDISLE PSU Figure 1 4 Electronic cabinets A and B 1 5 Stand alone EHT cabinet The stand alone EHT SAEHT cabinet Figure 1 5 contain all the power supplies for the
25. and Global OffsetX and Global OffsetY are positive the indices of the global alignment marks are as shown in the diagram The order in which the marks are visited is global cell 1 followed by global cell 2 followed by global cell 3 followed by global cell 4 A good set of marks would be 1 1 1 212 223 and 124 as they are all outside the area to be aligned and exposed Part Number 878275 Vectorbeam Operator Manual Page 235 global cell 4 123 global cell 3 124 Substrate Global alignment mark global cell 2 121 Layout of dies Layout of 2 x 2 global cells Centre of global cell 1 1 global cell 1 2 12 Optional reference feature Referenceoffset GlobalPitchx Figure 20 2 Diagram showing the meaning of global alignment mark layout parameters The mark indices are correct assuming that Global Offset X is negative Global CDCX is positive GlobalPitchX is positive and GlobalPitchY is positive 20 3 1 1 Global cell A global cell refers to one of the rectangular areas on the substrate defined by the global layout parameters The parameters defining the layout of global cells can often be chosen in several different ways so that they contain some of the global alignment marks on the substrate The global cells layout should be chosen to make the required global marks on the substrate available to the layout command The global cell layout may be an array across the substrate Figure 20 2 or just one cell Figur
26. corrections gain rotation keystone to be summed into the main deflection signal Such corrections are generated digitally prior to conversion to the analogue scan signals The subfield deflection system positions the beam within the subfield and also deflects the beam at high speed during the writing of pattern shapes The pattern generator provides these signals Data to correct the position of the beam within the subfield is also generated by the pattern generator system Imaging The object under the beam is imaged by scanning the beam and varying the image brightness with a detector signal Normally the BSE detector signal is used as the image signal The BSE detector on HR machines consists of 4 scintillators and photomultipliers and on UHR machines a PN junction The magnification of the SEM image may be varied corresponding to changing the amplitude of the scanning signal to the main deflector Image averaging and further video functions are also available The image is displayed on the video monitor Part Number 878275 Vectorbeam Operator Manual d 3 1 3 2 Page 23 Vacuum system Conversion of different pressure units As an aid to comparing pressures reported using different pressure units 1 Inthe British system pressure is usually measured in pounds per square inch PSI 2 In international usage pressure is usually measured in kilograms per square centimetres or in atmospheres 3 In the international
27. vi All Fast focus and Lens C3 magnetic Am in clemantc Substrate Figure 2 2 UHR column optics 2 1 Beam production gun The electron beam used to write the pattern is generated in an electron gun by a cathode emission process This is a thermally assisted field emission source TFE Electrode structures are used to control the emission process and form the beam The electrons are then accelerated to the operating beam energy by an electric field and pass through the anode of the electron gun assembly The cathode used for TFE in the Vectorbeam is heated ZrO coated tungsten An electrode known as the suppressor surrounds the cathode apart from the cathode tip to suppress emission from the cathode shank A high electric field strength is produced between the tip and an electrode known as the extractor Part Number 878275 Vectorbeam Operator Manual 2 2 2 2 1 2 2 2 2 2 3 2 2 4 Page 19 Electrons which are thermally exited to higher energy levels within the cathode may then be accelerated out of the cathode surface This is called Schottky emission and results in a virtual source of a few nm in size The electrons are accelerated towards and pass through the extractor A further electrode between the extractor and the anode used for focusing the beam The beam is then finally accelerated towards the anode The gun can focus the beam over a range of distances below the anode The high tension supply delivers 20
28. 0801AFCA SVDDIVZHY Division by zero in SVD matrix arithmetic hy 0801AFD2 SVDDIVZX Division by zero in SVD matrix arithmetic x 0801AFDA SVDDIVZZ Division by zero in SVD matrix arithmetic z 0801AFE2 LMP2FEWDATA Can t fit polynomial to data too few good points 0801AFEA LMPBADDWMAP The locates at all 4 corners must succeed but didn t 0801AFF2 LMPBADORDER Bad lens map order found when calculating polynomial 0801AFFA BADLENSID qCal Lens lens map number is out of range 0801B002 BADLENSSTEP qCal Lens Y step size is negative or out of range 0801B00A BADLENSGRID qCal Lens Grid size is too large 0801B012 BADLENSGRY qCal Lens Grid size in Y is too large 0801B01A LENSGRDSMAL qCal Lens Grid size too small order 1 is minimum 0801B022 LENSYGRDSMA qCal Lens Y Grid size too small order 1 is minimum 0801B02A BADLENSORGX qCal Lens Origin offset in X is out of range 0801B032 BADLENSORGY qCal Lens Origin offset in Y is out of range 0801B03A BADLENSSEND Transmission of lens map coefficients to PG failed 0801B042 BADLMDATSND Transmission of lens map assignment to PG failed 0801B04A BADLENSORD Lens map order specified is out of range 0801B052 PGCFGCFAIL PG configuration check failed 0801B05A PGCFGBAD PG setup is inconsistent bad mix of processor types 0801B062 PGCFGNOTOK PG unusable processors are not OK or missing reboot 0801B06A PGCFGNLOGS PG setup no of S
29. 11 Pattern generator Grid shapplilg s cea urit iet taret tiende 211 19 1 12 CHOOSING thie pattermlimite acu cci euren tonais 213 19 1 19 N gatvebIasifg scurccceteac esee teta see etur e ex tiras eret bnc AEEA 214 195 74 XEXalnpleSssnesesca dte tescud cities eae tb dap eet de siet 214 18 2 Di itremoval USING Omme ce riore Hirda ti precii ninian PRI esl dedo PHI uta YO 215 19 2 4 ONIME TUR CHOE iis occi orato erene HET EFE EE Et usseceesceceagenepantens ens 216 Part number 878275 Vectorbeam Series Vectorbeam Operator Manual 20 21 22 Part number 878275 19 3 19 4 18 5 19 6 Page viii 19 2 2 Resetting the datum co ordinate system origin 216 19 2 9 JEXample JoDIlIo tette th ie R 216 19 224 JExample comnallu scii e n iter Oe 218 d wefo och mM UEEE 218 19 3 1 Layout parameter file set up 218 19 3 2 Check that the required writing frequency is within range 221 Machine preparo E E o o OE ES 222 19 4 1 Machine parameter configuration eese 222 19 42 Load holder on Stage iui etta siete xps dba oe abdo xn net 222 19 4 9 Gonfidence cheeks uec sioe nete dba ioa see exer Enc O x Res S nMAM aues 222 mrisscq idet T dite is 223 195 12 Intermptilg eXpOSUl e ua sioe dior Eee abere E a deudas ies 223 19 52 WIVGICONM orsina ERE LER Eod RE EN TSHRE NU edie ER SE Aa 223 PSR UNS ici ricis dipin Lid ad s
30. 12 2 2 Total database The total database contains everything by definition 090909020 12 2 2 1 Parameters selected pattern name pattern file size Number of pattern fields X Y block size pattern resolution X Y pattern limits VRU Current X Y position in natural coordinates X Y stage load position X Y datum offset Y mainfield pivot point Y subfield pivot point stigmation wobble status invert mode calibrated height map working height map height measurement mode datum height X Y datum position X Y current position in direct write mode The parameters mapping order X Y origins X Y offsets X Y scales X Y rotations X Y keystones for each of the direct write modes absolute relative wafer die general chip usr1 usr2 usr3 usr4 22 holder ID 23 holder name 24 substrate type 25 substrate size 26 platesize 27 centre position 28 FM focus mark position 29 DP datum mark position 30 FFS fine focus mark position 31 GT gun target position 32 FC Faraday cup position 33 KE knife edge position LD CAANDMNR WN a N NR RR e e RR e lA OOmMANIANPWNKY OS 34 spare position 35 MM spare position 12 2 2 2 Partial databases 12 2 3 Top level partial databases 12 2 3 1 EO The EO partial database contains 12 2 3 1 1 Parameters 1 All current values of the parameters as listed for the coarse partial Part Number 878275 Vectorbeam
31. 2 Check that the required writing frequency is within range Check that the selected dose beamstep and beam current result in a pattern generator frequency within the machine s range and adjust the conditions if necessary Type cfreq in the job control window VB_OPER gt CFREQ Respond to questions VB5 and VB6 exposure conditions calculator Part Number 878275 Vectorbeam Operator Manual 19 4 19 4 1 19 4 2 19 4 3 19 4 3 1 19 4 3 2 19 4 3 3 19 4 3 4 Page 222 Version V01 01 date 8 Jan 2001 Copyright c Leica Microsystems Lithography Ltd Cambridge UK Enter required calculation F requency C urrent D ose or B eamstep f Enter required dose type A rea L ine or P oint a Enter beam current nA 1 Enter area dose uC cm2 400 Enter beam step size nm 25 Frequency 400 000 kHz Machine preparation Machine parameter configuration Load all the required machine parameters See Section Database selection prior to exposing in Chapter Databases Load holder on stage See Section Holder loading unloading for description of how to load the holder containing the substrate on the stage Confidence checks Substrate height map check Although this check is run automatically prior to exposure by wlvd com it may be useful to run it so that errors can be corrected See Section Height map measurement for details Beam current check Set up beam for exposure or load all the
32. 4 1 26 QSET DOSE REL 5 1 30 QSET DOSE REL 6 1 33 QSET DOSE REL 7 1 35 QSET DOSE REL 8 1 38 QSET DOSE REL 9 1 42 QSET DOSE REL 10 1 45 QSET DOSE REL 11 1 48 QSET DOSE REL 12 1 52 QSET DOSE REL 13 1 57 QSET DOSE REL 14 1 65 QSET DOSE REL 15 31 2 00 18 bit and 20 bit machines running V2005 01 or later can define the VRU to be associated with a particular clock number and override the default set by qset VRU e g QSET DOSE REL 0 1 2 VRU 2 Exposing proximity corrected patterns The jobfile WLVD COM allows the exposure of proximity corrected patterns using either of the alternatives described above see Section wivd com in Chapter Exposing a substrate Dose distribution transferred by clock file If a clock file with the same name as the pattern file but with extension clk is in the pattern file directory it will be run before pattern exposure This sets up the required clocks Dose distribution transferred by pattern file If the pattern file contains relative doses then the dose controller is set up automatically on selecting the pattern Part Number 878275 Vectorbeam Operator Manual 14 14 1 14 2 14 3 Part Number 878275 Page 127 Detectors The detectors provide information about the electron beam beam current back scattered electron signal etc This information is used among other things for automatic calibration purposes mark locate autofocus stigmation or for p
33. 77 10 6 4 Setting the height meter offset ececcec cesses eeeeeeceeeeeeaeteeneeesneees 77 10 7 Fine tuning the height dependent field scaling sesssessss 79 10 8 Fine tuning the height dependent field rotation esses 80 10 9 Heig tmeter early tead rosenn 82 10 10 Height map readmgs by JODIE aico eia 82 TOC aE EE o AE E E E T E A date end ia Uc EDT EE E 82 107115 Reakimemode rissin d iei bo nee danaa 82 40 14 2 Heightmap Mode errien eaei ead 82 10I Heini map CallbEAll oll o oe c ord Ter A terre ADAT 82 10 13 Mini height map 18 bit and 20 bit pattern generators only 83 10 19 1 Height identifiGre eiie terroir eot data ees 83 10 13 2 Creating a mini height map ciiita orco chevy 84 10 13 3 Mini height map use with standard layout parameter file template 85 1L Me Ce o rcino Mil Riva or TINIE ai sie allgudainds da aa AEA ALKENES 87 Part number 878275 Vectorbeam Series Vectorbeam Operator Manual 11 2 Designing alignment marks for direct write registration ssss 88 01 271 MaK YPE aont rebel ete t C TEE TENERE d 88 Taz VEI IA ITT 88 11 23 Maik Position sse detti tee c m tei HI NER ERE RS 89 11 24 Marte COMMU assunti rt ERE TREE fas 90 DECEM lod eer eii ET 90 11 4 Definition of parameters for mark location 91 TAT Syaa R IET 91 11 4 2 Choosing geometry parameters
34. 8 11 8 1 Expected mark centre position to start spiral search Page 96 To obtain a diagnostic plot from a mark locate function use the qualifier adc with the mark locate command The video data sampled during the mark locate scans are written to the file vb dat adc dump dat and may be used directly or plotted by a PVWave routine To use the PVWave plot routine set the default directory to vb wave and type wave r adc plot Pit locate algorithm The pit locate algorithm is useful for marks which produce detector signals as shown on the right side of Figure 11 2 The pit locate algorithm is separated into the spiral search bisection search and fine search The following diagram shows how the algorithm switches between these locate fine 7 7 7 7 7 7 I i B 1 Qaem aou em Y uus locate o spiral bisection fine success il 4 fail Figure 11 5 Pit mark locate algorithm search flow diagram Spiral search A spiral search is used to locate marks roughly as quickly as possible The spiral search examines points along a spiral path separated by a distance equal to 65 of the marks smallest dimension Each point is sampled from 1 to 255 times depending on the mark definition OVSAM parameter After sampling each point the video level is compared with the background video level and if there is sufficient contrast CT parameter t
35. EHT ramping up EOERROR EO Control error message in console window EOFAULT EO Control fault message in console window EOGLBSECT EO global data section is not available EOVXERR EO Control VxWorks error message in console window EOWARN Warning message from EO Control in console window ETHDEADCI Subsystem s Command Interpreter process suspended ETHEMMASUPCMD A slave processor received an Emma command ETHHOOKADD Subsystem couldn t set up Ethernet packet intercept ETHIFUNIT Subsystem Ethernet ifunit failure ETHIPQCRE Subsystem couldn t create Ethernet input queue ETHIPQFAIL Subsystem couldn t queue Ethernet message for input ETHIPQOVF Subsystem rejected command its i p queue would overflow Reset subsystem and retry ETHLONGMSG Subsystem long Ethernet message send failure ETHOPQCRE Subsystem couldn t create Ethernet output queue ETHOPQFAIL Subsystem couldn t queue Ethernet message for output ETHOPSEND Subsystem couldn t pass message to Ethernet output ETHOPSPAWN Subsystem couldn t spawn Ethernet output process EXPMAT BAD POS Expose matrix stage move outside limits EXPMAT BAD REP Expose matrix bad repeat value FADDR Pattern file offset address invalid FILBLOWN Filament blown Check the filament power supply Check the contact of the gun cable in the gun socket Check the continuity of the filament by probing the contacts in the gun before opening On FEG machines
36. Emma Status Window 19 5 1 2 Stop The job file commands can be interrupted by CTRL C first and then the pattern exposure can be interrupted by pressing abort in the Emma Status Window Continue must then be pressed to allow Emma to continue 19 5 2 Wlvd com Issues commands to expose the layout defined in the parameter file All Emma commands are written in the full form beginning with q to avoid additional time overheads associated with translating the mnemonics OPEN APPEND SHARE POST EMMA MAILBOX EMMA This line is necessary so that Emma commands can be issued from a jobfile called from another jobfile SET ON Enable DCL error checking ON CONTROL C THEN GOTO FINISH WLVD ON CONTROL Y THEN GOTO FINISH WLVD Define the actions taken on early termination ON ERROR THEN GOTO FINISH WLVD Set up error handling ORIGINALFILE F TRNLNM VB LOGFILE Save existing logfile name before redirecting logging to new file JOB F TRNLNM VB JOBFILE Obtain jobfile name from logical vb jobfile which is set using sjob command JOBFILE F ELEMENT 1 JOB JOBFILE F ELEMENT O0 JOBFILE Extract jobfile name from string containing complete directory and file name Part Number 878275 Vectorbeam Operator Manual Page 224 DEFINE GROUP USER NOLOG VB LOGFILE MCLOG JOBFILE LOG QSET OUT OFF QSET OUT OFF QSET OUT ON Redirect the output to the new logfile with the same name as the job IF LA
37. Emma logfile only contains the Emma commands issued interactively or by a jobfile For full logging of DCL and Emma commands the following notes may be useful DECTerm All Emma commands will be typed up in the DECTerm by defining the ELOG logical as follows VB OPER define group user ELOG ON This can be switched off by typing VB OPER define group user ELOG ON This facility can be used together with the DECTerm logging described below Notes on OpenVMS logging For a full description please see the DEC documentation OpenVMS has 3 ways Part Number 878275 Vectorbeam Operator Manual Page 248 to log the DECTerm output 1 DECTerm logging during batch queue operation 2 DECTerm logging using the SET HOST command 3 DECTerm logging using the Options menu Note that if the SET VERIFY command is issued all commands will be typed up in the DECTerm before being executed and a complete log of everything that has been done can be created This can be cancelled using the SET NOVERIFY command 22 2 44 DECTerm logging during batch queue operation The DECTerm output which is produced by any DCL or Emma commands issued by jobfile being run by the batch queue can be logged The command used to start a job would be VB OPER SUBMIT lt disk directory filename com gt This will submit the job to the batch queue and automatically create a log file with the same name as the jofile but extension log in the vb log dire
38. Error Setting Destination Stage Control Output error Stage Control Input error X Stage Not Moving Y Stage Not Moving Stage Electronics Failure Stage Power Failure X Stage Not Ready Stage No Cooling Air Pressure Stage Not Free To Move Y Stage Not Ready Y Stage Error X Stage Error X Stage Retry Error Y Stage Retry Error X Stage Not At position Y Stage Not At position X Position Error initialisation Y Position Error initialisation Error turning X BEF on Error turning Y BEF on Error turning X BEF off Error turning Y BEF off X Stage at 1st left limit switch X Stage at 2nd left limit switch X Stage at 1st right limit switch X Stage at 2nd right limit switch Y Stage at 1st left limit switch Y Stage at 2nd left limit switch Y Stage at 1st right limit switch Y Stage at 2nd right limit switch Reference Signal Error at X PEO Reference Signal Error at Y PEO Measurement Signal Error at X PEO Measurement Signal Error at Y PEO Position Overflow Error at X PEO Position Overflow Error at Y PEO Stage Velocity Error at X PEO Stage Velocity Error at Y PEO Undefined Error at X PEO Undefined Error at Y PEO Backoff Error Testing X Limits Backoff Error Testing Y Limits Error Finding X Left Limit Error Finding X Right Limit Error Finding Y Left Limit Error Finding Y Right Limit Too Many Retries Moving Stage Yaw Interferometer Measurement Signal Error Yaw Interferometer Position Overflow Error Vectorbeam Operator Manu
39. FRE ENDIF ELSE WFILE WORKFILE Sets the workfile equal to grating keyword ENDIF QSET RESIST STARTDOSE Sets the resist sensitivity to startdose defined in the parameter file IF HOLDER NE 999 THEN VB SEQ HOLDER_TABLE HOLDER If a holder number is specified then intialise the holder IF HOLDER NE 999 THEN VB ACCS ACC_PLATEHEIGHT If a holder number is specified then map the substrate height Part Number 878275 Vectorbeam Operator Manual Page 225 IF BEAM NE 999 THEN IF FSELEMENT 1 beam NES THEN QFILE LOAD MCDB beam F ELEMENT 1 beam QSET MODE FAB RESTORE F ELEMENT 1 beam QFILE LOAD MCDB beam TEST TEST COL ELSE QSET MODE FAB RESTORE TEST QSET MODE FAB ENDIF If a beam is specified then select that beam if the extension is not specified it defaults to test otherwise set fab mode IF CALIBRATE EQS 1 OR BEAM NES 999 THEN QMOVE SPO FM VID_P WAIT 00 00 03 VID_H Do video adjustment until jobcal can cope by itself JOBCAL Run jobcal before exposure if so defined in the parameters file DEFINE GROUP NOLOG USER VB JOBFILE JOB Reset logical which sjob command in jobcal has overwritten ENDIF CLOCKFILE VB PATS WORKFILE CLk Creates a local symbol containing the filename with extension CLK if fSsearch CCLOCKFILE nes THEN QDISPLAY
40. Holder already in crane pouch Choose alternative destination for holder Or first transfer holder from crane pouch to front pouch or library and then retry HOLDINFP Holder already in front pouch Choose alternative destination for holder Or first transfer holder from front pouch to crane pouch or library and then retry HOLDONSTAGE Holder already on stage Choose alternative destination for holder Or first transfer holder from stage to crane pouch front pouch or library and then retry HSTCOMM Host communication not initialised Use incm sys eo or pg or stage to initialise communication HWFLT Hardware fault ILLMAXMIN Illegal max and min values ILLPRONUM Illegal profile number ILLVALUE Illegal parameter value entered Check and adjust parameter Part Number 878275 Vectorbeam Operator Manual Page 278 INVALIDFMT The command was not in a valid format Check the format of the command INVALIDRES Invalid communication response Check the subsystem cabling and power Reset subsystem INVBLKALN Invalid block alignment INVFILHDR Invalid file header INVPOSID Invalid position identifier ITRUPACT Interrupt is already active ITRUPEST Interrupt established for end of current command ITRUPIGN Interrupt ignored no command active IVECALLOC ivector allocation failure LENSGRDSMAL qCal Lens Grid size too small order 1 is minimum Increase grid points LENSYGRDSMA qCa
41. If the beam current has been changed by adjusting lens 1 or if the fine focus value is not correct then lens 2 must be adjusted The lens 1 C1 setup must have already been carried out and is dealt with in the section Beam current adjustment in this chapter The lens 3 C3 setup must have already been carried out and is dealt with in the section Conjugate blanking setup in Chapter Machine setups The final aperture should be aligned and the stigmator balance should be set up before finding the best fine focus and stigmation values Once the beam has been focused using the procedures described here the focus on a substrate during exposure is maintained During jobcal a fine focus setting and a height meter reading are obtained from the calibration mark and are saved as a point of reference by the qcal main command During pattern exposure a height meter reading is taken after every stage move and the fine focus is adjusted based on a known relationship focus for height calibration so that the pattern remains in focus over the entire substrate The substrate surface should normally be within 10 um Manual focus and stigmation adjustment A manual focus and stigmation adjustment is carried out as follows Manual lens 2 C2 adjustment Note that when the lens 2 value has been reduced the current automatically is set to zero then back to the correct value to ensure that the magnetisation of the lens is always the same in spite of t
42. Moving the SHIFTS will have changed the TILT positions for maximum current so this step will have to be done again Vectorbeam Operator Manual 17 1 3 17 1 4 17 2 Page 160 13 Once the maximum current has again been found by ONLY ADJUSTING THE TILTS the upper column should now be aligned 14 It is now just a case of introducing the required final aperture and aligning it as described in the Section Final aperture alignment The procedure above should align the column reasonably well In order to get the ultimate performance the following further setup should be done to minimise the change in aperture position with beam current 1 Find the pairs of lens 1 and lens 2 values which give 1nA and 10nA beams 2 Set the 10nA beam and align the aperture 3 Change to 1nA beam Run the align_apert routine but do not change the aperture position Stop the routine after a few measurements If the above procedure has been followed then the alignment error should be lt 100nm You can adjust the shifts a little to reduce this further 4 lf the X error is negative on going from high to low current then the X SHIFT should be adjusted POSITIVE 5 Ifthe Y error is negative on going from high to low current then the Y SHIFT should be adjusted NEGATIVE 6 Changing the shifts by 0 01 changes the error by about 50nm so the changes are quite small Automatic The Emma command qadjust gun was intended to do an automat
43. One measures the position in line with the point where the electron beam hits the substrate The other measures the position a distance of 0 5 inch away and enables changes in the rotation of the superplate to be measured The pattern generator reads the yaw at each stage position The yaw is zeroed at position 0 0 when a qmove home command is done In an analogous way to the datum offset the yaw is measured at the datum point focus mark by the holder_table com jobfile when a new holder is initialised This yaw datum must be set on the datum mark used for calibration after a qMove Home but before any size alignment calibrations are done The yaw value is saved within the DCP and subtracted from all yaw readings before applying a correction A rotation correction to the mainfield subfield and BEF is made whenever a qadjust field command is issued unless scor off yaw has been selected which is equal to the measured yaw minus the yaw offset stored This yaw correction deals with the limitations of the magnetic map correction The pattern generator will only read the yaw from the stage electronics if the following line is added to the file WFDCP_CONFIG VW only possible if logged in under the emma username G_EnableYawCorrn 1 The pattern generator will use a zero value of yaw unless the yaw reading has been switched on in this way That is unless the command QADJUST YAW DATUM is used to force in a non zero value of the yaw datum
44. Operator Manual o0 9090 amp co Page 107 database 8 sets of runup profile parameters profile number min filament power max filament power min filament current max filament current min filament voltage max filament voltage min wehnelt voltage max wehnelt voltage min wehnelt current max wehnelt current filament stabilisation drive filament stabilisation eht filament stabilisation wehnelt filament stabilisation period EHT at which runup is valid EHT overdrive filament drive step size filament drive step interval time filament normalised drive step filament ramp end drive value EHT drive step size EHT drive step interval time EHT normalised drive step EHT ramp end drive value wehnelt drive step size wehnelt drive step interval time wehnelt normalised drive step wehnelt ramp end drive value suppressor drive step size suppressor drive step interval time suppressor normalised drive step suppressor ramp end drive value extractor drive step size extractor drive step interval time extractor normalised drive step extractor ramp end drive value filament rundown dive step size filament rundown drive step interval time Fine focus Axial stigmator setting Diagonal stigmator setting Axial stigmator balance X Axial stigmator balance Y Diagonal stigmator balance X Diagonal stigmator balance Y 10 Detector gain 11 Detector backoff 12 Detectors in use 12 2 3 1 2 Partial databases 1 2 Th
45. REGISTRATIO N 10UM_SQUARE_COARSE DESCRIPTION STEPPER1 DIAG D EBUG 2 qdisplay comment Finished calibrating stepper lens map F TIME FINISH Return to original log file DEFINE GROUP USER NOLOG VB LOGFILE logfilename SLOF SLON EXIT Checking the calibration The procedure for checking a lens map against the calibration substrate is described in the Emma command set part no 878274 Note that the underlying direct write mapping must be accurate Vectorbeam Operator Manual Page 190 18 10 4 Use of lens maps A lens map is used by assigning it to a direct write mapping using the qmap lens command The overall accuracy when using the lens map depends equally on the direct write alignment accuracy and on the stepper lens map accuracy For this reason it makes more sense to use die by die alignment together with a stepper lens map rather than global alignment with a stepper lens map In any case die by die alignment alone would normally be used rather than global alignment before considering using the stepper lens correction Note 1 Die by die alignment should be done with no lens map assigned 2 The stepper lens calibration done previously used the outermost 4 marks of the measured array to create a private direct write mapping This private mapping was used to eliminate the lower order corrections from the stepper lens map Therefore the positions of the die by die alignment mar
46. Sort strategy CFSP strategy writing strategy not implemented Sleeve width not implemented 12 2 4 2 Pos cal The pos cal partial database contains 1 The stage mapping coefficients scale X scale Y rotation X rotation Y keystone X keystone Y scale linearity X scale linearity Y bow X bow Y for all the stage modes natural absolute machine 2 current stage map mode natural absolute machine 12 2 4 3 Mag cal The mag cal partial database contains 1 EHT 2 The magnetic map calibration 12 2 4 4 Dist cal 1 PGresolution 2 EHT 3 max fieldsize 4 aperture 5 mainfield field corrections as a function of subfield index x subfield index y e XY distortions e XY scales e XY rotations e XY keystones e XY rate of scale change e XY rate of rotation change e XY rate of keystone change 6 subfield field corrections as a function of subfield index x subfield index y e XY scale e XY rotation e XY rate of scale change e XY rate of rotation change 7 beffield corrections as a function of subfield index x subfield index y e XY scale e XY rotation e XY keystone e XY rate of scale change e XY rate of rotation change e XY rate of keystone change 8 date and time of last mainfield field corrections calibration 9 date and time of last subfield field corrections calibration Part Number 878275 Vectorbeam Operator Manual Page 109 10 date and time of last BEF field co
47. The number of die by die marks required across the substrate depends on the accuracy required the accuracy of manufacture of the 1st pattern and the distortions introduced in the substrate since the 1st pattern was manufactured Part Number 878275 Vectorbeam Operator Manual 11 2 4 11 3 Page 90 3 The minimum separation of marks can be as small as about 4 um measured between the marks edges but only if the uncertainty in each mark s position is lt 1m and the fine scan length is 1 um Mark contrast The mark contrast should be maximised in order to minimise the amount of signal averaging required for location and thereby the time For processing reasons however this may not be practical and some compromise must be found In order to maximise the mark contrast it is important to consider the following Material contrast e The mark and substrate should have widely separated Z numbers gt 30 is a good objective although the system can be made to work with less The marks may appear as either bright features on a dark background or dark features on a bright background Both types are located in the same way with the input signal being inverted as necessary Topographical contrast from deep marks e The marks should be as deep as possible anisotropic etch The mark depth must be proportional to the size to obtain this type of contrast and a depth of greater than 1 um per 5 um of width or height is required In this case a
48. When the 18 bit pattern generator exposes a FRE file an 18 bit physical block is produced by combining 4 x 4 16 bit logical blocks The physical blocksize is the size of the rectangle centred on the centre of the deflection field within which the beam may deflected during exposure see Figure 19 2 The logical blocks correspond to the blocks of the standard 16 bit pattern file Emma takes care of combining the appropriate blocks at exposure time Logical blocksize Physical blocksize Maximum fieldsize Figure 19 2 Diagram showing the arrangement of logical blocks and the physical block The logical blocksize is defined in the pattern converter and contained in each pattern file The maximum logical blocksize that can be set in the converter for a FRE file is given by Max logical blocksize in um 2 6 32 x PG resolution Equation 19 8 Expressed in terms of exels Max logical blocksize in exels 255222 Equation 19 9 Part Number 878275 Vectorbeam Operator Manual 19 1 5 3 19 1 5 3 1 19 1 5 3 2 Page 204 This results in Max physical blocksize in um 25 128 x PG resolution Equation 19 10 The reason not all 2 deflection bits can be used for a physical block is because the FRE file format uses unsigned short integers to represent the coordinates of the shapes in outline format Such representation is compact but can only define 2 different numbers As the shapes are defined in outli
49. and 32 respectively The single pass lines are not missing and the other Part Number 878275 Vectorbeam Operator Manual Page 265 features in the pattern are unaffected This could have been done by redefining the width of the lines in the pattern Figure 24 6 Diagram showing the exposed exels for pattern accan125 sl10 00 fre at VRU 2 and erosion nodiscard Figure 24 7 Diagram showing the exposed exels for pattern acc8n125 sl10 00 fre at VRU 32 and erosion nodiscard 24 10 3 None or Noerode option The none or Noerode option causes no shape erosion carried out The pattern outline format is treated as exel format Figure 24 8 shows the exels exposed with the noerode none option and VRU 32 The letters are oversized and have overlapping Part Number 878275 Vectorbeam Operator Manual Figure 24 8 24 11 Page 266 shapes Diagram showing the exposed exels for pattern acc3n125_s 10_00 fre at VRU 32 and erosion none Zero dimension shapes Zero dimension shapes in the FRE or VEP pattern data i e those with zero height or zero width or both will be exposed with a single exel in that dimension Such shapes can be introduced in the pattern file by the converter The lines yes command in Cats will cause paths with zero widths in the input file to be written to the output file as zero width shapes The following is an example piece of Cats ctxt to create a zero width line STRUCT zero width line WID
50. and lens 2 produces the minimum change in the observed image position This corresponds to the beam passing into lens 2 on axis and parallel to the axis Coarse gun alignment HR and UHR final lens Coarse gun alignment is required when no SEM image can be obtained The procedure is as follows 1 Load a holder and move to the datum plate preferably the gun target area VB OPER MVSP GT If the location of the datum plate is not known accurately enough then load the autostitch plate in a holder and move to the centre of the plate 2 Select SEM scanning mode VB OPER SSEM 3 Set Emission Image by clicking in the box on the set gun aligner panel See Figure 7 8 4 Find the beam by changing tilt and shift on the set gun aligner panel See Figure 7 8 5 Turn Emission Image off by clicking in the box on the set gun aligner panel 6 Adjust tilt and shift until the beam is brightest Disable the auto gain and auto contrast on the soft front panel adjust the tilt and shift to increase the brightness of the SEM image enable the auto gain to bring the video level back down Repeat as necessary When it is no longer to possible increase the brightness of the SEM image the alignment is correct DO NOT ALTER THE LENS SETTINGS Fine gun alignment Fine gun alignment may be required if an SEM image can be obtained a beam is already reaching the substrate but the interval since the last adjustment has been long enough that d
51. and the field focus and stigmation calibrations the beam current and magnetic map These job specific set ups are also likely to be used together in different combinations for different jobs e g a 10 nA beam current may be used with several fieldsizes Note that a total database also contains further job specific set ups such as the fieldsize which are adjusted by jobcal or the holder initialisation sequence Since these sequences are always run before any job these can be ignored for the purposes of this scheme Part Number 878275 Vectorbeam Operator Manual 12 5 1 12 5 2 Part Number 878275 Page 112 Recommended scheme 1 For each beam energy 20 50 and 100 kV there should be one basic total database which contains all the set ups and calibrations which remain valid until mechanical work is done on the machine If ever any of these set ups are changed the basic database for the beam energy must be resaved 2 For each fieldsize in use there should be separate partial databases each containing only field correction calibrations 3 For each current in use there should be separate partial databases each containing only column setting or alternatively a spot table database can be used 4 If specific combinations of holders are to be used see below then there should be separate partial databases each containing only magnetic map calibrations After loading the databases jobcal must be run to adjust parameters such a
52. command Delete any unwanted marks from the database using the qmark del command MRKDEFFAIL Mark definition failed MRKEDG Mark edge data insufficient or of poor quality MRKFILT Mark filter parameter out of range Define filter to be in range MRKFSLIM Mark fine search limit MRKFSRCH Mark fine search MRKFULLFAIL Mark locate exhausted full field MRKFXFIT Mark fine x fit MRKFXSRCH Mark fine search x MRKFYFIT Mark fine y fit MRKFYSRCH Mark fine search y MRKHEIGHT Height of feature found does not correspond to param Check using SEM mode that the correct mark is being scanned and that it is undamaged Check that the height width and tolerance mark locate parameters are correct MRKHGTPRM Mark height parameter out of range Part Number 878275 Vectorbeam Operator Manual Page 280 Define the mark height to be within range MRKLEDGE Failed to find left edge MRKLINEAV Mark line averaging parameter out of range Refer to Emma command manual for valid range MRKLMBWID Mark cross limb width parameter out of range MRKLOOK Mark look option unknown MRKMFDAC Mark main field DAC MRKMFLDP Main field position out of range MRKMSRHGT Mark measurement height parameter out of range MRKMSRLEN Mark measurement length inappropriate for mark size MRKMSRWIDTH Mark measurement width parameter out of range MRKNONSPEC Error occured in module not specific marklo
53. computer This computer communicates with the pattern generator electron optics and stage subsystem microprocessors via Ethernet Vectoroeam commands are input at the operator terminal and the appropriate commands are issued to the subsystems by the program running on the Operator Terminal Part Number 878275 Vectorbeam Operator Manual 4 3 4 3 1 4 3 1 1 4 3 1 1 1 Page 38 called Emma During pattern exposure the pattern data are transferred from the operator terminal to the pattern generator via the ethernet under the control of Emma The workstation consists of e Alpha workstation with system and user hard disks colour monitor CD ROM drive tape drive thinwire ethernet interface keyboard mouse and an optional terminal or modem link The specification of each of these components varies from system to system due to a continuous upgrade of machines in production in line with the latest models available e The following software e Two user OpenVMS licence e PV WAVE run time licence e TCPWARE TCP IP run time licence Computer controlled subsystems Pattern generator The pattern generator will be either a 16 18 or 20 bit version 16 bit pattern generators were produced up to about 2000 18 bit up to about 2004 and 20 bit from 2005 onwards 16 bit pattern generator hardware The 16 bit pattern generator consists of a crate of digital electronics and a crate of analogue electronics The digital part of the pat
54. discarded because NOUPDATE set 0801807B DCNEWCURFRQ Current frequency not retained was outside new band Part Number 878275 Vectorbeam Operator Manual Page 284 08018083 PGINFO Successful Message from PG in console window 0801808B SVINFO Successful Message from Stage in console window 08018093 EOINFO Successful Message from EO Control in console 0801809B 080180A3 080180AB 08018FA0 08018FA8 08018FBO 08018FB8 08018FCO 08018FC8 08018FD0 08018FD8 08018FE0 08018FE8 08018FF0 08018FF8 08019000 08019008 08019010 08019018 08019020 08019028 08019030 08019038 08019040 08019048 08019050 08019058 08019060 08019068 08019070 08019078 08019080 08019088 08019090 08019098 080190A0 080190A8 080190B0 080190B8 080190C0 080190C8 080190D0 080190D8 080190E0 080190E8 080190F0 080190F8 08019100 08019108 08019110 08019118 08019120 08019128 08019130 08019138 08019140 08019148 08019150 08019158 08019160 Part Number 878275 window NUPPOSDISP Couldn t update stage position display correctly LMAPDEASS Lens Map s deassigned from changed DW transform s QUALIGNORE Illegal qualifier s ignored ITRUPACT Interrupt is already active ITRUPEST Interrupt established for end of current command ITRUPIGN Interrupt ignored no command active EOGLBSECT EO global data section is not available STGLBSECT Stage global data section is not available PGGLBSECT PG global data section
55. distortion calibration Do not use until further notice Zero any existing corrections using VB_OPER gt QCAL BEF DIST INIT LOAD Jobcal and fullcal errors and warnings Jobcal and fullcal are jobfiles that issue standard Emma commands If an error or warning occurs it is likely to be a standard Emma one There are however a few additional warnings that jobcal or fullcal may issue Additional warnings Warnings are issued 1 If any of the sequence of calibrations is switched off e g warning gun alignment switched off gt No action necessary 2 If logicals which define parameters for calibrations such as vb_calibrate_points are out of range gt The parameter must be changed 3 If the fine focus value is not correct for conjugate blanking gt The conjugate blanking set up should be checked 4 If the datum height is outside of 10 um or datum tilt is gt 1 jum mm gt Action only required if values 3 times these levels 5 The calibration coefficients are gt 10 000 gt No action required unless this was the result of a large sudden change 6 The final aperture is not centred gt Centre the aperture 7 The shift for fine focus calibration was not accurate gt Re run jobcal and then investigate Additional errors An error is issued if the maximum fieldsize parameter is not within the machines range Part Number 878275 Vectorbeam Operator Manual 18 7 18 7 1 18 7 2 18 7 3 18 7 4
56. dmax is to be defined then type VB OPER QSET BAND ABSOLUTE DOSE dmin dmax If a range of line doses dmin to dmax is to be defined then type VB OPER QSET BAND LINE DOSE dmin dmax These commands will calculate and set the lower and upper frequencies Alternatively type VB OPER QSET BAND REL drelmin drelmax Part Number 878275 Vectorbeam Operator Manual 13 2 2 13 2 3 13 2 3 1 13 2 3 2 Page 120 This will set the upper and lower frequencies based on the resist sensitivity and the minimum and maximum relative doses drelmin and drelmax This is useful if the resist sensitivity is subsequently varied as the dose controller band will be adjusted also Update of pattern generator frequencies on exposure When the qexpose pattern command is issued any frequencies for clocks to which a dose has been assigned are recalculated All clock frequencies are compared with those set in the pattern generator and if any differ the pattern generator is updated Note This only happens at the start and if an ontime procedure has remeasured the current during an exposure the operator must see that a qadjust command is issued VMS logicals The following VMS logicals are set in order to allow jobfiles to automate operation VB ACLK STATUS The logical VB ACLK STATUS is set to one of the following strings by the qadjust clock command 1 OK 2 WARNING Clockband too wide worst dose accuracy x xx An actual val
57. do not open gun but contact Vistec Engineer FILDISABLED Filament disabled Part Number 878275 Vectorbeam Operator Manual Page 277 FILEMPTY Pattern file empty FILFAULT Filament hardware fault FILMAPSIZ File mapping error FILOFF Filament off FILRAMPDWN Filament ramping down FILRAMPON Filament ramping on FILRAMPUP Filament ramping up FLDADRERR Pattern field address transfer failed FLDHDRERR Pattern field header transfer failed FLDSIZERR Field sizing error Check that the fieldsize is within range for the EHT set This error will occur if the EHT is zero FLTCNVVAX VME to VAX floating point conversion number too big FLTCNVVME VAX to VME floating point conversion number too small FNCNTAVAIL Function is not implemented FPE Floating point arithmetic error HGTMEASERR Height Measurement Error HGTPRVRDNG Using previous height reading HGTTBLDNLD Height Table Download Failed HLDRPOSINIT Cannot initialise holder Centre FM DP not defined HOLDERINIT Cannot initialise this holder HOLDERTHERE Airlock already occupied Choose alternative destination for holder Or first transfer holder from airlock to pouch or library or vent and remove and then retry HOLDINAL Holder already in airlock Choose alternative destination for holder Or first transfer holder from airlock to pouch or library or vent and remove and then retry HOLDINCP
58. during global alignment MAN requires a manual stage alignment using the joystick before each mark locate AUTO just uses the automatic mark locate at the expected global mark positions REFERENCEOFFSET 20 19 The distance from the centre of the layout to a reference feature which is usually well away from the dies to be exposed to avoid accidental exposure during the initial search on the substrate This is useful for the initial coarse global alignment before the final global alignment is carried out If 999 is used then no reference feature search will be carried out Global_cell_ 1 221 Global_cell_ 2 421 Global cell 3 241 999 for only 2 mark global Global_cell_4 441 999 for only 3 mark global alignment These symbols define the alignment marks used for the global alignment The first 2 symbols must be defined as a minimum of 2 marks are required for global alignment The 3 and 4 marks are optional If 999 is entered for Global_cell_3 then only 2 marks are used for global alignment If 999 is entered for Global_cell_4 then only 3 marks are used for global alignment The first two numbers of each symbol are the X Y indices of the global cells The third number is the alignment mark number within a cell Global cell 1 POS 55 746 32 495 999 for no stored global position Global cell 2 POS 999 999 for no stored global position Global cell 3 POS 999 999 for no stored global positio
59. error UNRECCMND Sub system received but did not recognise command UNSOLINP Unsolicited input on Ethernet link XFERERROR Substrate transfer error If the loader arm on a VB5 has stopped anywhere but in the fully retracted position due to the emergency switch being activated then extreme caution is required A Vistec engineer should be contacted If the cause of the error is clear and not a mechanical fault with the loading mechanism such as the destination position being occupied or the stage not being able to move to the load position due an interferometer error correct the command parameters or clear the error Issue the qdisplay air read command and retry 27 6 5 Emma error numbers 08018009 SUCCESS Command Successful 08018013 EOINIT Initialising EO communication link 0801801B EOLNKACT EO communication link active 08018023 EOLNKNRDY EO communication link is not ready 0801802B STINIT Initialising Stage communication link 08018033 STLNKACT Stage communication link active 0801803B STLNKNRDY Stage communication link is not ready 08018043 PGINIT Initialising PG communication link 0801804B PGLNKACT PG communication link active 08018053 PGLNKNRDY PG communication link is not ready 0801805B LNKALRDYCON Host is already connected to the sub system 08018063 LNKESTB Host to sub system communications link established 0801806B NEWMRK New mark definition 08018073 CALNOUPDATE New data
60. focus the beam decrease the current in lens 2 5 Repeat from step 3 until the difference is within the range described in the Section Correct fine focus value on datum for conjugate blanking Manual fine focus and stigmation adjustment Normally the automatic fine focus and stigmation adjustment will give more accurate results in less time and so should be used instead of this manual adjustment 1 Move a feature such as the calibration mark under the lens and select SEM mode 2 Adjust the sliders for the fine focus and stigmation to obtain the sharpest image If the fine focus is outside the range described in the Section Correct fine focus value on datum for conjugate blanking then lens 2 C2 adjustment must be carried out Automatic focus and stigmation adjustment Automatic lens 2 C2 adjustment Automatic lens 2 adjustment is only possible if the spot table has been calibrated If the beam current or spot size has been changed by using either the QSET CURRENT or QSET DIAMETER commands the lenses 1 and 2 will be set up automatically based on the spot table calibration If the spot table has not been calibrated then use the manual lens 2 adjustment Automatic fine focus adjustment The qcal focus command only uses 1 edge of a mark for adjusting the fine focus and so unless the stigmation is accurate will not give the correct result Therefore use the automatic stigmation adjustment instead as described in the
61. global marks 4 Optional die by die alignment Global alignment Global alignment must be carried out if any alignment is required It may be used as the only alignment before exposure or as the setup in order to be able to use die by die alignment Steps 1 to 3 above carry out the global alignment The uncertainty in the position of the substrate on the holder after a substrate has been loaded will be several tens of microns at best and may be as much as a few mm This uncertainty will be mainly in X and Y as the rotation of the substrate will have been matched to the stage using the alignment microscope Therefore the first step is to find one mark or feature and eliminate this X and Y shift so that subsequent marks can be more easily found If the uncertainty in the positions of any of the mark is more than the maximum coarse search range of the mark locate or if a reference feature is used then manual global alignment must be carried out in SEM mode Optional alignment of reference feature As an option a reference feature may be used as the first global alignment feature This feature does not have to be an alignment mark and must be aligned manually in SEM mode This option is chosen to avoid accidental exposure of device areas during the initial search on the substrate as the reference feature is usually well away from the dies to be exposed Once the reference feature has been found the X and Y substrate insertion errors are eliminate
62. gun Part Number 878275 Vectorbeam Operator Manual 1 6 1 7 Page 11 lt Hotbox containing Filament supply Suppressor supply Extractor supply Lens 1 supply Rack containing from left to right PSU aa Z SAEHT microprocessor B Gun supply controller Gun controller interface Gun supply driver PSU Hotbox controller 9 WISIS E Pi Dee Gul H O0 E 100 kY supply controller 100 kY supply Bertan S Figure 1 5 Stand alone EHT supply Power supplies The stabilised power supplies deliver output voltages independent of mains fluctuations within the mains tolerances The power supplies switch off in case of output current or voltage exceeding specified limits The computer automatically checks the presence of the output voltage of most of the power supplies Operator console Video monitor The video monitor located on the operator console is used mainly to display the SEM image when the machine is in SEM mode The video monitor also displays graphical and other information produced by the video processor electronics Part Number 878275 Vectorbeam Operator Manual Page 12 1 8 Temperature control water baths
63. height meter needs resetting If a QCAL HEIGHT reports an error it should be retried once more if the error persists then the most probable cause is that the holder is positioned where the height sensor will not work If this is the case then the holder should be moved to a position where it is known to work e g the middle of a plate and it should be re calibrated and height reading taken as a check If it still gives errors on calibration then the height sensor should be reset using the button on the height sensor PCB located in the top of Cabinet A and the calibration repeated If it still will not work then connections and power supplies etc should be checked These are the status values which can be reported by the Laser Height Sensor and Part number 878275 Vectorbeam Series Vectorbeam Operator Manual 27 5 Page 270 picked up by the pattern generator At present the pattern writing will continue if any of these faults appear except the stage fault 9 the last measured height being used instead e 0 good reading e 1 poorreading from poor video signal Possible holder shadowing or wrong LHS calibration table being used e g LHS calibrated on a bright focus mark but now measuring a dark anti reflective chrome plate e 2 LHS atlimit Laser gone to max or min brightness Ignore recalibration desirable but not essential e 3 LHS overrange Height too great small or faulty CCD Recalibrate or investigate
64. is desired to set the offsety to be the distance to the centre of the mark instead of the corner CDCX 2 5 The X pitch in mm of the alignment marks within the die see Figure 20 4 CDCY 2 5 The Y pitch in mm of the alignment marks within the die see Figure 20 4 MARKER Mark_10 The name of the mark definition for the alignment marks If the name starts with then it is taken to be the name of a jobfile to be run at each die alignment mark instead of mark locate e g MARKER directory jobfile_1 com RECOVERY ADJUST FIELD continue RECOVERY LOCATE continue RECOVERY MAP continue Defines the error recovery actions Can be set to either continue manual redo skip stop or user 9QJOB NAME VB SEQ WLVD_DW COM optional DCL command or Emma command This line calls the file that carries out the necessary operations to expose the layout as defined If no command is put after the wlvd com the pattern is exposed at each layout cell If an Emma command or DCL command is put after the wlvd com e g vb seq wlvd_dw com qdisplay height or vb seq wlvd_dw com directory expose_both_patterns com this command will be run at each layout cell instead of the qexpose pattern command The file expose both patterns com must have Part Number 878275 Vectorbeam Operator Manual 20 6 20 7 20 8 Page 241 OPEN APPEND SHARE POST EMMA MAILBOX EMMA as the first line and then any usu
65. kV to 100 kV to the cathode of the electron gun In addition it provides the suppressor bias voltage and the filament current Safety interlock circuits inhibit the operation of the gun in poor vacuum or when the gun is mechanically removed from the emission chamber Spot formation Gun aligner The electron beam that emerges from the gun passes through two sets of magnetic deflection coils which can tilt and shift the beam to align it to the electron optical axis of the following lenses Apertures The beam divergence is limited by intermediate apertures at various positions down the column The final aperture is used to define the final beam convergence angle to the substrate On systems with the HR final lens an adjustable final aperture blade positioned within the lens is provided that accommodates four apertures on a VB5 and six apertures on a VB6 On systems with the UHR final lens an adjustable final aperture blade positioned between the 2 and final lens is provided On HR machines and early UHR machines the apertures are selected and adjusted manually using the knobs on the outside of the column On later UHR machines the apertures are selected and adjusted automatically using motors The beam convergence angle affects both the final spot size and the current density of the final spot at the substrate surface A range of different sized apertures enables the spot size and current density to be optimised The minimum spot size achieva
66. lower video level is obtained from the mark centre than from the background Topographical contrast from shallow marks e The mark edge step should be very sharp anisotropic etch and depth of at least 1um ideally although the system can work with less In this case features exhibit no apparent difference in video level between the mark centre and the background instead they present a change in contrast as the beam is scanned across the edges of the feature In addition e The mark must be stable under exposure by the electron beam e Beam scans across the edges should give reasonably sharp transitions from one level to the other e The same form of profile is required in both the X and Y directions Mark locate algorithms The detectability and reproducibility of the measured mark position is a complex subject and apart from the considerations above will also depend upon the beam diameter and current and accelerating voltage used material types and thicknesses and can be degraded by various sources of noise The accuracy of the mark location will be affected among other things by the edge roughness of the mark For these reasons the Vectorbeam mark locate algorithms are provided with a number of user tuning options which with training and experience enables the user to optimise mark detection accuracy by trade off with mark detection speed The marks may appear as either bright features on a dark background or dark
67. lt N gt VB6 Single holder airlock To transfer a substrate from one position to another enter in the job control window VB_OPER gt QSUBSTRATE TRANSFER lt SRC gt lt DEST gt This moves the stage to the load position first and then transfers a substrate between two positions SRC source position DEST destination The positions are Airlock AL Front Pouch FP Crane Pouch CP and the Chamber Stage CH Ten holder airlock The operation is the same as for the single holder airlock except that any transfer involving the airlock must specify an airlock position between 1 and 10 as the third parameter e g VB OPER QSUBSTRATE TRANSFER AL CH 3 Stopping a substrate transfer Control Y and Control C should not be used to abort the qsubstrate transfer SUBX command but rather the ABORT button on the Emma Status Window should be clicked Remember to also click on the CONTINUE button as instructed in the OPER window after clicking on the ABORT button Note that in general these buttons are dimmed but once the qsubsirate transfer SUBX command has been invoked they become highlighted and available for use If a COM file with a qsubstrate transfer SUBX is running and needs to be aborted then the procedure is a Look at the ABORT button and if it is dimmed then qsubstrate transfer SUBX is not currently running in the COM file so Control Y C can be used to abort the COM file b If the ABORT button is highlighted then
68. maximum can be measured directly by scanning the beam across a sharp edge of a feature and noting the distance between the points at which the detector signal is at 12 and at 88 of the maximum A sharp edge means one that has a length significantly smaller than the beam diameter to be measured Some of the sharpest edges which can be made practically are produced by anisotropically etching silicon These may have an edge sharpness of lt 20 nm By comparison the best edge sharpness of the metal calibration marks is around 30 nm If the beam diameter to be measured is similar to the edge sharpness the edge sharpness will make a significant contribution to the measured diameter and the result will not be accurate This problem can be alleviated by making use of the fact that the directly measured diameter is the quadrature sum of the actual diameter and the edge sharpness If the edge sharpness is known the beam diameter can be calculated as follows d 4d de dp where d is the beam diameter d is the measured diameter d is the edge sharpness and d is the noise on the beam position assumed to be negligible here Display diameter The function qdisplay diameter measures the beam diameter by the above method The distance between the 30 and 70 levels is noted and a factor is used to give the diameter corresponding to Full Width Half Maximum The output shows the direct measurement for each individual edge the a
69. metric system SI pressure is usually measured in newtons per square meter also called Pascal with abreviation Pa 4 The pressure exerted by 1 mm Hg has been defined as 1 Torr after the physicist Evangelista Torricelli 5 750 Torr 1000 mbar 100 000 Pa 6 The unit atmosphere atm is defined as a pressure of 1 03323 kg sq cm 14 696 Ib sq in which in terms of the conventional mercury barometer corresponds to 760 mm 29 921 in of mercury abbreviated mm Hg 7 1atm 760 Torr 1013 mbar 101 300 Pa 1 03323 kg sq cm 14 696 Ib sq in Principles of operation Ultra high vacuum is needed in the gun to avoid damage to the cathode and in the column and chamber to minimise the risk of collision of electrons with residual gas molecules The airlock is used to enable the operator to load unload substrates It enables fast substrate exchange by maintaining vacuum in the column gun and main chambers The pressure required in the gun emission chamber is lower than that required elsewhere The pressure in the gun can be maintained at a lower pressure than in the column section below by the use of differential pumping The gun is pumped by an ion pump and isolated from the column section beneath it by an aperture which restricts the amount of gas flow In addition UHV techniques to reduce the amount of gas being emitted into the chamber are used wherever possible These include the use of copper knife edge seals and a bake out cycl
70. minimise subfield stitch errors for some applications RECOVERY_DO continue RECOVERY_DOSE continue RECOVERY MOVE continue RECOVERY PATTERN continue Defines the error recovery actions Can be set to either continue manual redo skip stop or user oGJOB NAME VB SEQ WLVD COM optional DCL command or Emma command This line calls the file that carries out the necessary operations to expose the layout as defined If no command is put after the wlvd com the pattern is exposed at each layout cell If an Emma command or DCL command is put after the wlvd com e g vb seq wlvd com qdisplay height or vb seq wlvd com directory expose_both_patterns com this command will be run at each layout cell instead of the qexpose pattern command The file expose both patterns com must have OPEN APPEND SHARE POST EMMA MAILBOX EMMA as the first line and then any usual Emma and DCI commands Multiple layouts can be exposed by one parameter file by repeating the above sequence in the same file The parameters can be thus redefined and must be followed each time by the line running the file wlvd com The wlvd com file is described below FINISH ALL This is the line to which the software jumps when control c or control y is pressed This line is placed at the end of this file so that if the file contains more than 1 layout all following layouts are not executed and operation stops immediately 19 3
71. movement of the stage to be corrected Part Number 878275 Vectorbeam Operator Manual Page 143 for scaling rotation keystone scaling linearity and bow errors so that the stage matches some other machine The set up uses the same procedure as for the absolute stage map mode set up except that the array of marks must have been produced on the machine to be matched and the errors from the plate are extracted and applied 15 9 5 Select stage map mode The command qMAP STMODE enables the user to select the specified mode as the current mode VB OPER qmap stmode abs load Current stage mode is ABSOLUTE Coefficients for ABSOLUTE mode are X Y SCALE 0 00001 4213917000 0 000014529809000 ROTATION 0 001296107700000 0 001379073780000 X SQUARED 0 000000007655390 0 000000007075946 XY 0 00000000281 3361 0 000000015719697 Y SQUARED 0 000000072221504 0 000000020973950 Part Number 878275 Vectorbeam Operator Manual 16 16 1 16 2 16 2 1 16 2 2 Page 145 Substrate loading and unloading Systems with the Brooks handling option See the manual VB6 PMC Operator Manual 893157 Systems with the single chuck or 10 chuck airlock handling option For systems with the single or 10 chuck handling option substrates must be loaded manually into the holders Use gloves while handling the holder and substrate For mask plates use an earth clip VB5 Attach the appropriate Perspex jig to the base of the substrate exc
72. nm 1 310720 mm 5 nm Depending on the hardware it may be possible to use a larger range as shown below but this is not guaranteed Max fieldsize PG 20 kV 50 kV 100 kV resolution Minimum 0 16384 mm 0 625 0 16384 mm 0 625 0 131072 mm 0 5 nm nm nm 2 930 mm 1 12 nm 1 850 mm 7 06 nm 1 310720 mm 5 nm 19 1 4 1 5 20 bit pattern generator with UHR lens Max fieldsize PG 20 kV 50 kV 100 kV resolution Minimum 0 16384 mm 0 16384 mm 0 131072 mm 0 125 0 15625 nm 0 15625 nm nm Maximum 1 310720 mm 1 25 1 310720 mm 1 25 1 310720 mm 1 25 nm nm nm Part Number 878275 Vectorbeam Operator Manual Page 202 Depending on the hardware it may be possible to use a larger range as shown below but this is not guaranteed Max fieldsize PG 20 kV 50 kV 100 kV resolution Minimum 0 16384 mm 0 16384 mm 0 131072 mm 0 125 0 15625 nm 0 15625 nm nm 2 930 mm 2 79 nm 1 850 mm 1 76 nm 1 310720 mm 1 25 nm 19 1 5 19 1 5 1 Block height and width The block height and width are the dimensions of the rectangle centred on the centre of the deflection field within which the beam may be deflected during exposure see Figure 19 1 Block height Block width Maximum fieldsize Figure 19 1 The block in relation to the maximum fieldsize The height and width of the block are defined in the pattern converter and contained in the pattern file 16 bit pattern generator additional
73. number 878275 Vectorbeam Series Vectorbeam Operator Manual 27 etd 27 2 27 3 27 4 Page 269 Recovery from exception conditions Job control window locks up If the job control locks up but the other DECterms are still operating normally select Exit from the file menu in the job control DECterm Start another DECterm from the Session Manager which should appear with the VB_OPER prompt Enter start to set up the new DECterm Top half of the screen goes black If the top half of the screen goes black usually displaying information about Ethernet communication type control F2 to return to normal VAX Alpha screen locks up If the VAX or alpha screen locks up there is a failure in the VMS window handler To rectify the problem you should log into the VAX remotely Type SHOW SYSTEM to display a list of running processes You need to delete the right process and this will clear the fault VUE VECTORB is the probable cause of the trouble Type STOP VUE VECTORPB or the process that you think is causing the problem and look at the VAXStation screen to see if it is running again Height meter error recovery If any height meter errors are reported in response to a QDISP HEIGHT command then this may be due to one of the following 1 The holder is positioned at a point where the height sensor will not work because no surface is present 2 The height sensor table is not correct for the surface 3 The
74. number of subfields and the maximum blocksize The number of subfields and the maximum blocksize are the same in both X and Y The number of the subfields that span the maximum blocksize must be defined by the operator The size of a subfield bit is always set to be the same size as a mainfield bit The maximum number of subfields is 64 and the default number is also 64 The minimum number of subfields possible is limited by the maximum subfield size and the maximum blocksize that must be spanned The maximum subfield size has two limits Part Number 878275 Vectorbeam Operator Manual 19 1 8 3 Page 209 1 A bits limit of 2 or 16384 bits 2 Asize limit of 20 um at 100 kV 28 um at 50 kV and 28 um at 20 kV For example 32 subfields with a maximum blocksize of 0 8192 mm anda Maximum Fieldsize of 0 8192 mm gives a subfield size of 25 6 um which is within the deflection limit and 262144 32 81 92 bits which is within the bits limit Another example would be 25 subfields with a Maximum Blocksize of 0 500 mm and a Maximum Fieldsize of 0 524288 gives a subfield size of 20 um and 10000 bits The Maximum Blocksize set on the Vectorbeam must be larger or equal to the pattern block height and width otherwise the exposure cannot be carried out There must be an integer number of bits in each subfield and this means that the result of dividing the number of bits in the Maximum Blocksize by the number of subfields must be an integer Both examp
75. on choosing blocksize The blocksize must be within the limits of the hardware The range of physical blocksizes is given in the Sales Specification for the particular machine The Part Number 878275 Vectorbeam Operator Manual Page 207 following additional details should be taken into account when choosing the blocksize for the 20 bit pattern generator 1 The table below shows typical values Smallest 160 um 160 um 128 um 1000 um 1300 um 1300 um 19 1 7 Subfield fracturing The fracturing of the pattern data into subfields is carried out by the pattern generator at exposure time 19 1 7 1 16 bit pattern generator subfield fracturing The subfields span the maximum fieldsize There are 64 x 64 subfields and each subfield contains 1024 bits 1024 64 65536 19 1 7 2 18 bit pattern generator subfield fracturing N Maximum fieldsize Maximum N blocksize Subfields Pattern blocksize Figure 19 4 Diagram showing the arrangement of subfields Maximum Blocksize and the pattern blocksize for the 18 bit pattern generator The subfields span a square area known as the Maximum Blocksize whose size can be set by the operator and which is centred on the field centre The default size of the Maximum Blocksize is the Maximum Fieldsize but the Maximum Blocksize can be set
76. on the MUP VB OPER qset reg 3 53 1 Next use the following command to set the offset 0 3 um for example for the previously selected table VB OPER qset reg 3 54 0 3 10 7 Fine tuning the height dependent field scaling The dependence of field size on height is calculated by the software from the geometry of the deflection system i e the nominal height of the pivot point above the substrate A compensating correction is applied for the difference in height between the substrate and the calibration height by default the height corrections are on However due to inaccuracies in the bench calibration of the heightmeter and non linearities of the height meter system the compensation will not be perfect Therefore further coefficients are available to take out any inaccuracies in the scale for height correction calculation as shown in the table below These coefficients are applied to the height meter reading that is used for the scale for height correction and calibrated on the basis of measurements on the Vectorbeam There are coefficients for a 2 order polynomial for both X and Y allowing independent adjustment These coefficients allow fine tuning of the scale for height and non linear scale for height The constant terms effectively produce a scale offset but it recommended that the calibration offsets are used instead see Section Calibration offsets in Chapter Calibration X Y Nominal value NonlinearGainConstX Nonlinear
77. or stopped LED X Y is lit when the stage is at the load position LED HOLD is lit when a holder is in the current cassette position LED CASS is lit when a cassette is in the airlock LED PRESS red is lit when the compressed air pressure is insufficient LED WATER red is lit when the water level in the thermostatic unit is too low 3 7 2 1 Modifications to FEG vacuum system Recent modifications to the column pumping are Part Number 878275 The gun penning G5 has been removed The valves V6 and V9 and the nitrogen admit have been removed IGP3 is connected directly to the column with no branches The gun and the IGP4 are connected to the chamber with valves V17 and V18 for isolation respectively The airlock vent button has been disconnected The airlock full vent must be used Vectorbeam Operator Manual Page 28 3 7 3 VB6 FEG with 2 roughing pumps vacuum control panel Figure 3 2 VB6 TFE vacuum control panel Part Number 878275 Vectorbeam Operator Manual Page 29 3 7 4 VB6 FEG with 1 roughing pump and no bypass vacuum control panel VACUUM IGPe pa Nar IGP1 CONTROL WARNINGS TRANSPORTER e COMPRESSED AIR o COLUMN WATER SYSTEM WATER BAKEDUT COOLING AIR o SYSTEM SYSTEM PUMP VENT e RESET Figure 3 3 VB6 FEG with 1 roughing pump and no bypass vacuum control panel Part Number 878275 Vectorbeam Operator Manual Page 30 3 7 5 VB6 FEG with 1 roughing pump vacuum con
78. order to read and plot the data a PV WAVE compiled procedure must be run as follows See also jobfile documentation manual number 892777 for more detailed information VB SUPER wave r temp monitor A postscript plot vb disk vb resultstemp monitor dd mmm ps is produced Temperature set up Ensure that the clean room environmental temperature is stable and set to the desired temperature usually 22 0C VB5 No procedure yet available VB6 Use the background temperature monitoring to assess the control and settings continuously for a period of 24 hours If there is significant temperature variation of the superplate X stage table Y stage table airlock or ring then the stage temperature control system and the temperature controller set up should be checked Matching temperatures requires storing the holder and substrate at the same temperature as the airlock stage the substrate and its holder does not change significantly whether the holder is left in the airlock or on the stage Changing the beam accelerating voltage kV The VB can be operated at either 20 50 or 100 kV and the choice of beam accelerating voltage depends on the application Choosing the beam accelerating voltage Note 1 At higher beam accelerating voltages the current density is higher but the resist sensitivity is lower These effects cancel each out to the point that no significant change in beam on time is obtained at different beam accelerating volt
79. other types of substrates The holder with substrate is placed in the airlock and before loading the holder on the stage the airlock must be evacuated The holder is then put onto the stage by a substrate handler mechanism Prior to removing a holder out of the airlock it is vented to atmospheric pressure with dry filtered nitrogen rather than air to reduce pump down times and avoid contamination For systems with the Brooks handling option wafers in standard cassettes are placed manually onto the load stations The wafers are taken by the first wafer handler under automatic control and placed onto a prealigner inside an airlock After evacuation of the airlock and prealignment the wafers are taken by a second wafer handler and placed into a holder on a loading station The second handler then picks the holder from the loading station and places in onto the stage Each holder is provided with a target mark array which is used in several system calibration routines Substrates are held against 3 front face reference points which are preset to the same height as the target mark array The substrate can be height mapped in order to relate beam deflection accuracy to the plane of the substrate VB5 There are sapphire discs on the underside of the holders which sit on sapphire discs on the stage In addition there are sapphire discs on the side of the holders which are pressed against sapphire discs mounted on the stage by springs On VB5 mac
80. parameters to those defined in the parameter file IF DROPOUTS EQ 0 THEN GOTO ENDDROPOUT Define the dropout and plugin cells in sequence if any LAYOUTDROPOUT COUNT 1 LAYOUTDROPOUT LOOP DROPOUTCELLS DROPOUTCELLS COUNT Creates local symbol containing the name of the global symbol defined in the parameter file LAYOUT DROPOUT DROPOUTCELLS Issues command to define the dropouts to be those contained by the global symbol in the parameter file IF COUNT LE PLUGIN THEN PLUGINFILE PLUGINWORKFILE COUNT IF PLUGINFILE NES THEN IF FSSEARCH VB PATS PLUGINFILE FRE EQS THEN QDISPLAY COMMENT ERROR CANNOT FIND VB PATS PLUGINFILE FRE GOTO ERROR_WLVD ENDIF LAYOUT PLUGIN VB PATS PLUGINFILE FRE ENDIF ENDIF IF COUNT EQ DROPOUTS THEN GOTO ENDDROPOUT COUNT COUNT 1 GOTO LAYOUTDROPOUT LOOP ENDDROPOUT Defines any plugins as those cells just dropped out using layout DISPLAY 0 32 DISPLAY DISPLAY F EDIT DISPLAY COLLAPSE IF DISPLAY NES THEN LAYOUT DISPLAY DISPLAY Part Number 878275 Vectorbeam Operator Manual Page 227 RECOVERY DO 0 32 RECOVERY DO RECOVERY DO F EDIT RECOVERY_DO COLLAPSE IF RECOVERY DO NES THEN LAYOUT RECOVERY DO RECOVERY DO RECOVERY DOSE 0 32 RECOVERY DOSE RECOVERY DOSE F EDIT RECOVERY_DOSE COLLAPSE IF RECOVERY DOSE NES THEN LAYO
81. pattern generators To display the current status type gt showTableOffsets Offsets added to Height Sensor output Table no Offset microns Mechanism switched off Currently using table 7 with offset 0 000 value 43 0x2b To set an offset type gt setTableOffset 1 0 3 value 0 0x0 The following is also required to set the mechanism on gt G_TableOffsetsOn 1 _G TableOffsetsOn Oxfab4c8 value 1 0x1 This will give the following gt showTableOffsets Offsets added to Height Sensor output Table no Offset microns Currently using table 7 with offset 0 000 value 43 0x2b 10 6 4 2 Height meter offset using qset register Using the qset register command enables automation via jobfiles 10 6 4 2 1 16 bit pattern generator The height meter offset can be set using Emma commands First use the following command to select heightmeter table 1 for example and to switch the Part Number 878275 Vectorbeam Operator Manual Page 79 offsets on This sets G_TableOffsets 1 on the MUP VB_OPER gt aset reg 4 19 1 Next use the following command to set the offset 0 3 um for example for the table which was previously selected VB OPER Qset reg 4 20 0 3 10 6 4 2 2 18 bit and 20 bit pattern generators The height meter offset can be set using Emma commands First use the following command to select heightmeter table 1 for example and to switch the offsets on This sets G_TableOffsets 1
82. plate has holes at regular intervals into which substrate clips are screwed The height of the plate is set using three grub screws there is no front face reference Normally when the sample has been mounted on the plate a heightmeter is used to measure the height of the front face of the substrate and the height of the plate is adjusted so that the substrate matches the height of the datum plate The maximum substrate thickness which can be used is roughly 1 2 mm Part Number 878275 Vectorbeam Operator Manual 16 3 16 3 1 16 3 2 16 3 2 1 16 3 2 2 16 4 16 4 1 16 4 2 Page 146 Alignment of substrate for direct write The substrate must be rotated so that the pattern is roughly aligned to the stage axes if direct write is to be carried out This is in order for the machine to be able to place the pattern with the best accuracy and in order to find the global marks easily and possibly automatically Clamp the substrate to the table on which it sits using the small screw operated clamps around the outside of the table These ensure that when the table is rotated the substrate moves with the table and does not stick to the front face referencing points Measuring the rotation Place the holder on the alignment microscope which is provided for this purpose and clamp it into position Focus on the pattern on the substrate Find a feature which is repeated across the substrate at intervals in X or Y Move the sta
83. qualifier is omitted the block number defaults to 0 Blocks will be exposed in numerical order from 0 upwards Block numbers need not be consecutive Offset The offset qualifier defines the stage position relative to the position for block O in millimetres i e not the position relative to the last block to be exposed The offset qualifier will not be accepted for Block 0 generates warning and is ignored The stage positions will not be corrected for the scales applied it is the user s responsibility to ensure that adjacent blocks written at different scales will stitch correctly The Offset qualifier obviously applies to the whole block rather than to any one segment therein However unfortunately a segment number still has to be supplied even if the qSet Grating command contains only Offset or Scale qualifiers Scale The scale qualifier allows additional scaling factors to be applied for the block relative to the current D W mode where 1 0 implies no change of scale Note that the sticky re interpolations are applied i e if the command qSet DWMode Part Number 878275 Vectorbeam Operator Manual 24 9 2 24 9 3 24 9 4 24 9 5 Page 262 Load Wafer Field Main was used to set up the current D W mode then the system will re interpolate only the Main corrections when it applies the Scale adjustment The Scale qualifier obviously applies to the whole block rather than to any one segment therein Ho
84. reply to stage start ready Possible hardware fault on stage to PG fibre optic link Reboot PG and stage 0x400f0015 Bad reply to stage end ready Possible hardware fault on stage to PG fibre optic link Reboot PG and stage 0x400f0016 Can t reexpose insufficient SLCs Possible PG config file error Reboot PG 0x400f0017 Headers not allocated Reboot PG Vectorbeam Operator Manual 27 6 6 13 27 6 6 14 27 6 6 15 27 6 6 16 Part Number 878275 0x400f0018 Page 292 Can t reexpose SLC buffer too small Don t use reexpose qualifier 0x400f0019 Can t reexpose an SLC has more than 1 field Don t use reexpose qualifier 0x400f001a 0x400f001b 0x400f001c 0x400f001d 0x400f001e LHS task errors 0x40100001 0x40100002 0x40100003 0x40100004 0x40100005 No room in Pipe Field Table Widefield Field not found in Pipe Field Table Widefield Can t create field header address table Widefield Can t find field number in PPFT Widefield Can t find field in pattern Widefield LHS receive error from qRoot LHS Reboot LHS LHS error sending to qLHS Root Reboot LHS LHS error sending to qPatternIP Reboot LHS LHS invalid command Reboot LHS LHS stage init error Possible LHS cabling fault Reboot LHS 0x40100006 Height Sensor init error Possible LHS cabling fault Reboot LHS 0x30100007 Error reading height for field 1 If pattern exposure started off the substrate or where the height meter is blocked by subs
85. reserved for the datum The suggested height meter tables for typical substrates are Substrate Table Bright chrome masks Dark chrome masks Si SiO2 CGS Table selection in Emma In order to select a particular table to be used either specify the table with the calibrate command or with the display command e g VB_OPER gt QCAL HEIGHT TAB 7 VB OPER QDISPLAY HEIGHT TAB 7 Once the table has been selected in this way do not specify the table using the tab qualifier with further qdisplay height commands as the command will take longer The last table selected will continue to be used for further height meter commands Height meter calibration Each height meter table to be used should be calibrated first Calibrating the height meter adjusts the laser drive and detector gain using a fast search method to give the correct signal level Calibrate the datum plate height meter table by loading the holder and typing VB OPER MVSP FM move to the Focus Mark VB OPER QCAL HEIGHT TAB 7 calibrate the height sensor Calibrate the appropriate substrate table by loading the holder moving the stage to position the surface under the beam and typing VB OPER QCALIBRATE HEIGHT TAB specified height table The height meter tables for the datum and substrate are calibrated automatically by the holder sequence see Section Holder initialisation The holder sequence should therefore always be run after loading a holder
86. resolution to avoid stitching errors 19 1 6 1 16 bit pattern generator additional details on choosing blocksize The following additional details should be taken into account when choosing the blocksize for the 16 bit pattern generator 1 The blocksize must be within the limits of the hardware The range of physical blocksizes is given in the Sales Specification for the particular machine The tables below show typical values 20 kV 50 kV 100 kV Smallest 128 um 128 um 128 um Largest 1000 um 800 um 560 um 19 1 6 2 18 bit pattern generator additional details on choosing blocksize The blocksize must be within the limits of the hardware The range of physical blocksizes is given in the Sales Specification for the particular machine The following additional details should be taken into account when choosing the blocksize for the 18 bit pattern generator 1 The table below shows typical values for the standard fieldsize pattern generator with HR final lens ae 20 kV 50 kV 100 kV Smallest 128 um 128 um 128 um 1000 um 800 um 560 um 2 The table below shows typical values for large fieldsize pattern generator with HR final lens MED 20 kV 50 kV 100 kV 160 um 160 um 128 um 1000 um 1300 um 560 um 3 The table below shows typical values for large fieldsize pattern generator with UHR final lens 20 kV 50 kV 100 kV 160 um 160 um 128 um 1000 um 1300 um 1300 um 19 1 6 3 20 bit pattern generator additional details
87. resolution does not match calibrated resolution Change fieldsize and hence the pattern generator resolution to match the pattern Remember to load the appropriate field corrections recalibrate and retry Use the noresol qualifier Change the resolution in the converter to match ythe machine and reconvert the pattern QUALBADVAL Bad value supplied for qualifier RBOUTTOL Readback out of tolerance Retry If error persists there may be a hardware fault RKOFFHGT Mark offset height parameter out of range Define mark offset height to be within range SBERR Structure Block transfer failed SINGMATRIX Singular matrix SLVCPUERR Slave Processor is not initialised for communications SPCALFAIL Spot Table Calibration Failed Note the first error message to be reported as this will be the underlying problem to be solved SPOTFAIL Failed to set requested spot Check that the current spot and demagnification tables are correct Check that the mark position and locate parameters are correct Check that the Faraday cup position is correct Check that the autofocus and autostigmation functions work correctly with the current parameters STGLBSECT Stage global data section is not available STGMOVBLK Stage Move Block Transfer Error Reboot PG and retry STIGERROR Error occured during auto stig Check that a normal SEM image of the mark can be obtained Check that the mark position and locate parameters
88. s 163 17 3 3 Defining the aperture adjustment mechanism positions 164 17 3 4 Controlling the automatic aperture adjustment mechanism 164 iza Focusing ihe Deam sciis aa era 165 17 4 4 Manual focus and stigmation adjustment essssss 165 17 4 2 Automatic focus and stigmation adjustment ss 166 17 4 8 Debugging problems with automatic focus stigmation adjustment 168 17 4 4 Correct fine focus value on datum for conjugate blanking 168 WAALS gadjUst fleld 25 aon DR DEDOS 169 17 4 6 How the fine focus setting and adjust field combine to drive the fine focus 169 17 5 Beam curent messHremiel i esce citia dec tiere beni t ev eS Re rk ku nC Eve dE Aia ad 170 17 8 Beam diameter rmeasuremelil uus iceoica soceri pete ine acaba i NEGO 170 17 6 Background issii breed be tus deste dias A 170 17 62 DiSplaydialWiBler ipao De ERE xd e E ONERE a 170 176 8 Mark slope calibldtlOlr uice soie Ee dens pese ER SuSE aer ada sd ee vla Sos 172 qr Besmoumen adpntsl telis on daase roce utt tide aa kit bruta 173 prev PME Won ETT 173 2 ENTIER 173 14 9 BesmidsmetePadlHistnellb d God iade kai toad aou ES cuta did esu Ene Fla pde rsue 174 APO Ak JAUORISlIG uiii evens riter iacet itr ee tanien tbi ba test bt catu haine ble 174 17 9 2 Mahal iiie esee ice ioc tenes ta encre traced a cereo ti sites 174 17 9 Theoretical tables of on axis beam diameters and
89. the EHT Set the ramp speed to avoid too rapid changes VB OPER qset reg 1 28 200 Type the command VB OPER SEHT 20 Set up for operation The following will need to be set up after the beam accelerating voltage has been changed if valid settings cannot be reloaded by restoring a database 1 The final lens C3 current 2 The conjugate blanking 3 The beam current 4 The gun alignment 5 The final aperture 6 The magnetic map 7 Field corrections Fullcal 8 The demagnification table 9 The spot table The procedures for these set ups are described in this manual Part Number 878275 Vectorbeam Operator Manual 15 4 15 5 15 5 1 15 6 Page 134 Final lens 3 C3 set up Lens C3 is set up by the Vistec Engineer so that the conjugate blanking condition is fulfilled See Section Conjugate blanking set up Stigmator balance set up Ensure that the final aperture is centred first In order to adjust the stigmator balance ensure that there is a holder on the stage and a feature such as the FM mark which is a 10 micron octagon is displayed in SEM mode on the video monitor Then type VB_OPER gt QSET WOBBLE ON AXIAL The image on the screen will wobble If there is vertical or horizontal movement minimise this by selecting Stigmator Balance in the Set menu Figure 7 8 and adjusting the Axial X and Axial Y sliders for minimum movement Then type VB_OPER gt QSET WOBBLE OFF AXIAL VB_OPER gt QS
90. the edge m Part Number 878275 Vectorbeam Operator Manual Ae 17 7 1 17 7 2 Page 173 d kl d d d where d is the constant aberration of the optical system d is the edge length and d is the noise on the beam position Hence a plot of beam current against the square of the measured spot size will be a straight line whose gradient depends on the brightness of the source with an offset i e a finite spot size at zero current determined by the noise assumed to be negligible here the aberrations and the width of the edge The function requires a valid spot table so that two currents can be set automatically If beam diameter measurements are to be done on the focus mark type VB_OPER gt QCALIBRATE SLOPE FM FM FC FM FM This will calculate the edge length of the focus mark for use with the beam diameter measurement function so that the corrected value will be the true beam diameter The function qset slope can used to manually change the edge length used by the beam diameter measurement if required Beam current adjustment Automatic Automatic beam current adjustment is only possible if the spot table has been calibrated The machine can adjust all the required lenses the gun alignment and the focus and stigmation to give the desired beam current based on the spot table calibration Type VB_OPER gt QSET CURRENT lt current nA gt lt mark name gt lt mark pos gt bcm FC Manual To change
91. the initial mode fab or sem QDISP MODE OLDMODE F TRNLNM VB MODE QSET MODE FAB Remember the initial position QMOVE POS REL SPO SDP ORIGINAL 0 0 Carry out adjustment QMOVE SPO MarkPos QLOCATE MarkType SDP TEMP rel STATUS F TRNLNM ESTATUS IF STATUS NES X00000001 THEN QDISPLAY COMMENT The mark could not be located and no datum offset adjustment has been done GOTO SKIP ADJUSTMENT ENDIF REL POS F TRNLNM SDP TEMP QDISPLAY COMMENT The datum offset was adjusted by F ELEMENT 1 rel pos QADJUST DATUM SDP TEMP REL SKIP ADJUSTMENT WRITE SYS OUTPUT QDISP POS FULL X f trninm VB_DCPO_OFF_X Y f trninm VB_DCPO_OFF_Y XY f extract 0 f locate X 1 X f extract f locate X 1 7 X f extract 0 f locate Y 1 Y f extract f locate Y 1 7 Y f time JOBFILE F ELEMENT 1 f trninm VB JOBFILE JOBFILE F ELEMENT 0 JOBFILE IF FESEARCH MCLOG JOBFILE _SDP LOG NES OPEN APPEND SDP MCLOG JOBFILE SDP LOG WRITE SDP XY CLOSE SDP ENDIF Vectorbeam Operator Manual Page 218 Now move back to start position resubmitting job if appropriate MVSP SDP_ORIGINAL QSET MODE oldmode IF TIMEINT EQS THEN EXIT QSET ONTIME Timelnt MCJOBS SDP COM MarkType MarkPos Timelnt 19 2 4 Example command The command to reset the datum position at 30 mi
92. the mechanical alignment of the fine focus coil to the optical axis of the final lens has a significant residual error the beam does not pass through the centre of the fine focus coil This means that the on axis beam position varies with the fine focus setting and a correction is applied in the X and Y directions to null this effect Corrections for matching Corrections can be applied in order to overlay a pattern accurately on a substrate which has already been patterned using another exposure tool Direct write correction See Chapter Direct write alignment Stage mapping machine mode This allows more accurate overlay of a pattern on a substrate which has already been patterned using an exposure tool whose stage positioning is repeatable but not accurate in an absolute sense Correction coefficients are applied to the stage in machine mode Correction coefficients can be determined either 1 From measurements with the VB of an array of marks exposed on the tool to which it is to be matched See chapter Machine set ups for further details 2 From measurements on a metrology tool such as the Leica IPRO Stepper lens correction This allows more accurate overlay of a pattern on a substrate which has already been patterned using a step and repeat exposure tool whose imaging system has repeatable distortions See chapter Calibration for further details Part Number 878275 Vectorbeam Operator Manual 10 Laser illumi
93. the required mark etc is located selection of the Save icon will store the position in the Saved Stage Position X Y If the stage is then moved elsewhere selection of the Restore icon will move the stage back to the Stored position It is possible to enable disable the SEM mag tracking by selection of the SEM mag tracking icon with the mouse left button Part Number 878275 Vectorbeam Operator Manual Page 60 7 3 7 4 Emma X04 05D pu 50 023872 mm us Y 0 000000 st Mag 2287 Saved Stage Position X 49 858276 Y 89 259984 Mag x 204 W Mag tracking Grid 0 01000 mm Absolute Figure 7 21 Toolkit Joystick Relative Mode Emma directory structure The directory structure on the VB is set out in the document Machine Managers Guide document no 892813 Emma help For access to Emma Help select with the mouse left button Help in the he top right corner of the STATUS window keeping the mouse button pressed down slide the arrow down to highlight On Commands then release the mouse button After approximately 30 seconds the Emma help window will appear The information is the same as that provided in the Vectorbeam Command Set manual 878274 The commands are grouped into general headings to provide a hierarchy from which the user will be able to quickly find the command they require For information double click on the subject required then selecting with the mouse le
94. the side effect of shifting the centre of the shape in exel format by exel size 2 with respect to the initial outline This does not create any difficulty as the position of the stage is shifted during exposure to compensate For some special applications it might be useful to change the erosion method The 16 bit 18 bit and 20 bit pattern generators allow the user to choose the mode of shape erosion using the command Part Number 878275 Vectorbeam Operator Manual Page 264 VB OPER qset erosion normal nodiscard none or noerode As an example the results of exposing the Vistec resolution pattern acc3n125_sl10_00 fre with the various options are described below 24 10 1 normal option Figure 24 4 shows the usual case for VRU 1 and normal erosion Figure 24 4 Diagram showing the exposed exels for pattern acc3n125_sl10_00 fre at VRU 1 and normal erosion Figure 24 5 shows the exels exposed VRU 2 Since the pattern contains single pass lines at a PG resolution of 3 125 nm the lines are missing at VRU 2 Figure 24 5 Diagram showing the exposed exels for pattern accan125 sl10 00 fre at VRU 2 and normal erosion 24 10 2 Nodiscard option The nodiscard option causes all shapes to be eroded as normal but shapes with heights or widths smaller than an exel but not zero will be exposed as a single exel high or wide as appropriate Figures 24 6 and 24 7 show the exels exposed with the nodiscard option at VRU 2
95. the yaw correction will have no effect In addition the yaw correction is only applied if it has been enabled using the QSET CORRECTIONS ON YAW or QSET CORRECTIONS ON ALL The yaw correction is only used by the jobfiles if the logical vb yaw enabled is set to TRUE but this is part of the system configuration and should not need to be changed Deflection field corrections The focus varies with deflection and the position of the beam does not vary linearly with the deflector drive Deflection field corrections remove these focus and non linear placement dependencies on the deflection of the beam These are e Focus e Stigmation e Main field distortion X and Y offsets e Sub field scaling and rotation e BEF scale and rotation Main field corrections are applied during pattern writing and mark location main or sub field They are not changed at other times such as in SEM mode Part Number 878275 Vectorbeam Operator Manual 9 2 9 2 1 9 2 2 9 2 3 Page 71 Shift scale and rotation corrections for fine focus A characteristic of the optical system is the rotation and change in scale of the deflection field when the fine focus is varied On 18 bit and 20 bit systems with the 1 2 mm fieldsize this effect is nulled out explicitly using dedicated correction coefficients On other systems the effect is compensated for by the mainfield distortions The beam is aligned to the optical axis of the final lens Since
96. this label is visible X Rays produced Removal or tampering with this guard will invalidate Radiological Protection Self explanatory Moving parts can crush and cut Wafer handler may move unexpectedly in maintenance mode Self explanatory Part Number 878275 Vectorbeam Operator Manual Page 17 2 Optical system The purpose of the optical system is to provide at the substrate an electron beam of suitable energy having a constant known appropriate diameter and scan the beam in a precise pattern The optical system uses a thermal field emission cathode and there are two configurations one with a high resolution HR final lens see Figure 2 1 and one with an ultra high resolution lens UHR see Figure 2 2 HR Column Optics Emitter Suppressor Extractor Focus Lens C1 electrostatic gun Anode Tilt Gun alignment jj M Shift M Lens C2 magnetic i M Blanking Blanking Upper main i N Subfield K Lower main i i Selectable aperture Fast focus and i Bu Lens C3 magnetic Xn elemens Substrate Figure 2 1 HR column optics Part Number 878275 Vectorbeam Operator Manual Page 18 UHR Column Optics Emitter Suppressor Lens C1 electrostatic gun l Extractor Focus Anode M Gun ali t Tilt un alignmen K Shift Lens C2 magnetic K i Blanking Blanking Selectable Upper Subfield H W Upper main N H Lower Subfield N H Lower main i i
97. ti eet teet ye vere repe ERE HE E lend 138 15 9 Stage MAPPING PIOS cis desees due cra rie tute i aeia 139 15 9 1 Absolute stage map mode set up sssssseseeese 139 15 9 2 Entering stage map coefficients sssssssesssse 142 15 9 3 Display stage map coefficients sessssssesseese 142 15 9 4 Machine stage map mode set up sseesssssseeeene 142 15 9 5 Select Stage map mode oiii rasan Sr eres E 143 16 Substrate loading and unloading LAS Part number 878275 16 1 Systems with the Brooks handling option iiic iiia 145 16 2 Systems with the single chuck or 10 chuck airlock handling option 145 DEI rm 145 1622 VBO iiie tite itn e plesedi cbe epe t epi eee Ebert 145 Vectorbeam Series Vectorbeam Operator Manual 16 3 Alignment of substrate for direct write ceecceceeeeceeeeeeeeeeceeeeeseaeeeeaeeseneeees 146 10 9 1 Measuring TNE ROLAMOM diit te oh ER ret RE 146 16 3 2 AGIUSTING TNE KOLAUOM sedia dto nene OSEE One 146 16 4 Holder loading unloading in from airlock sssssesseesss 146 WGA WBBOLsienmpiifieiiidtemiten ddnde HON UE Heins 146 16 4 2 VBG single holder alflook tici eth cochera eacte 146 16 4 3 VBG ten holder ainlOcK oit eroi eei ets 147 16 5 Holder loading unloading on from stage s
98. to be smaller in order to adjust the subfield size The Maximum Blocksize must be larger of either the block height or width of any pattern files to be exposed If the Maximum Blocksize is less than the default the total number of bits spanned by all the subfields will be less than 29 The number of subfields that span the Maximum Blocksize can be set by the operator The maximum number is 64 and this is the default The smallest number of subfields that can be set is limited by the maximum subfield deflection and the Maximum Blocksize The maximum number of bits in a subfield is 2 16384 but with 64 subfields and a maximum blocksize equal to the maximum fieldsize only 4096 bits are used The command to set for example 25 subfields with a Maximum Blocksize of 500 um is VB OPER qset block subfields 25 maximum 0 500 Part Number 878275 Vectorbeam Operator Manual Page 208 19 1 7 3 The blocksize must be within the limits of the hardware The range of physical blocksizes is given in the Sales Specification for the particular machine 20 bit pattern generator subfield fracturing X Maximum fieldsize Maximum blocksize Subfields Pattern blocksize Figure 19 5 Diagram showing the arrangement of subfields Maximum Blocksize and the 19 1 8 19 1 8 1 19 1 8 2 pattern blocks
99. used in E Beam lithography systems are raster and vector scanning In the raster technique the beam is scanned over the entire chip area and is turned on and off according to the desired pattern In the vector technique the beam is scanned only over the pattern areas requiring exposure and the usual approach is to compose the pattern from a list of simple shapes such as rectangles triangles and parallelograms Current Vectorbeam Series systems all use the vectorscan method of pattern generation Vectorbeam system overview The Vectorbeam system comprises of the following 1 Plinth supporting substrate chamber with stage and column 2 Control electronics racks 1 CER1 Part Number 878275 Vectorbeam Operator Manual 1 2 1 2 1 Page 6 3 Control electronics racks 1 CER2 4 Electronics cabinet A 5 Electronics cabinet B 6 Stand alone EHT cabinet 7 Operator console with operator terminal printer and video monitor 8 3 temperature control water baths 4 for machines with Brooks handling The layout of these components will vary from installation to installation Plinth The plinth supports the substrate chamber airlock and column on a vibration isolation table BAR CODE READER AIRLOCK RCW ALVES PLINTH SCALP B UAGES Figure 1 1 VB6 plinth Stage A motor driven X Y stage inside the substrate chamber carries a holder containing the substrate to enable its entire surface to be exposed The limit
100. ut otc Eon rid eic AEA 131 15 2 Temp rature moaselemeltl mdp don abides rud eta ER Er eta eave asta eee 131 15 2 1 Location of temperature measurement sensors 131 15 2 2 Obtaining temperature readings 5 2 coi tp eh raideri idana 131 15 23 emperature eb Upon eain oS 132 15 3 Changing the beam accelerating voltage kV ssssssssssss 132 15 3 1 Choosing the beam accelerating voltage sssss 132 15 3 2 Important notes before increasing the beam accelerating voltage kV 133 153 3 Gohaitionitig leui 3 9 3 2 reri eie e 133 1594 SEWN ERT sicot tct ictee Rr RE 133 15 9 5 XObupTOr operato Mesmos EHEORERA 133 154 Final lens 9 OS BOLUasiicisid tense pedit cia cse a e QU Ms aes 134 15 5 iSugiator balance sal Hp ice eet Ete erba ces DER HU d eres la ea RU SR Rega 134 15 5 1 Setups affected by stigmator balance esssssss 134 15 6 COMICS Banking Sel UD iiser tia ardua census elio piss dans dp d rdi 134 15 61 Lens CSS SetUp ie tarditas cete e tente a teret ine utu Ea 135 154 Magneie map CAN ISO ssi roter bad da Ye rige rb ana exe PRJad P essi desc a 136 15 7 1 GallbrationJOD IIlo c crt ceo deti rire ieget i eee edel toten 137 15 8 Demagnification and spot table calibration sse 138 1581 Demagniticalomtable cca theo tot i ipee s Ee te ero reete 138 15 9 2 SpoLtable dites
101. 0 62 500 Defines the stage position in mm of the centre of the layout of the dies This is the only parameter which fixes the relationship between the stage coordinates and the die co ordinates on the substrate even if direct write is used This parameter may be set to RELATIVE instead of an X Y coordinate in which case the layout centre is taken as the current position of the stage when the file WLVD COM is called This is useful for nested layouts see Section Nested layouts MAINSCAN Serpentine Defines the order in which the main dies are exposed This can either be serpentine or raster DROPOUTS Defines the number of groups of dies to be dropped out A dropped out die means that the workfile pattern is not exposed at that die The dropouts are arranged into separate groups in order to allow different plugins see below The dies belonging to each group must also be defined as shown below by defining the symbols dropoutcells_1 to dropoutcells_n where n is the same as the number of dropouts DROPOUTCELLS_1 1 3 10 3 9 9 DROPOUTCELLS 2 5 1 6 5 DROPOUTCELLS_3 5 3 Each symbol defines a group of dies to be dropped out A group may contain up to 10 cells listed as shown All the dies contained in the first n groups where n is the number of dropouts will be dropped out PLUGIN Defines how many of the groups of dropouts defined above are to be plugged in A plugin die means that a pa
102. 001 PatWrite receive error from qPatternIP Reboot PG 0x400f0002 PatWrite invalid response from DCP Reboot PG 0x400f0003 Invalid FISP Complete response Reboot PG 0x400f0004 Invalid SLC Complete response Reboot PG 0x400f0005 PatWrite invalid response from slave Reboot PG 0x400f0006 PatWrite invalid command Reboot PG 0x400f0007 PatWrite error sending to slave Reboot PG 0x400f0008 PatWrite error receiving from slave Reboot PG 0x400f0009 PatWrite wrong reply received from slave Reboot PG 0x400f000a Error on stage move Possible hardware fault on stage to PG fibre optic link Reboot PG 0x400f000b Invalid LHS response Possible hardware fault on LHS link Reboot LHS and possibly PG 0x400f000c Invalid Linewriter EOF interrupt Possible hardware fault Reboot PG 0x400f000d Invalid FISP transfer response Reboot PG 0x400f000e Invalid SLC transfer response Reboot PG 0x400f000f Stage move start not signalled Possible hardware fault on stage to PG fibre optic link Reboot PG and stage 0x400f0010 Stage move end not signalled Possible hardware fault on stage to PG fibre optic link Reboot PG and stage 0x400f001 1 Stage move abort not signalled Possible hardware fault on stage to PG fibre optic link Reboot PG and stage 0x400f0012 Stage move start timeout Possible hardware fault on stage to PG fibre optic link Reboot PG and stage 0x400f001 3 Stage move end timeout Possible hardware fault on stage to PG fibre optic link Reboot PG and stage 0x400f001 4 Bad
103. 0820008 FISP Invalid pointer Reboot PG 0x40820009 FISP Ethernet timeout waiting for pattern data Possible cabling fault Reboot PG 0x2082000A _ FISP Invalid sorting strategy received Reboot PG 0x2082000B _ FISP Bad subfield number calculated Reboot PG 0x2082000C _ FISP Invalid stripe number Reboot PG 0x1082000D FISP Info Buffer fill semaphore timeout Reboot PG 0x4082000E FISP Tried to spawn 2nd buffer fill task Reboot PG 0x2082000F _ FISP Illegal opcode in input pattern data Reboot PG 0x40820010 FISP Semaphore taken but Fisp buffer full Vectorbeam Operator Manual Page 293 Reboot PG 0x20820011 FISP Structure ID number out of range Possible pattern data error Reboot PG 0x30820012 FISP Ethernet timeout waiting for pattern data Used instead of ER FISP ETHERNET TIMEOUT when retry is possible If the retry is successful then no action is required Otherwise reboot PG 27 6 6 17 Slave DCP errors Part Number 878275 0x20830001 DCP Already wobbling stigs Reboot PG 0x20830002 DCP Not wobbling stigs Reboot PG 0x40830003 DCP Can t create wobble task Reboot PG 0x40830004 DCP Can t delete wobble task Reboot PG 0x20830005 DCP Bad stig value to set Check value being set Reboot PG 0x20830006 DCP Focus setting at max extreme Set the fine focus to 0 0 and check image in SEM mode Adjust C2 to give best focus and try again Check the autostig autofocus parameters 0x20830007 DCP Focus setting at min extreme Set th
104. 1 6 Page 205 pattern file The maximum logical blocksize that can be set in the converter for a FRE file is given by Max logical blocksize in um 2 6 32 x PG resolution Equation 19 12 Expressed in terms of exels Max logical blocksize in exels 2 32 Equation 19 13 This results in Max physical blocksize in um 27 512 x PG resolution Equation 19 14 The reason not all 2 deflection bits can be used for a physical block is because the FRE file format uses unsigned short integers to represent the coordinates of the shapes in outline format Such representation is compact but can only define 2 6 different numbers As the shapes are defined in outline format this limits the number of exels which can be defined in a logical block to 2 1 The maximum logical blocksize is limited to 2 6 32 to avoid possible stitching problems on the 16 bit pattern generator if the blocksize were not an integer number of beamsteps As 16 x 16 logical blocks are combined into a physical block the physical blocksize is therefore limited to 2 512 Choosing the blocksize The following should be taken into account when choosing the blocksize 1 The blocksize which is set in the converter must be several microns smaller than the maximum allowed blocksize to ensure that only the accurately calibrated part of the deflection field is used for exposure The accurately calibrated part of the deflection field is the
105. 11 1 3 11 2 3 11 2 1 3 11 2 2 3 11 3 Page 35 isolated vent Consequently when switching PICS off either a full vent must be done or an isolated vent can be done after arranging an external compressed air supply on the close line Vacuum system start up Please see Vectorbeam Customer service procedures manual document number 893116 Vacuum system shut down Please see Vectorbeam Customer service procedures manual document number 893116 Vacuum system monitoring Some of the pressure gauge readings can be displayed in Emma using the qdisplay vac command For PICS versions X10a VB6 and V6 VB5 or later a choice can be made between the set of gauges read by the stage subsystem directly or the set read by the stage subsystem from the PICs subsystem For previous versions of PICs only the set of gauges read by the stage subsystem directly can be displayed In order to select which set of gauges is displayed the variable TheGuagesType g sic in SV_CONFIG VW must be defined For stage gauges it is set to 0 and for PICs gauges it is set to 1 Stage Gauges The gauges that can be read directly from the stage processor are as follows IGP1 IGP2 IGP3 IPI 1 2 Gun Pe System Pe Airlock Pe Airlock Pi Vacuum gauges The gauges that can be read from PICs by the stage processor or by the Brooks CTC computer are as follows VB5 IGP1 IGP2 IGP3 IGP4 G1 G3 G4 VB6 IGP1 IGP2 IGP3 IGP4 Pi1
106. 18 bit machines Format VB VB Minimum Maximum Minimum Maximum VB Maximum Maximum Field Max field Resolution Resolution VB Resolution VRU default um um um um Resolution um field pum count VB Not Not used 0 000125 0 00390625 Not used Not used 512 1048576 used 1024000 VB20 1024 1024 0 000125 0 5 0 000125 0 0009765625 512 1048576 1024000 VB50 819 2 819 2 0 000125 0 4 0 000125 0 00078125 512 1048576 1024000 VB100 589 824 589 824 0 000125 0 288 0 000125 0 0005625 512 1048576 1024000 Table 19 3 Default limits for 20 bit machines 19 1 1 1 2 VB The VB command types up the various VB specific parameters such as VB Machine VB Field VB VRU etc 19 1 1 1 3 VB machine The VB machine must be set to 16 18 or 20 corresponding to the number of mainfield bits of the Vectorbeam pattern generator for which the pattern is to be generated If the VB Machine is set to 16 the output file format used by the Writefile will be FRE If the VB Machine is set to 18 19 or 20 the output file format used by Writefile will be VEP 19 1 1 1 4 VB field The VB Field parameter is only used for the VB20 VB50 and VB100 formats and defines the largest Maximum Fieldsize that can be set on the Vectorbeam Part Number 878275 Vectorbeam Operator Manual 19 1 1 1 5 19 1 1 1 6 Page 196 pattern generator A separate value is maintained for each of the formats VB20 VB50 and VB100 The default VB field values are
107. 30AA0F72 Ox30AAOF73 Ox30AAOF74 0x40AA0F75 0x10AA0F81 0x10AA0F82 0x10AA0F83 0x10AA0F84 0x10AA0F85 0x40AA0F86 Ox30AAOF87 Ox30AAOF88 Ox30AAOF89 Ox30AAOF8A Ox30AAOF8B Ox30AA0F8C Ox30AAOF8D Ox30AA130E 0x30AA130F 0x40AA1409 0x20AA140C 0x20AA140D 0x20AA140E 0x20AA1410 0x20AA1 411 0x20AACO000 0x20AACO01 Ox30AAC002 0x20AACO003 Ox30AAC004 0x40AAC005 0x40AAC006 0x20AACO007 0x20AACO08 0x30AACOO09 Ox40AACO0A 0x20AACOOB 0x20AACOOC 0x20AACOOD 0x20AACOOE 0x40AACOOF 0x40AACO10 0x40AACO 1 1 0x40AAC012 0x40AACO17 0x40AAC020 0x40AAC021 0x40AAC022 0x40AAC023 0x40AAC024 0x40AACO25 Page 300 SAEHT Hardware fault SAEHT Control mode not yet OK SAEHT No extractor current SAEHT param name not specified SAEHT value not specified SAEHT Vacuum OK again SAEHT Vacuum poor SAEHT Vacuum timed out SAEHT Vacuum bad SAEHT Vacuum failure SAEHT UPS exhausting SAEHT Could not spawn Set process SAEHT Maximum less than standby value Cannot connect to SAEHT Cannot disconnect from SAEHT SAEHT communications link error SAEHT not yet initialised SAEHT Runup started SAEHT Rundown started SAEHT Monitor mode set ON SAEHT Monitor mode set OFF SAEHT Mains failure SAEHT Hot box detected on supposedly LaB6 system SAEHT EHT analogue reading unstable SAEHT Emission current analogue reading unstable SAEHT Bias voltage analogue reading unstable SAEHT Filament voltage analogue reading unstable SAEHT Filame
108. 3A laser product Part Number 878275 Vectorbeam Operator Manual Page 14 5 Crush Hazard Keep hands clear Self explanatory Crush Hazard Keep hands clear of Airlock Door Self explanatory Crush Hazard Keep hands clear of Plinth when switching off Airfeet Possible hazard during service operations refer to service manual No hazard in normal operation HIGH LEAKAGE CURRENT Connection to earth essential before connection of supply A Self explanatory A DANGER Electrical shock hazard Equipment powered by multiple sources A Possible hazard during service operations refer to service manual No hazard in normal operation Possible hazard during service operations bakeout No hazard in normal operation Protective conductor terminal do not disconnect any lead connected to this A D l Earth ground terminal Frame or chassis terminal Part Number 878275 Vectorbeam Operator Manual Page 15 Accessible Functional Earth Equi potentiality On Supply Off Supply Warning Electro Static Discharge ESD sensitive service engineers take precautions 1 10 2 Labels only found on machines with wafer loading automation X Rays produced If label is visible then Radiological Protection is Missing The machine should not be operated with beam if
109. 400b0104 Couldn t spawn a task 0x400b0105 Couldn t spawn SLC task Reboot PG 0x400b0106 Couldn t spawn FISP task Reboot PG 0x400b0107 Couldn t spawn FISP buffer fill task Reboot PG 0x400b0108 SLC task has already been spawned 0x400b0109 FISP task has already been spawned 27 6 6 3 C library errors 0x400b0201 Malloc failed in pattern reception Part Number 878275 Vectorbeam Operator Manual 27 6 6 4 27 6 6 5 27 6 6 6 27 6 6 7 27 6 6 8 27 6 6 9 27 6 6 10 Part Number 878275 Semaphore errors 0x400b0301 Page 290 Error creating semaphore Reboot PG 0x400b0302 Error granting semaphore Reboot PG 0x400b0401 Memory allocation errors Error allocating memory Reduce any data gathering processes Reboot PG 0x400b0402 0x400b0403 0x400b0500 Reboot PG Reboot PG Error freeing memory Can t generate Physical logical table Register access errors This register is write only Use qset reg command not qdisp reg Reboot PG 0x400b0501 This register is read only Use qdisp reg command not qset reg Reboot PG 0x400b0502 This register does not exist If using qset reg command check parameters Reboot PG 0x400b0503 Wrong data type for this register If using qset reg command check parameters Reboot PG 0x400b0504 Data out of range If using qset reg command check parameters Reboot PG 0x400b0505 Illegal function attempted If usi
110. 5 9 SuD eld GIStOMON caei a 182 18 5 4 Beam error feedback distortion sess 182 18 6 Jobcal and fullcal errors and warnings essen 182 18 64 Additional WalhllgjS cci tocco tert to i Pacers oreste deese tese ritade tigi 182 19 6 2 Additional eitOIS sieoicoes ii der aer eie ein oe k sh beiden tore eat roe a Ebo eaa 182 18 7 Checking deflection field corrections seseseseeeeene 183 16 71 Field focus and stigmatioh iiiciii eerte rhe nes 183 19 72 Mamield GISLOMION scccssteecads stecaassersacatscicedead race da eversa da o pier Enion ied 183 19 73 JSubfielddiStOFtOTi occi roce areae epe e d 1D road RET o oda rie er eroe dad 183 18 7 4 Beam error feedback distortion eese 183 18 8 Fine tuning deflection field corrections essesseeeee 184 16 94 Field FOCUS and stigmatiOnhi uo rire ette ctor pote etrnaitnevdces 184 19 8 2 MainfielddislOFtlOlt iiie toro rotos vay ess acueyY tha eyes exar EENE 184 19 8 3 Subfield diStortlOric saves ribi y Reese ce virer Erba pope des Ete EE 185 18 8 4 Beam error feedback distortion sssssesssss 185 19 9 dalbratol Ols rannen ip orb DE oEEg dere Et here i Pep UR E publ o PE DMR pO M dine 185 19 91 T6 BIE macies sco sies copa up epa epo age xp P tubo an Fx R 186 16 92 T8 bit and 20 bIE MACHINES isana 186 18 10 Stepper lens Calibration os ie s
111. 5 MHz DCBADFRQADJ Frequency can t be adjusted out of hardware range Check VRU current clocks doses and resist sensitivity parameters Change conditions if necessary DCBANDNARR Clock band not set too narrow Set wider clock band DCBANDNSET Couldn t set clock band DCBANDWIDE Clock band wide clocks may be inaccurate Set narrower clock band DCCLKSOUTB Clock s outside clock band Set wider clock band to cover clock or change clock DCCMDFAIL Clock command failed Check clock hardware Reboot PG DCCMDNOGOOD Clock command irrelevant to installed hardware Check clock hardware configuration file Reboot PG DCFRQOUTBND Frequency can t be set outside clock band Set clock band to cover required frequency DCUNKNERR Unknown clock command error code was returned Retry Reboot PG DCUNKNHW Can t find what PG clock hardware is in use Check clock hardware configuration file Reboot PG DEMAGFAIL Demag Table Calibration Failed Note first error message given and rectify DIAGDRIVE Diag stig drive beyond range DMATRIX1 dmatrix allocation failure 1 DMATRIX2 dmatrix allocation failure 2 DOSADJFAIL Dose adjustment failed Check VRU current clocks doses and resist sensitivity parameters Change conditions if necessary DOSEFAIL Failed to set requested dose Check VRU current clocks doses and resist sensitivity parameters Change conditions if necessary DOSTYPINV Invalid dose ty
112. 5 na 999 don t restore beam pitchinx 50 lin mm pitchiny 50 lin mm cellsinx cellsiny layoutcentre 77 500 90 500 X Y position in mm mainscan Serpentine Serpentine default or raster dropouts No of groups of dropouts dropoutcells 1 2 2 plugin 0 No of groups of plugins pluginworkfile 1 plugin 00 reset datum interval 999 calibrate interval 999 VB SEQ WLVD COM VB USERS JOB_FILES SECOND_LEVEL COM FINISH_ALL Second level layout parameter file The name of this jobfile is SECOND_LEVEL COM ON CONTROL_C THEN GOTO FINISH_ALL ON CONTROL_Y THEN GOTO FINISH_ALL TYP FSENVIRONMENT PROCEDURE SJOB F bENVIRONMENT PROCEDURE holder 999 6 inch mask 999 dont init holder calibrate 0 calibrate workfile VERNCHECKA 12 5 nm res workfile_vru 12 5 nm beamstep vru res startdose 135 uc cm 2 20 KV dosestep 1 x overall operator 0 no update or beam 999 2na 999 don t restore beam pitchinx 0 75 lin mm pitchiny 0 75 in mm cellsinx 3 cellsiny layoutcentre relative X Y position in mm mainscan Serpentine Serpentine default or raster dropouts 1 No of groups of dropouts dropoutcells 1 3 1 2 2 3 2 plugin No of groups of plugins pluginworkfile 1 plugin 00 reset datum interval 99
113. 9 calibrate interval 999 VB SEQ WLVD COM FINISH ALL Part Number 878275 Vectorbeam Operator Manual 20 Page 231 Direct write alignment Direct write alignment is the process that aligns one pattern to an existing pattern on the substrate All the direct write methods used on the VB are based on mapping the machine s absolute co ordinate system to match the substrate s co ordinate system This allows substrate design distances to be used directly for positioning the stage and the beam and so when a pattern which is laid out in CAD coordinates is exposed it will be accurately positioned relative to the existing pattern In order for the VB to be able to calculate the mapping coefficients both the design dimensions and the dimensions measured on the machine must be known There are two mapping modes called absolute One is a stage mode which uses mapping coefficients to compensate for any positioning errors from absolute The other is a direct write mode for which the mapping coefficients are all zero The appropriate stage mode is used in addition to the direct write mode Offset scaling Rotation Keystone Figure 20 1 Schematic of the various errors to be corrected with direct write mapping coefficients The black squares are the expected and the white squares are the measured 20 1 positions of the alignment marks Direct write alignment methods Many direct write
114. 9 Calibration offsets Sometimes the exposed lithography may have consistent mainfield or subfield errors of several nanometres in size in spite of the usual calibrations having been done correctly These can be corrected manually by setting additional empirically determined calibration offsets Part Number 878275 Vectorbeam Operator Manual 18 9 1 18 9 1 1 18 9 1 2 18 9 1 3 18 9 2 Page 186 16 bit machines Mainfield scaling offsets Offsets can be applied to the mainfield X and Y scalings by adding height offsets to the measured height before the scale for height adjustment is carried out This is done with the following commands for X and Y respectively VB_OPER gt qset register 3 24 value in microns VB_OPER gt qset register 3 27 lt value in microns These commands change the NonLinearGainConst terms of the height dependent field scaling described in the Chapter Height meter The effect of a height offset on scaling can be calculated from the usual scale for height dependence of about 20 nm um at the edge of a 1 mm field These offsets must be removed before mainfield calibration as otherwise their effect will be calibrated out and then re applied This is done automatically by jobcal Mainfield rotation offsets Offsets can be applied to the mainfield X and Y rotations by adding height offsets to the measured height before the rotation for height adjustment is carried out This is done with the fol
115. ALTIME QSET MODE FAB Move to stored position m5 substrate centre MVSP M5 Calibrate the height sensor on the substrate table 1 QCAL HEIGHT TAB 1 Calibrate the height sensor on the datum table 7 MVSP FM QCAL HEIGHT TAB 7 Calibrate a heightmap where FM is the datum position M1 is the start position close to the origin 0 0 on the substrate the step and repeat for the grid is 9 mm and the number of points in the grid is 9 x 9 QCAL HEIGHT MAP FM M1 9 GRID 9 AVERAGE 2 DIAG TABLE DP 7 SUBS 1 Set the height map OFF in height mapping mode QSET HEIGHT MAP OFF Calibrate JOBCAL Set the height map ON in height mapping mode QSET HEIGHT MAP ON Mini height map 18 bit and 20 bit pattern generators only A mini height map can be generated prior to exposure and used over a limited area during exposure instead of realtime height measurements This is particularly useful for direct write because Height readings can be taken at defined positions on patterned substrates where the heightmeter readings are not disturbed by the patterning The typical size of dies exposed by a stepper 25 x 20 mm is suitable for use with a mini map Height readings can be taken close to each alignment mark during die by die alignment in addition to the mark locate and this does not increase the time overheads very much The increased time to make up to 4 height readings required for t
116. CBANDNARR Clock band not set too narrow DCBANDWIDE Clock band wide clocks may be inaccurate DCUNKNERR Unknown clock command error code was returned DCCLKSOUTB Clock s outside clock band Vectorbeam Operator Manual Page 288 08019760 PGWARN Warning message from PG in console window 08019768 SVWARN Warning message from Stage in console window 08019770 EOWARN Warning message from EO Control in console window 08019778 LMPROWHOLE Data has consecutive failed locates looking along rows 08019780 LMPCOLHOLE Data has consecutive failed locates looking down columns 08019788 LMPPERCENT Significant percentage of mark locates failed 08019790 BADHOLDERINPUT Incorrect holder ID specified 08019798 NONEWHOLDER Could not create a new holder definition 080197A0 UNKNOWNHOLDER Unknown holder specified 080197A8 MRKFULLFAIL Mark locate exhausted full field 0801AEE2 ETHOPSPAWN Subsystem couldn t spawn Ethernet output process 0801AEEA ETHIPQCRE Subsystem couldn t create Ethernet input queue 0801AEF2 ETHOPQCRE Subsystem couldn t create Ethernet output queue 0801AEFA ETHHOOKADD Subsystem couldn t set up Ethernet packet intercept 0801AF02 ETHLONGMSG Subsystem long Ethernet message send failure 0801AF0A ETHOPQFAIL Subsystem couldn t queue Ethernet message for output 0801AF12 ETHOPSEND Subsystem couldn t pass message to Ethernet output 0801AF1A ETHIFUNIT Subsystem Ethernet ifunit failu
117. CO2D OxA40AACO2E Ox40AAC02F 0x30AAC030 Ox30AAC031 0x30AACO032 0x30AACO33 Ox30AAC034 0x30AACOS35 0x30AACOS36 0x30AACO37 0x30AAC100 0x40AAC101 0x40AAC102 0x40AAC102 0x40AAC103 0x40AAC201 0x40AAC202 0x40AAC203 0x40AAC204 0x40AAC206 0x40AAC207 0x40AAC208 0x30AAC209 0x30AAC210 0x30AAC212 0x10AAC213 0x10AAC214 0x30AAC215 0x40AAC216 0x40AAC217 0x40AAC218 0x20AAC219 0x20AAC21A 0x20AAC21B 0x40AAC21C 0x40AAC21D Ox40AAC21E Ox40AAC21F 0x40AAC230 0X40AAC231 0X40AAC232 0x20AB0301 0x20AB0302 0x20AB0303 0x20AB0304 0x40AB0308 0x40AB0309 0x40AB030C 0x40AB030D 0x20AB030E 0x40AB0400 Page 301 Power failure of C1 lens driver 15 V supply C1 lens driver board power failure C2 power amplifier temp exceeds 80 C Failure of C2 lens driver current source Power failure of C2 lens driver 15 V supply C2 lens driver board power failure FL power amplifier temp exceeds 80 C Failure of FL lens driver current source Power failure of FL lens driver 15 V supply FL lens driver board power failure CO lens readback tolerance error C1 lens readback tolerance error C2 lens readback tolerance error FL lens readback tolerance error CO range register error C1 range register error C2 range register error FL range register error Invalid EHT value for lens EHT alignment Lens out of range Lens setting out of range Lens mode out of range Lens demand is non zero when EHT is off Ellink primary status register s
118. COMMENT Executing clock file CLOCKFILE CLOCKFILE If clockfile exists then run it See Example clockfile in Chapter Exposure dose ELSE QDISPLAY COMMENT No clock file setting all clocks to relative dose 1 QSET DOSE 0 31 1 ENDIF If no clockfile is present then the clocks are set to give a relative of 1 If the pattern contains realtive doses then this will be overriden by the next step QSET PATTERN WFILE Select pattern This will set any relative doses contained in the pattern LAYOUT CLEAR Clears any existing definitions LAYOUT PATTERN WFILE Define pattern file MAINSCAN 0 32 MAINSCAN MAINSCAN F EDIT MAINSCAN COLLAPSE Create symbol mainscan in case one was not defined in the parameter file but keep any string if one was defined Part Number 878275 Vectorbeam Operator Manual Page 226 IF MAINSCAN NES THEN LAYOUT MAIN MAINSCAN ELSE QDISPLAY COMMENT NOTE THE SYMBOL MAINSCAN HAS NOT BEEN DEFINED QDISPLAY COMMENT NOTE THE DEFAULT SERPENTINE WILL BE USED ENDIF Define order of dies IF LAYOUTCENTRE EQS RELATIVE THEN LAYOUT RECT 0 0 PITCHINX PITCHINY CELLSINX CELLSINY CENTRE RELATIVE ELSE LAYOUT RECT LAYOUTCENTRE PITCHINX PITCHINY CELLSINX CELLSINY CENTRE ENDIF Defines rectangular layout LAYOUT DOSE STARTDOSE DOSESTEP OPERATOR Sets the dose
119. Check for sensible command parameters Contact Vistec Part Number 878275 Vectorbeam Operator Manual Page 283 SVDDIVZHY Division by zero in SVD matrix arithmetic hy Check for sensible command parameters Contact Vistec SVDDIVZX Division by zero in SVD matrix arithmetic x Check for sensible command parameters Contact Vistec SVDDIVZZ Division by zero in SVD matrix arithmetic z Check for sensible command parameters Contact Vistec SVDNOCONV SVD matrix arithmetic didn t converge Check for sensible command parameters Contact Vistec SVERROR Stage error message in console window Check for sensible command parameters Contact Vistec SVFAULT Stage fault message in console window Check for sensible command parameters Contact Vistec SVVXERR Stage VxWorks error message in console window Reboot stage SWERR Sub system software error Reboot subsystem TRAPDEFLIM Trapezium deflection limit exceeded UABLTODB Unable to open machine database file Exit Emma run Emma environment and then restart Emma Check that the logical vb db exists Check the directory and file exists UABLTOUPD Unable to update machine database file Check the directory and file exists using the directory command UNABLCLOSE Unable to close file UNABLEOPN Unable to open file UNKNOWNHOLDER Unknown holder specified UNKNSUBSYS Message from unknown subsystem in console window UNLOADERROR Substrate unload
120. EON Invalid database section BADTIME Unable to get system time date UNABLCLOSE Unable to close file DBBADMC This file was created on a different machine DBEHT EHT in file different to current value DBFLD Field size in file different to current value DBMFSENS Main sensitivity in file different to current value DBTFSENS Trap sensitivity in file different to current value DBBEFSENS Bef sensitivity in file different to current value DBMVPIVOT Main pivot point in file different to current value DBTVPIVOT Trap pivot point in file different to current value DBFOCSENS Height focus factor in file different to current value DBHRSENS Height rotation factor in file different to current value DBPARSE Failed to parse file specification DBSEARCH Could not decode file specification or no file COLRESTORE Unable to restore column settings Vectorbeam Operator Manual 08019168 08019170 08019178 08019180 08019188 08019190 08019198 080191A0 080191A8 080191B0 080191B8 080191C0 080191C8 080191D0 080191D8 080191E0 080191E8 080191F0 080191F8 08019200 08019208 08019210 08019218 08019220 08019228 08019230 08019238 08019240 08019248 08019250 08019258 08019260 08019268 08019270 08019278 08019280 08019288 08019290 08019298 080192A0 080192A8 080192B0 080192B8 080192C0 080192C8 080192D0 080192D8 080192E0 080192E8 080192F0 080192F8 08019300 08019308 08019310 08019318 08019320
121. ET WOBBLE ON DIAGONAL The image on the screen will wobble If there is movement minimise this by selecting Stigmator Balance in the Set menu Figure 7 8 and adjusting the Diag X and Diag Y sliders for minimum movement VB_OPER gt QSET WOBBLE OFF DIAGONAL The stigmator balance should only need to be set up once for each beam accelerating voltage These values will not change unless the column characteristics change Save the set up in the appropriate database Setups affected by stigmator balance Changing the stigmator balance will changed the mainfield distortion corrections and these will need to be recalibrated using fullcal Conjugate blanking set up The beam blanking is conjugate when the beam focal point after the 2 lens C2 is exactly half way along the blanking plates Conjugate means that there is no movement of the spot at the substrate as the voltage on the blanking plates is changed but only a change in current This is important for eliminating possible errors in shapes at the points where the beam is blanked and unblanked Such error would result in tails on each shape at the point where the beam blanks or unblanks The fundamental conjugate blanking setup is done once at installation or after mechanical work on the column The current in the 2 lens C2 is adjusted to place the crossover half way along the blanking plates Beam movement is minimised while gradually varying the blanking voltage using the potenti
122. F 2 Write all Emma commands in a DCL command file in the full form beginning with q to avoid additional time overheads associated with translating the mnemonics Part Number 878275 Vectorbeam Operator Manual 24 7 24 8 24 8 1 24 8 1 1 Page 258 3 Do not iconise the Status Window 4 Control Emma through a control program which is installed in VMS Pattern sleeving Pattern sleeving is a technique which is used to advantage when exposing a pattern containing both large and small shapes The pattern data are manipulated so as to obtain two separate pattern files The first contains the borders of all the shapes with a border width of at least the minimum dimension The second contains the remaining parts of the shapes The first pattern is exposed with a small beam diameter to obtain high fidelity of all edges The second pattern is exposed with a larger beam diameter to quickly fill in the centres of the shapes This technique enables the faster exposure of patterns while retaining the fidelity of using a small beam diameter Fieldsize adjustment resolution Very small changes in the field size are possible up to the maximum range limit However all the corrections such as rotation keystone magnetic map etc are applied to the same DAC and so the maximum range for scaling alone only applies if all the other corrections are zero 16 bit pattern generator The mainfield fine scaling DAC which is 18 bit has a
123. FIL Not a pattern file Check that the file specified is correct NOVIDEO No video signal detected OUTVBNL Outside virtual block limits PACKORERR Ethernet packet order error PARAMRNG_ Parameter passed was out of range Check validity of parameter PGCFGBAD PG setup is inconsistent bad mix of processor types Check PG configuration file Reset PG PGCFGCFAIL PG configuration check failed PGCFGNLOGS PG setup no of SLAVE logicals not same as no of FISPs Check PG configuration file Reset PG PGCFGNOTOK PG unusable processors are not OK or missing reboot Wait for 1 minute and retry Reset PG Check PG configuration file PGERROR Pattern Generator error message in console window PGFAULT Pattern Generator fault message in console window PGGLBSECT PG global data section is not available PGNRDY Pattern Generator is not ready for exposure PGVXERR Pattern Generator VxWorks error message in console window PGWARN Warning message from PG in console window PTNBADFIRSTXY First x y badly specified PTNBADLASTXY Last x y badly specified PTNBADRANGEAREA Bad pattern Range Area PTNCANTFIND Cannot find fields in given Range Area Part Number 878275 Vectorbeam Operator Manual Page 282 PTNDATA Pattern data transfer failed PTNHDRERR Pattern header transfer failed PTNLASTFIRST Last field lt First field PTNNSLC Pattern file is not selected PTNRESERR Pattern
124. GainConstY NonlinearGainScaleX NonlinearGainScaleY 1 000 NonlinearGainSquareX NonlinearGainSquareY The use of the coefficients to adjust the height meter reading can be enabled or disabled using NonlinearGainOn In order to access the coefficients on the DCP directly login to the DCP VB SUPER rlogin dcp Part Number 878275 Vectorbeam Operator Manual Page 80 Vectorbeam Pattern Generator Master login pgdiag Password gt To display the current value type gt double NonlinearGainScaleX To change the current value type for example gt double NonlinearGainScaleX double NonlinearGainScaleX To ensure the values are set when the DCP is rebooted put them into the DCP_CONFIG VW file for 16 bit systems or WFDCP_CONFIG VW for 18 bit and 20 bit machines In order to set and display the coefficients through Emma commands use the qset reg and qdisplay reg commands together with the numbers in the table below X Register Register Y Register Register for 16 for 18 for 16 for 18 bit PG bit and bit PG bit and 20 bit 20 bit PG NonlinearGainOn NonlinearGainScaleX NonlinearGainSquareX 10 8 NonlinearGainOn NonlinearGainScaleY NonlinearGainSquareY PG Fine tuning the height dependent field rotation Only UHR systems have height dependent field rotation A rotation for height coefficient is defined and a compensating correction is applied for the difference in height bet
125. ITAL that first 1 the focus is set up and 2 a height meter reading is obtained on the mark used This is because the reference height for field size and focus is reset Jobcal does this automatically 18 4 18 4 1 Interpretation of jobcal fullcal on axis calibrations The main field subfield beam error feedback focus and stigmation on axis calibrations are each followed by further measurements which do not alter the calibrations but enable the accuracy of the calibration to be assessed They have a similar output format Correction coefficients The correction coefficients are the fine corrections applied after calibration Part Number 878275 Vectorbeam Operator Manual Page 179 18 4 2 Residual coefficient errors The residual coefficient errors are measured after completion of the last calibration iteration Four measurements are done of the remaining errors in the calibration and mean and sigma values are calculated The residual coefficient error means are the changes that would need to be applied to the correction coefficients based on these measurements The residual coefficient error sigma values show the uncertainty in the means 18 4 3 Correction errors nm The correction errors are the same measurements as the residual coefficient errors except that the results are presented as distances in nm at the corners of the field being calibrated 18 4 4 Main field calibration An accurate calibration shoul
126. L NOT CORRECT GOTO FINISH WLVD ENDIF Check syntax of time interval QSET ONTIME calibrate interval MCJOBS RESET_ONTIME_JOBCAL COM calibrate_interval ENDIF Set up ontime for running jobcal QDISPLAY COMMENT f time Started exposing workfile Log the start time IF P1 EQS THEN LAYOUT DO QEXPOSE PATTERN NOPOSTMOVE IF P1 NES THEN LAYOUT DO P1 Expose the layout If a string was passed to wlvd com then issue this string should be a valid command at each layout position instead of QEXPOSE PATTERN QDISPLAY COMMENT f time Finished exposing workfile Log the finish time ERROR Success error code GOTO FINISH WLVD ERROR WLVD ERROR 44 Fatal or severe error code FINISH WLVD IF LAYOUTCENTRE EQS RELATIVE THEN QSET PATTERN NONE If a relative layout has been processed then set the pattern back to none to avoid offsets being introduced for pattern centring by next top level layout move QSET NOONTIME Part Number 878275 Vectorbeam Operator Manual Page 229 QSET NOONTIME Switch off ontime timers each requires a separate command DEFINE GROUP USER NOLOG VB LOGFILE ORIGINALFILE Reset original logfile as output QSET OUT OFF QSET OUT OFF QSET OUT ON EXIT ERROR 19 6 19 6 1 Pass the error status back to the parameter file Nested layouts It is possible to define a layout of a layout An example of
127. LAVE logicals not same as no of FISPs Part Number 878275 Vectorbeam Operator Manual Page 289 Reboot PG Reboot PG Reboot PG Reboot PG Reboot PG Reboot PG Reboot PG Reboot PG Reboot PG Reboot PG Reboot PG Reboot PG Reboot PG Reboot PG 27 6 6 Pattern generator error messages The following error messages are generated by the processors in the pattern generator and passed to Emma As a result Emma may generate and report its own error number and message as listed above The original processor error number is only visible on a terminal which is logged into that processor A message also appears in the job control window which is preceeded by PG SYS E FAULT or PG SYS E ERROR 27 6 6 1 Queue creation errors 0x400b0001 Couldn t create qErrors queue 0x400b0004 Couldn t create pattern input queue Reboot PG 0x400b0005 Couldn t create pattern output queue 0x400b0006 Couldn t create height sensor gt Root queue 0x400b0007 Couldn t create Root gt height sensor queue 0x400b0008 Couldn t create SuP Mgr gt Root queue 0x400b0009 Couldn t create SuP Mgr gt patWrite queue Reboot PG 0x400b000a Couldn t create SuP Mgr command input queue 0x400b000b Couldn t create PatWrite gt Root queue 0x400b000c Couldn t create Root gt patWrite queue 27 6 6 2 Task spawn errors 0x400b0101 Couldn t spawn Slave Manager task 0x400b0102 Couldn t spawn Height Sensor task Reboot PG 0x400b0103 Couldn t spawn Pattern Writing task 0x
128. LC operations on one board The FISP Flattening Instancing Sort Processors operation takes the structured and unstructured shapes from the CFPS Central Field Pattern Store on a field or part field basis Structured shapes are flattened into individual shapes and an instance of each shape for each subfield that the shape appears in is generated The shapes are sorted into subfield order and output to the Shape to Line Converter The timing and destinations for the output data are determined by the MUP The subfields are processed in a boustrophedon order The SLC Shape to Line Converter operation takes a part field buffer of shapes from the FISPs and places them in the input buffer of the shapewriter 4 3 1 2 3 3 DCP The DCP Digital Correction Processor applies real time corrections for field and subfield perturbations Part Number 878275 Vectorbeam Operator Manual 4 3 1 3 4 3 2 4 3 3 4 3 4 4 3 5 Page 42 The analogue crate transforms the digital scan and correction data into analogue signals 16 18 and 20 bit pattern generator operation The pattern generator receives the pattern data from the operator terminal i e Alpha workstation sequentially during exposure The pattern data are stored in either a FRE format file with extension FRE or a VEP format file with extension VEP file Such FRE and VEP format files will have been produced by the pattern converter software see Section Pattern data prepa
129. LEMENT 1 reset datum interval SS F ELEMENT 2 reset datum interval IFF TYPE HH NES INTEGER OR F TYPE MM NES INTEGER OR F TYPE SS NES INTEGER THEN QDISPLAY COMMENT RESET DATUM INTERVAL SYMBOL NOT CORRECT GOTO ERROR WLVD ENDIF Check syntax of time interval OPEN WRITE SDP MCLOG JOBFILE SDP LOG WRITE SDP START NEW SDP LOGFILE JOBFILE SDP LOG F TIME CLOSE SDP Create new separate logfile for periodic adjust datum using ontime RA AAGA AAA Part Number 878275 Vectorbeam Operator Manual Page 228 MCJOBS SDP COM FM FM reset datum interval ENDIF Set up ontime for resetting the datum CALIBRATE INTERVAL 0 32 CALIBRATE INTERVAL CALIBRATE INTERVAL F EDIT CALIBRATE INTERVAL COLLAPSE Create symbol calibrate interval in case one was not defined in the parameter file but keep any string if one was defined IF CALIBRATE INTERVAL EQS THEN QDISPLAY COMMENT NOTE THE SYMBOL CALIBRATE INTERVAL HAS NOT BEEN DEFINED QDISPLAY COMMENT NOTE THE DEFAULT NO REPEAT JOBCAL USING ONTIME WILL BE USED ENDIF IF calibrate interval NES 999 AND CALIBRATE INTERVAL NES THEN HH F ELEMENT O0 calibrate interval MM F bELEMENT 1 calibrate interval SS FSELEMENT 2 calibrate_interval IFF TYPE HH NES INTEGER OR F TYPE MM NES INTEGER OR F TYPE SS NES INTEGER THEN QDISPLAY COMMENT CALIBRATE INTERVAL SYMBO
130. PPERCENT Significant percentage of mark locates failed Check video levels and mark locate parameters give good reliability on good marks Check mark quality of array of marks being used LMPROWHOLE Data has consecutive failed locates looking along rows LNKHSTNCON Host is not connected LNKTIMOUT Link timeout for receive Ethernet packet Part Number 878275 Vectorbeam Operator Manual Page 279 LNKWNGHST Subsystem is already connected to another host LOADERROR Substrate load error LOCFAIL Locate mark failed LUMATDEC DW LU matrix decomposition error MAINDEFLIM Main field deflection limit exceeded MAPSIZERR Map size error MISSPKT Missing ethernet packet during block transfer MRKABORT Operator aborted mark locate MRKARITH Arithmetic error during mark locate MRKBDW Mark bandwidth MRKBEDGE Failed to find bottom edge Check the video levels and mark locate parameters Using SEM mode check that the mark is not damaged MRKCSLEN Mark cross coarse search length parameter out of range Reduce the coarse search length parameter MRKCSLIM Mark coarse search limit MRKCSOFF Mark cross coarse search offset parameter out of range MRKCSRCH Mark locate exhausted coarse search area MRKCT Mark contrast parameter out of range MRKCTT Mark contrast tolerance MRKDBFULL Mark definition data base full List the marks in the database using the qmark list
131. S is the x scale coefficient for the data for 0 degrees R yoo is the y rotation coefficient for the data for 90 degrees etc NOTES 1 Scaling If the measured array is too large then the output is a positive scaling error coefficient Uncorrected scaling errors should be small providing that the temperature control is good 2 Rotation Rotation error sense is defined as follows A positive X axis rotation error is an anti clockwise rotation of the X axis A positive Y axis rotation error is a clockwise rotation of the Y axis Uncorrected rotation Part Number 878275 Vectorbeam Operator Manual 15 9 2 15 9 3 15 9 4 Page 142 errors will usually be very large as they depend on the mechanical location of the mask plate in the holder 3 Orthogonality orthogonality error is not used in the Stage Position Correction process For our own purposes it is defined here as X axis rotation Y axis rotation So an acute angle between the positive directions of the X and Y axes gives a positive orthogonality error Uncorrected orthogonality error will normally be large and is due to the orthogonality of the zerodur mirror block It should however be very consistent 4 Keystone Consider the sloping sides of a square with keystone distortion to be the head of an arrow Then A downwards pointing arrow gives a positive X Keystone error A left pointing arrow gives a positive Y keystone error Uncorrected Keystone e
132. STIGSNOTFND Unable to find optimum stig settings STIGERROR Error occured during auto stig DEMAGFAIL Demag Table Calibration Failed CORBIG Calculated corrections within 1096 of maximum CORTOOBIG Calculated corrections greater than maximum SPCALFAIL Spot Table Calibration Failed MSLOPEFAIL Marker Slope Calibration Failed SPOTFAIL Failed to set requested spot CURRFAIL Failed to set requested current DOSEFAIL Failed to set requested dose DOSADJFAIL Dose adjustment failed CLKADJFAIL Clock adjustment failed CLKADJNEG Invalid unset nominal dose CMDNCOMPLETE Command not completed CMDABORT Command Aborted ADCFILEOPENING Could not open file for ADC plot ADCFILEERROR File error in file for ADC plot LOADERROR Substrate load error Vectorbeam Operator Manual 08019568 08019570 08019578 08019580 08019588 08019590 08019598 080195A0 080195A8 080195BO0 080195B8 080195CO0 080195C8 080195D0 080195D8 080195E0 080195E8 080195F0 080195F8 08019600 08019608 08019610 08019618 08019620 08019628 08019630 08019638 08019640 08019648 08019650 forced 08019658 08019660 08019668 08019670 08019678 08019680 08019688 08019690 08019698 080196A0 080196A8 080196B0 080196B8 080196CO0 080196C8 080196D0 080196D8 080196E0 080196E8 080196F0 080196F8 08019700 08019708 08019710 08019718 08019720 08019728 08019730 08019738 08019740 08019748 08019750 08019758
133. ScaleX To change the current value type for example gt double NonlinearRotScaleX double NonlinearRotScaleX To ensure the values are set when the DCP is rebooted put them into the DCP_CONFIG VW file for 16 bit systems or WFDCP_CONFIG VW for 18 bit and 20 bit machines In order to set and display the coefficients through Emma commands use the qset reg and qdisplay reg commands together with the numbers in the table below X Register Register Y Register Register for 16 bit for 18 bit for 16 bit for 18 bit PG and 20 PG and 20 bit PG bit PG NonlinearRotOn NonlinearRotOn NonlinearRotConstX NonlinearRotConstY NonlinearRotSquareX NonlinearRotSquareY Part Number 878275 Vectorbeam Operator Manual 10 9 10 10 10 11 10 11 1 10 11 2 10 12 Page 82 Heightmeter early read On 18 bit and 20 bit pattern generators the heightmeter is usually configured to do an early read This is where a heightmeter reading is started immediately after a stage move but during the stage settling time An early read reduces the time overheads and increases throughput The early read can be configured on the DCP by setting the variable as follows gt G_ExpectMoveSig 1 To ensure this value is set when the DCP is rebooted put it into the DCP_CONFIG VW file on 16 bit systems or WFDCP_CONFIG VW for 18 bit and 20 bit systems Height map readings by jobfile A simple height map consisting of a 3x3 ma
134. Signal Error X Interf Position Overflow Error Vectorbeam Operator Manual Page 295 Reboot stage Reboot stage Reboot stage Reboot stage Reboot stage Reboot stage Reboot Reboot stage Reboot stage Reboot stage Reboot stage Reboot stage Reboot stage Reboot Part Number 878275 0x400e000a 0x400e000b 0x400e000c 0x400e000d 0x400e000e 0x400e000f 0x400e0010 0x4006001 1 0x400e0012 0x400e0013 0x400e0014 0x400e0015 0x400e0016 0x400e0017 0x400e0018 0x400e0019 0x400e001a 0x400e001b 0x400e001c 0x400e001d 0x400e001e 0x200e001f 0x200e0020 0x400e0021 0x400e0022 0x400e0023 0x400e0024 0x400e0025 0x400e0026 0x400e0027 0x400e0028 0x400e0029 0x400e002a 0x400e002b 0x400e002c 0x400e002d 0x200e002e 0x400e002f 0x200e0030 0x400e0031 0x200e0032 0x400e0033 0x200e0034 0x400e0035 0x400e0036 0x400e0037 0x400e0038 0x400e0039 0x400e003a 0x400e003b 0x400e003c 0x400e003d 0x400e003e 0x400e003f 0x400e0040 0x400e0041 0x400e0042 0x400e0043 0x400e0044 0x400e0045 0x400e0046 0x400e0047 0x400e0048 Page 296 X Interf Stage Velocity Error X Interf Undefined Error Y Interf Measurement Signal Error Y Interf Position Overflow Error Y Interf Stage Velocity Error Y Interf Undefined Error Y Interf Reference Signal Error Y Interf Reference Not Present Y Interf Laser Not Ready X Interf Reference Signal Error X Interf Reference Not Present X Interf Laser Not Ready Y Interf Error Setting Destination X Interf
135. TH 0 0 P 0 0 100 0 ENDP ENDSTRUCTAlgorithmic programming Algorithmic programming is an alternative method of defining the pattern to be exposed by means of a program which generates each shape concurrently with the exposure No pattern file is required This is useful when 1 The pattern can be easily generated mathematically for example a spiral 2 The pattern would require a large file for example those containing curved shapes will be fractured by the converter into many trapezoids of various sizes 3 Many similar patterns are required which can be generated mathematically by varying only a few start parameters for example gratings with different pitches For more information see the Vectorbeam Algorithmic Programming Software User Guide part number 878612 Part Number 878275 Vectorbeam Operator Manual Page 267 25 Routine maintenance and servicing Service operations should only be carried out by personnel trained by Vistec engineers Generally only Vistec engineers will carry out service operations However a limited set of service and maintenance operations may be carried out by suitably trained customers based on the guidelines described in Vectorbeam Customer service procedures manual document number 893116 D Electrical and mechanical hazards exist when carrying out these WARNING maintenance operations They are intended to be carried out only by personnel trained by Vistec The table below de
136. This is useful for recording images of resolution samples or marks The detectors can be either the backscattered or transmission detectors As an example the following command scans a 10 x 10 um area VB_OPER gt qimage size x 10 y 10 skip x 4 y 4 filt 8 points 8 frames 1 file 10um_are a The image can be viewed on screen by using a PVWave program VB_OPER gt wave r qimage utility This program reads the img file which is output by the pattern generator and allows the image to be written as a tif or postcript file Part Number 878275 Vectorbeam Operator Manual 24 24 1 24 1 1 24 1 2 Page 251 Advanced operation Calculating throughput Calculating the pattern time is necessary for scheduling jobs and for optimising the throughput Total pattern time Beam on time main field deflector settling time subfield deflection settling time shape synchronisation time stage movement time stage settling time heightmeter measurement time data processing time alignment mark registration time load direct write map time A jobfile which does this calculation can be run as follows VB SUPER run vb seq vb exposure time exe The jobfile vb exposure time exe is documented in the manual Acceptance tests and operator jobfiles part number 892777 Beam on time The beam on time total number of exels clock frequency total pattern area to be exposed beamstep size x beamstep size cloc
137. UT RECOVERY DOSE RECOVERY DOSE RECOVERY MOVE 0 32 RECOVERY MOVE RECOVERY MOVE F EDIT RECOVERY_MOVE COLLAPSE IF RECOVERY MOVE NES THEN LAYOUT RECOVERY MOVE RECOVERY MOVE RECOVERY PATTERN 0 32 RHECOVERY PATTERN RECOVERY PATTERN F EDIT RECOVERY PATTERN COLLAPSE IF RECOVERY PATTERN NES THEN LAYOUT RECOVERY PATTERN RECOVERY PATTERN Sets the display and error recovery behaviour Symbols are created in case they weren t defined for backwards compatibility WRITE SYSSOUTPUT LAYOUT INFO WRITE SYSSOUTPUT Types the current layout command settings IF LAYOUTCENTRE EQS RELATIVE THEN QMOVE SPO RELCENTRE ELSE QMOVE POS CENTRE ENDIF QDISPLAY HEIGHT TAB SUBSTRATE TYPE Set the height meter to the table for the substrate jobcal sets the table to 7 RESET DATUM INTERVAL 0 32 RESET DATUM INTERVAL RESET DATUM INTERVAL F EDIT RESET DATUM INTERVAL COLLAPSE Create symbol reset datum interval in case one was not defined in the parameter file but keep any string if one was defined IF RESET DATUM INTERVAL EQS THEN QDISPLAY COMMENT NOTE THE SYMBOL RESET_DATUM_INTERVAL HAS NOT BEEN DEFINED QDISPLAY COMMENT NOTE THE DEFAULT NO DATUM RESET USING ONTIME WILL BE USED ENDIF IF RESET DATUM INTERVAL NES 999 AND RESET DATUM INTERVAL NES THEN HH F ELEMENT O0 yreset datum interval MM F E
138. YOUTCENTRE EQS RELATIVE THEN QMOVE POS REL SPO RELCENTRE 0 0 Create position symbol containing current position if the layout is defined as being relative to current position This is for nested layouts QSET CORR ON ALL Makes sure that all the corrections will be applied QSET HEIGHT REALTIME Makes sure that the heightmeter is working in real time mode QMAP DWMODE ABSOLUTE LOAD Makes sure that exposures are carried out in absolute mode and not mapped mode QSET SORT NORMAL Makes sure that the exposure of shapes is carried out subfield by subfield QSET VRU WORKFILE VRU Sets the vru to workfile vru defined in the parameter file WORKFILE F EDIT workfile UPCASE IF WORKFILE NES GRATING AND WORKFILE NES NONE Grating and none are options for the qset pattern command that don t require pattern files THEN Deal with possible VEP or FRE file and any underlapping or overlapping with FRE file WORKFILE F EDIT WORKFILE COLLAPSE Underlapping or overlapping specified only used for FRE files IF FSELEMENT 1 workfile NES THEN WFILE VB PATS f element 0 workfile FRE f element 1 workfile No underlapping or overlapping specified default to VEP but if not present then select FRE IF FSELEMENT 1 workfile EQS THEN WFILE VB PATS f element 0 workfile VEP IF f search WFILE eqs THEN WFILE VB PATS f element 0 workfile
139. _C THEN GOTO FINISH_ALL ON CONTROL_Y THEN GOTO FINISH_ALL Defines operation when jobfile is cancelled TYP F ENVIRONMENT PROCEDURE Types parameter file so it can be checked SJOB FSENVIRONMENT PROCEDURE Types up name of this parameter file in Emma status window SJOB sets the logical vb_jobfile which is used later to automatically create a logfile with the same name as the parameter file G EEEE AARAA ERRERA pae SEE End header kkkkkkkkkkkkkkkkkkkkkkkkkkkkkkk HOLDER H_069K Defines which holder initialisation sequence is to be carried out If 999 is given then no initialisation sequence is carried out CALIBRATE Determines whether jobcal is carried out If 0 is given then no jobcal is carried out unless a beam is defined to be loaded see below WORKFILE DEVICE_LAYER_1 Defines the name of the pattern file to be exposed No FRE or VEP extension The pattern file must be in the VB PATS directory WORKFILE_VRU Defines the VRU STARTDOSE 14 Defines the start dose in uC cm DOSESTEP Defines the dose interval between successive dies if the operator see below is either or If the operator is 0 then each die is exposed with the start dose and the dosestep is not used OPERATOR cell Defines the method of changing the dose within a layout and comprises of two parts The first part of the operator can be either 0 or The second part of the operator is opti
140. a long measurement time due to both reasons Height meter warnings and error messages The height meter subsystem returns a quality assessment with each reading This results in Emma reporting one of the following Nothing This implies a good height value 2 Laser height sensor reading poor 3 Laser height sensor brightness at limit 4 Laser height sensor over range For error recovery suggestions see Section Emma error messages in Chapter Recovery from exception conditions Height meter offsets Height meter offsets can be applied in order to compensate the for resist thickness effects and to adjust the field scaling if required On UHR systems a height meter offset will also result in a change in field rotation A more user friendly way to adjust scale and rotation is to use the calibration offsets see Section Calibration offsets in Chapter Calibrations and this is the recommended method Resist thickness compensation By considering the optics of the heightmeter an estimate of the position of measurement relative to the surface of the mask can be made The following examines the case of PMMA on 85 nm thick Cr The wavelength of the laser diode is about 780 nm The reflection coefficient from the PMMA surface is about 0 28 The beam transmitted into the PMMA with relative intensity 0 72 is reflected from the Cr The reflection coefficient for this is 0 57 giving a beam with relative intensity 0 41 At t
141. able contrast is produced over the entire area to be used later for exposures The marks should have a pitch of around 0 5 to 2 mm as although measurements are typically carried out every 10 to 20 mn if a mark is damaged the VB has the ability to use the nearest neighbour to complete the calibration The absolute positioning accuracy of these marks it not important as long as they are within the mark locate coarse search range The magnetic map calibration is then carried out in the usual way using this substrate Save the magnetic map see Section Database management separately so that it can be used in combination with any other set ups 5 When calibrating the magnetic map use should be made of the facility to use neighbouring marks if a particular calibration mark cannot be located This will enable the calibration to complete successfully on a less than perfect array of marks Use the qualifier alt_mark_grid 0 5 for marks on a 0 5 mm grid Calibration jobfile The required sequence of commands is contained in the file cal_magmap com 1 Locate two marks on the autostitch plate one at about the lower left corner and one close to the centre of the array of marks for the calibration and store their positions in the identifiers M1 and M5 respectively 2 Type VB OPER QG vb seq cal magmap com 3 Respond to the prompts and the calibration will be carried out Two iterations will be performed The residual errors of the final meas
142. actly on axis By typing VB_OPER gt VID_H the automatic video level will be switched off and the manual gain level is set to that last found by the automatic video level adjustment The backoff will need to be manually set to achieve a background level of around 10 or higher UHR machines There is one function that automatically adjusts the gain and backoff together called peak The video gain and backoff can be switched into automatic adjustment mode by typing VB_OPER gt VID_P This command switches the machine into SEM mode and selects the automatic peak level adjustment using the command VB OPER qset video vmode peak mark 0 8 This command is the same as selecting the peak switch in the Set Video Level panel Figure 7 9 Auto peak level will adjust the gain until the peak video level reaches the value set either in the Set Video Level panel next to the peak option or by the mark command option usually about 7096 8094 It also adjusts the backoff until the background video level reaches the value set either in the Set Video Level panel next to the mean option or by the bckgrd command option Video Calibration for UHR machines Machines with a UHR final lens have a silicon 4 quadrant PN junction backscattered electron detector and a video amplifier with 4 coarse gain ranges The video amplifier has the complication that it is necessary to adjust the backoff level as the gain changes To facilitat
143. ade deca ads ci n ad cesa c dese claeteg duci ce our Er 35 Sils Va cuumsystem MORTON arasina kkaa ii ahaaa esai E AN RAA 35 94d Stage GallgeS cetdi ite nce iteriocten eitis tate eaa ttai femora 35 ode Vacu m dali eS o titan eei iine echan aaae aaant 35 en ks MM Noo 35 4 Computer SUHIDBE ida Spon FREE EM DEI EM RM NME M MR MEN f 4 1 Pattern data preparation GORDLIIBrus iio auca iir cid dd etii duxi deeds sut ced dux adds 37 4 2 scudo o Lo LESE 37 4 3 Computer controlled SUDSySEOlTIS casas i ioa dea i tla ela qc Freue ua ERR ires deae 38 4 3 4 Pattern genero asoan eroe Fees eta s onim eis spaces since 38 4 3 2 Stage COMOT uices atlas pro E EE e pee tr pas ot era cium eiae deed 42 43 9 Omas Conttollel 5 eir pao ela doo o Fes kai raa OR ORE Masa 42 434 PIGS ICOMMONEN siietiesscesaccacay asin idus taie S 42 4 3 5 Logging in to the subsystem controllers eese 42 5 Machine start up from cold eee petto tana sp a aap step senda roin D 6 Machine shut CO ae Te Emma control software sssossssseesssssessssscosssssssssssseesssssssssssseesssss E7 Tals IDE aieo nda ire ada exi dba nein Sura cLRIM E MU E I 47 2 2 Emma NIDODUS OVOFVIBIU searing asia Guida reda ea aia Fel du Feds Naa MUR Ra ad 48 Pasi SUMMON Sadacctdinssisnetsiss EENE tutta dua Uto tench toda Skala ufa Lou dta Hd ARE du 50 T22 Display IROIU tice deacons ieee eee iut dec a ARE 54
144. ages for similar spot sizes 2 Electrons deposit most energy into the substrate at the point they stop For very sensitive substrate layers the appropriate beam accelerating voltage depends on the depth of the layer Advantages of lower voltage 1 Larger maximum fieldsize available which enables higher throughput 2 More signal contrast for locating marks 3 Can obtain more undercut of resist sidewall which aids lift off 4 Damage see above Advantages of higher voltage 1 Smaller spotsize available Part Number 878275 Vectorbeam Operator Manual 15 3 2 15 3 3 15 3 4 15 3 5 Page 133 2 Higher beam positional stability 3 Backscattered electrons spread out over larger area which can reduce proximity effect for some patterns 4 Thicker resist can be exposed with near vertical sidewalls 5 Damage see above Important notes before increasing the beam accelerating voltage kV Before increasing the accelerating voltage the following conditions must be met 1 The pressure in the gun is within the range for operation 2 f the gun has not been operated at the target voltage since it was last vented to atmospheric pressure then it will require conditioning see below Conditioning the gun It is vital that the gun is conditioned after it has been vented before normal operation as otherwise a flashover could occur which may destroy the cathode This is carried out by the Vistec Engineer Setting
145. al 11 10 2 Page 101 position of a horizontal limb This scan is of length twice the cross height If nothing is found then a second attempt is made but this time the vertical scan is carried out on the other side of the expected position If the vertical scans fail to confirm the presence of the cross then a mark not found error is returned If the confirmation scan finds the horizontal limb then two further coarse scans are made about the bottom and left limbs to measure their positions also Each scan checks that the limb width is correct and measures the video levels of the cross and background and the edge width The video levels and edge width values are used to calculate MLEN RT and CT which are passed to the fine search The edge positions found by the coarse scans are also passed to the fine search routine CSLIM CSLEN CSOFF W Figure 11 12 Diagram showing coarse search raster scans for cross mark Fine search If the fine search is performed first the fine qualifier was used then RT CT and MLEN are taken from the mark definition otherwise these are supplied by the raster search A series of parallel scans is made about each of the observed edge positions from the raster search or about the expected positions if no raster search was done because the fine qualifier was used The number of scans is given by PARASCANS and the length is given by MLEN The length of edge ove
146. al 27 6 7 7 Part Number 878275 0x400e0049 0x400e004a 0x400e004b 0x400e004c 0x400e004d 0x400e004e 0x400e004f 0x400e0050 0x400e0051 0x400e0052 0x400e0053 0x400e0054 0x400e0055 0x400e0056 0x400e0057 0x400e0058 0x200e0059 Vacuum Errors 0x400f0001 0x200f0002 0x200f0003 0x400f0004 0x400f0005 0x200f0007 0x200f0008 0x400f000d 0x200f000e 0x400f000f 0x400f0010 0x200f001 1 0x400f0017 0x400f0018 0x400f0019 0x400f001a 0x300f001b 0x400f001c 0x400f001e 0x400f0025 0x400f0026 0x400f0027 0x200f0030 0x200f0031 0x200f0032 0x400f0033 0x400f0034 0x400f0035 0x400f0036 0x400f0037 0x400f0038 0x200f0039 0x200f003a 0x400f003b Ox200f003c 0x200f003d 0x400f003e 0x400f003f 0x200f0040 0x200f0041 0x200f0042 0x400f0043 0x200f0044 Page 297 Yaw Interferometer Stage Velocity Error Yaw Interferometer Undefined Error Yaw Interferometer Reference Signal Error Yaw Interferometer No Reference Signal Yaw Interferometer Laser Not Ready Yaw Interferometer Error Setting Destination Yaw Position Error Card Initialisation Error Too Many Retries Stage at Travel Limit Too Many Retries Stage at X 2nd Limit Too Many Retries Stage at Y 2nd Limit Too Many Retries Moving Stage in X Too Many Retries Moving Stage in Y Stage settling Time Not Found BEF Rollover Error in X BEF Rollover Error in Y Error finding limit difference Function not valid for current stage HW Error receiving from qVac queue Abort arrived too late In
147. al Emma and DCI commands Multiple layouts can be exposed by one parameter file by repeating the above sequence in the same file The parameters can be thus redefined and must be followed each time by the line running the file wlvd dw com Machine preparation for direct write See Section Machine preparation Substrate preparation for direct write See Section Alignment of substrate for direct write Expose Type filename com The correct sequence of operations to expose the layout will be carried out assuming that the parameters are correct by the file wlvd dw com Wlvd dw com Issues commands to expose the layout defined in the parameter file This file is similar to WLVD COM except it contains some additional items to carry out global alignment and set up the die by die alignment parameters Part Number 878275 Vectorbeam Operator Manual 21 21 1 21 2 Page 243 Remote operation The most common requirement when the operator is located away from the machine for example outside the cleanroom is to check the status of a job which has been started locally The Vectorbeam can however be operated from any computer running a terminal emulation program which can connect to the Vectorbeam operator terminal For example it is possible to operate a VB which is networked at its installation site with a computer with an internet connection from a PC which can be located anywhere in the world connected via a
148. alignment methods are possible and the best method will depend on the application The differences between the methods relate to the number and proximity of marks used per exposure Regular rectangular arrays of marks are the ones most commonly encountered and a template jobfile which supports such arrays is described below and supplied as part of the Applications jobfiles release However the underlying Emma mapping command can operate with marks which are not on a regular grid and other template files are available which allow for irregular grids irregular layouts irregular sequencing and alternative marks Part Number 878275 Vectorbeam Operator Manual 20 1 1 20 1 2 20 1 3 Page 232 Global alignment Global alignment is normally always carried out as the first step in any alignment scheme Global alignment may also be the only alignment step Global alignment is faster because only a few marks are located before exposing all the substrate but consequently less accurate than die by die alignment Die by die alignment Die by die alignment is slower but more accurate than global alignment As several marks close to each die are aligned localised substrate or pattern distortions will be taken into account and this method is therefore more accurate Direct alignment commands on the VB The implementation of all the various methods involves the appropriate use of just two commands 1 The first command is the generation o
149. am Om UME sesiis aa aaa adaa a aaaea 251 24 1 2 Mainfield deflector settling time sessseseeesss 251 24 1 3 Subfield deflector settling time sesesessssssss 252 24 1 4 Shape synchronisation TIME i uiii cccs sacco ttai toa teat abet tbt eoe 252 24 1 5 Stage moVvemiehlL MMe misrin cte te seine tit ente tcu baa cce tta depu iaze 252 213 1 5 Stag seting ME sissantina anen dataan 252 24i THEIGIMUMELCN UME naait 252 241 8 Data processing TIME siaradai naiinitan 253 24 1 9 Die by die alignment time cece eeececeeeeeeeeeeeeeeeeeeeeeeeaeseeneeenenees 253 24 2 Displaying and setting the settling times sesessseeeeseeesse 253 24 2 1 Mainfield deflector settling times essssseesses 253 24 2 2 Subfield deflector settling time ssssssseesss 254 24 2 9 Stage seting NE naianei 255 24 3 Under or overlapping mainfields sparse tiling ssssssss 256 24 4 Random field placement sparse tiling seseesesesseeeesss 256 24 5 Selectable field correction interpolations eeeseesseeeese 257 246 ODUNISIBU EDFOUG IL sacado kid asp oto RR RA HG dasha Ra ER Ku d ton dog ook Sa drain 257 24 6 1 Increasing current and spot size sesesssssseeeeee 257 24 6 2 Reducing command processing times
150. amber when the electron beam interacts with some customer substrates e g resists D CAUTION Customers must assess the potential hazard from any such process and take appropriate ventilation precautions if required Base Process chemicals used The base process of the VB6 utilises a silicon substrate coated with resist The resists developers and solvents that are utilised in the base process are listed in the table below PPE Personal Protective Equipment required for using these chemicals are described in the hazardous substances information chart manual no 931447 The PPE requirements in this document must be adhered to when handling the chemicals described in 931447 The chemicals used in the base process do not produce any hazardous by products when used in the VB6 tool Any additional chemicals introduced to the system should be assessed by the user for hazardous emissions Chemical MSDS Number NEB22A2 amp 31A2E MSDS 5000911 A5 amp A7 PMMA 495 MSDS 5000844 Electron Beam Resist 950 PMMA A11 MSDS 5001283 EBR 9 Ebeam Resist MSDS 5000903 HSQ Fox R flowable oxide MSDS 5000919 Chrome etch MSDS 5000416 LDD26W Developer MSDS 5001119 Microposit MF CD 26 MSDS 5000907 IPA Propan 2 ol MSDS 425069 Methyl iso butyl Ketone MSDS 425157 Acetone MSDS 425138 List of chemicals used in the base process Contact Vistec for datasheets MSDS Material Safety Data She
151. an electron beam sensitive resist which is exposed with the pattern and subsequently developed A focused beam of electrons is used to write the patterns principally by deflection over a limited area of the substrate known as the main field a moveable stage supports the substrate enabling exposure to be made over the range of the stage in field by field step and expose manner Several specialised detector systems are used to control this process including sensors for substrate position x y and z beam current and beam position All user control of the system is provided through an HP Compaq Alpha workstation An off line software capability is used to transform CAD generated pattern data into a machine readable format After job control preparation loading of pattern data for a number of substrates and loading of the physical holder with the substrates the machine will execute the writing process fully automatically For the purpose of this manual beam establishment spot formation deflection substrate positioning detection systems computer software and support systems will each be referred to as system each system may contain several sub systems Each sub system consists of several hardware modules often distributed throughout the machine and or a number of supporting modules After the beam has been formed and focused it must be moved scanned over a substrate using a beam writing technique The two basic beam writing techniques
152. and also displays the frequencies which are set in the pattern generator If these differ from the frequencies computed above a warning is printed that the clocks need adjusting Beam current The beam current used for the qadjust clock command is shown in the Emma status window and will be the last value produced by any of the following 1 The measured beam current using qdisplay bcm 2 The beam current set and measured by qset current using the spot table 3 The beam current set and measured by qset diameter using the spot table 4 The beam current specified for qadjust clock with curr or bcm Part Number 878275 Vectorbeam Operator Manual 13 1 7 13 2 13 2 1 Page 119 Exposure grid The exposure grid size used for the qadjust clock command is the beamstep size described in Section Beam step size and VRU in chapter Exposing a substrate Dose controller The dose controller provides a band of exposure frequencies or clocks All shapes in a pattern are tagged with a number clock in the pattern file which defines which one of these frequencies is to be used for exposing that shape If no explicit assignments are done in the converter then all shapes are by default tagged with the same clock 0 and all the pattern is exposed with the same dose However each shape can be exposed with a different clock which is useful if proximity correction is required see Section Proximity correction The dose con
153. and e g VB OPER qset clock 0 25000 VRU 2 This gives more flexibility when exposing patterns with both small and large shapes with a range of doses 19 1 10 Choosing the spot size beamstep size The spot size and the beamstep size are related by Spot size Beamstep size x A Equation 19 16 where A can be thought of as an overlap factor Typically A is in the range 1 to 1 5 for suitable overlap making the spot size approximately equal to the beamstep size If A is too large then the shapes will be too large If A is too small then the overlap will be insufficient giving rough edges stamp effect The spot size is related to the minimum dimension in a pattern by Beamstep size minimum dimension B Equation 19 17 where B can be thought of as a pattern fidelity factor Typically B is taken as 4 or 5 B is chosen based on a compromise between exposure speed and pattern fidelity If B is increased more scans smaller spot then the pattern fidelity will improve as the corners of shapes will be less rounded In addition variations in dose due to proximity effects and variations in the development conditions will have a smaller effect on the final dimensions If B is increased larger spot the pattern fidelity will suffer but the exposure time will decrease In order to minimise the grid snapping when shapes are made to be an integer number of beamsteps in size it is necessary to know the design grid If you have designed t
154. ap Field X Y Emma 0 000 kV 0 00 um 000 000000 X 00 000000 3 00 000000 3 00 000000 3 00 000000 3 00 000000 3 00 000000 S 000 000000 Y mm 00 000000 R 00 000000 R 00 000000 R 00 000000 R 00 000000 R 00 000000 R 00 000000 K 00 000000 K Cancel Figure 7 16 Display Database Calibration The Display Database Clocks panel Figure 7 17 displays current clock frequencies and the upper and lower limits of the dose controller band 400 000 KHz 400 000 KHz 400 000 KHz 400 000 KHz 400 000 KHz 400 000 KHz 400 000 KHz 400 000 KHz 396 040 KHz 404 040 KHz Figure 7 17 Display Database Clocks 7 2 3 Toolkit menu The toolkit is also available from the menu in the Status Window Select with the mouse left button toolkit from this menu keeping the mouse button pressed down the tools available will be displayed When the EO monitor is selected Figure 7 18 the Column Status Panel is displayed which contains the Set Demand and Measured values for the EHT Filament Wehnelt and the Lenses It is continuously updated Part Number 878275 Vectorbeam Operator Manual Page 58 EO Monitor Demand Measured EHT Filament Wehnelt Figure 7 18 Tool Kit EO Monitor Panel When the PG monitor is selected the Pgmonitor Status Panel is displayed Figure 7 19 this displays the progress of the pattern data through the pattern generator during exposure It is conti
155. apping has a lens map assigned to it qnove pos will move to the corrected position map however the position displayed by Emma will the nominal one This is also true of the command qmove spo If the direct mode of the position symbol has a map assigned to it then the Part Number 878275 Vectorbeam Operator Manual Page 191 stage actually moves to the corrected position but the displayed position is the nominal one The lens map used is the one assigned to the mapping mode in the position symbol not the current mode If the lens map assignment to that mode has changed since the position symbol was set up the new assignment will be used The commands qmove home and qmove load always move to the true positions as this is essential to avoid mechanical damage Mark location will also return nominal positions although the position of the beam is corrected If the stepper map is accurate therefore the measured position of a mark in a stepper field will show no distortion 2 Lens maps cannot be assigned to absolute direct write mode This is because machine calibration is done is absolute mode A calibration with a lens map inadvertently assigned is likely to give wrong results so the possibility has been removed 3 Lens maps have limited areas of validity They are valid over the area of calibration from the location they are assigned to the full area of stage travel Outside this area the extrapolation of the corrections may give erroneous resu
156. arch length for crosses directly Due to the Offset for crosses CSOFF the CSLIM parameter as applied to crosses only places a limit on the total of the offset CSOFF and the coarse search CSLEN Mark location command The mark locate command requires a mark name specifying the parameters to be used to locate the mark and a name for the position identifier into which the position will be written Qualifiers may be used to influence the locate operation Refer to the Vectorbeam Command Set Manual 878274 for details of the syntax and qualifiers Mark administration functions Typing qmark list on the Emma terminal will display to the user the list of marks currently stored in the mark database Typing qmark delete mark type name will cause that mark to disappear from the mark data base The qmark show command is used to display the definition of a particular mark Type VB OPER qmark show mark type name Refer to the Vectorbeam Command Set Manual 878274 for details of the syntax and qualifiers Diagnostic output of the mark locate function Once the mark location routine has completed a value will be returned in the position identifier specified indicating the centre of the mark If the location routine has failed then an error message indicating the cause of the failure where identifiable will be output together with the highest contrast reading observed in the scan Part Number 878275 Vectorbeam Operator Manual 11
157. are correct Check that the autostigmation parameters are correct Adjust the stigmation manually and retry STIGSNOTFND Unable to find optimum stig settings Check that a normal SEM image of the mark can be obtained Check that the mark position and locate parameters are correct Check that the autostigmation parameters are correct Adjust the stigmation manually and retry STNATMODINV Natural mode is invalid for this command STSRCMAPINV Source map mode invalid in ConvertStagePosns The number that identifies the Stage mode is not in the range 0 to 2 Internal consistency check has failed Try reloading the Stage mode Try restarting Emma Contact Vistec if it persists SUBBUSY Sub system Busy Wait for sub system to finish previous command if required If the sub system will require too long to finish type CTRL C to exit from any jobfile and press the abort button in the Emma status window If the sub system does not respond or has stopped working reboot the subsystem and then type start at the VB OPER prompt SVD2FEWROWS Can t solve SVD matrix has fewer rows than columns Check for sensible command parameters Contact Vistec SVDDIVZAII Division by zero in SVD matrix arithmetic Aii Check for sensible command parameters Contact Vistec SVDDIVZAIL Division by zero in SVD matrix arithmetic Ail Check for sensible command parameters Contact Vistec SVDDIVZH Division by zero in SVD matrix arithmetic h
158. as in the following example my_program vb disk vb users my_directory my_program exe 2 entering the symbol my_program at the VB_OPER prompt When calling a compiled jobfile from a command file the line define user mode sys input sys command must precede the line my program Part Number 878275 Vectorbeam Operator Manual 9 1 Page 67 Corrections Corrections for absolute accuracy In order to focus and place the beam accurately in an absolute sense on the substrate various corrections are required These corrections are described below Stage mapping absolute mode Correction coefficients are applied to the stage to remove inaccuracies in the measured stage position due to mechanical limitations in the interferometer mirrors This enables the positioning of patterns closer to true grid a perfect x y Cartesian coordinate system Correction coefficients are determined either from measurements with an appropriate metrology tool of an array of marks exposed with the VB or from measurements from the VB of an array of marks The coefficients are applied to the absolute mode See Section Stage mapping modes Beam error feedback BEF correction The beam error feedback removes discrepancies between the desired stage position and the measured stage position There is typically a discrepancy of a few microns after each stage move due to mechanical limitations The laser interferometer measurements are used t
159. at 100 kV 28 um at 50 kV and 28 um at 20 kV For example 64 subfields with a maximum blocksize of 0 8192 mm and a Maximum Fieldsize of 0 8192 mm gives a subfield size of 12 8 um which is within the deflection limit and 1048576 64 16384 bits which is within the bits limit Another example would be 32 subfields with a Maximum Blocksize of 0 262144 mm and a Maximum Fieldsize of 0 524288 gives a subfield size of 8 192 um and 16384 bits Part Number 878275 Vectorbeam Operator Manual 19 1 9 Page 210 The Maximum Blocksize set on the Vectorbeam must be larger or equal to the pattern block height and width otherwise the exposure cannot be carried out There must be an integer number of bits in each subfield and this means that the result of dividing the number of bits in the Maximum Blocksize by the number of subfields must be an integer Both examples above meet this requirement but 48 subfields with a maximum blocksize of 0 4096 mm and a maximum fieldsize of 0 8192 mm would give 10922 66 bits per subfield which is not allowed In summary the following should be taken into account when choosing the subfields and the maximum blocksize 6 There must be an integer number of bits in each subfield 7 The maximum subfield deflection must not be exceeded 8 The maximum blocksize must be larger or equal to the physical blocksize required by the pattern 9 A larger number of subfields e g 64 may give higher accuracy but will be slowe
160. at the nearest subfield to the end of the previous line This reduces the total deflection settling overhead by replacing long subfield jumps with short ones For comparison the existing nosort mode always starts the exposure of a line at the subfield closest to the bottom left corner of the block Note that whether the number of passes making up each line is odd or even which determines whether the scanning of each line within a subfield ends adjacent to the next subfield is not relevant 2 Expose a specified number of sub shapes before moving to the next subfield This means that a number of lines will be exposed a group completely before starting the next group If the specified number is less than the total number of lines crossing a subfield the subfield will have to be revisited For each subfield the sub shapes exposed in a group before switching to the next subfield must come from the same set of shapes Part Number 878275 Vectorbeam Operator Manual Page 263 24 10 until those shapes have been completely exposed If after exposing several groups of lines in the same row or column of subfields the remaining number of lines is less than the specified number only these lines will be exposed as a group The next group of lines in the adjacent row or column will contain the specified number again The functionality described at points 1 and 2 above is configurable to be the same for both for X and Y gratings The command to
161. ation for direct Wite ussiisa 241 EXUDSS T ud num Et MM cU UI n 241 Remote operation ioioeiees pd itp ie pin In i pae ER eI 0d M EU pa Soo M n ps IS ne a Uu uS a 21 1 21 2 21 9 21 4 21 5 Checking status remotely asuncion 243 Obtaining the oper control DFOITIDIu aei coc sicci ane cott acera usar as erret 243 Login without USING Second NICENSS ieri tie et ten nera bases exi dea asa ines 244 Batch gieus operaio tI 244 Re booting the subsystems remotely sss 244 21 5 Paten geleralof seii tere ve cd ito ede bt Eus 244 21 52 Stageiicconsistietiie nein ien tana i itai es 244 21 59 JEO enudieaon inier en iiu en etes 245 Logfiles and loggmg qM vd 224 22 2 Emma ecd a b etna Terre tren hey edad id de rer eer 247 CRANE 247 2212 DEG ISMN Ss nce eerta ut puer iSo tiende ades adito eRa des 247 Notes oM OpenVMS AOJO es wud a unes vena na ceils aiden Sd ceils sca fau na e dua kk n du Edo 247 Vectorbeam Series Vectorbeam Operator Manual 22 2 1 DECTerm logging during batch queue operation 248 22 2 2 DECTerm logging using the SET HOST command 248 22 2 8 DECTerm logging using the Options menu ss 248 23 Creating scanning electron image files 249 24 Advanced Operation Me 9 L 24 Hed DIO sarsies aa ania aada 251 25 4 1 Be
162. aximum video level 140 Uu o9 4 ideo Intensity ADC units Vi E o TII p na Minimum video level 50 IL cos En a iore T3 Yr TT 9 Erici a dora qo cese caa pad 4c 4c pagg Diva icr cp pai gas 2 80 2 2 2 2 2 2 20 Deflection X um File VB DAT ADC_DUMP_0B_54_13 MAY 2005 dat Figure 11 3 Example plot of mark locate ADC data Alternatively the mark contrast can be found by locating with increasing contrast values until the locate fails Minimum contrast CT should be set to roughly half the actual mark contrast However it can be increased to close to the actual contrast of the marks if more immunity from contamination or other nearby features is required For low contrast marks the minimum contrast may be reduced to 0 05 although the backoff should be decreased first in order to Part Number 878275 Vectorbeam Operator Manual Page 93 increase the actual contrast 2 The rise time RT in spite of the name is a distance and is typically set between 0 1 um and 0 5 um The rise time will be approximately the sum of the following a The mark edge sharpness which is the edge slope multiplied by the thickness of the mark This is the distance between where the edge starts and where it finishes usually lt 0 1 um b The spot diameter usually 0 1 um Reducing the rise time below the sum a b will reduce the contrast seen by the mark locate algorithm and ma
163. base to enable them to be recalled the next time that the field size is required See Chapter Databases This removes the need to run the fullcal sequence each time the fieldsize is changed The results for the three sets of field corrections remain valid until the column is disturbed mechanically e g for column cleaning A beam current in range 2 20 nA should normally be used for fullcal The program is started by typing VB_OPER gt fullcal Jobcal By locating a calibration mark at different stage positions the translational rotational and scale errors of the beam deflection with respect to the stage can be measured and compensated in order to accurately match beam and stage co ordinate schemes This enables writing of patterns extending over many fields by stitching them together Many parameters relating to the deflection and focusing of the beam will vary with beam current In addition various components of the machine are liable to drift which may affect these parameters significantly after a period of a few hours The machine has various calibration routines to calibrate the deflection and focusing of the beam and these have been sequenced in a program called jobcal Jobcal should be carried out after changes in beam current and at regular intervals of a few hours It is also recommended after a new holder has been loaded on stage The program is started by typing VB OPER jobcal If the main field calibration is done it is V
164. bbu huge tka Es 105 12 22 otaldalabasQ iii ei tre n attese RR ces eee tera pa npe IR BRE 106 12 2 9 koplevel partial databases etr eontra eb ese xha sss 106 12 2 4 Bottom level partial databases essssssseeese 107 125 Dreba o SQUID oun Cerca nto Gu EA 111 led Julabasedsniig auch oig sevice Idus e A a Rd EE 111 12 5 Database management iic orienta e Drki dad dictt as rU E dE pn dla dent dee nda 111 125 1 Recommended Scheme sesioan a eene 112 125 2 Basio database GCNeratlo my accsstere hs cactizeseenateseectsanadvexicaeteaaieeusiee 112 12 5 3 Field corrections database generation eesssuss 113 12 5 4 Beam database getneratlon ici cot tree ttc paac e 113 12 5 5 Magnetic map database generation esses 113 12 6 Database selection prior to exposure ssesessssseeseeenene 114 12 7 Deflection field corrections confidence check sssssssssss 114 12 9 J OBDIXDITR d a oodd Handel Hines Finca i datis Hold edd adip 114 IET rl RR E Wide Area OS eiim ditio hei odieso e Lediesu isotope sedi eso tou cuam Peu df ocr Eie dE 117 419 52 duned0s65ccsoe ona De mE M MILL IEEE 117 Dee SEN POMEOOS urona TES 117 19 1 4 Resist Sensitivity pararmieter iiie te pe ect ieee 117 q3d5 U IDOKGU i etisicdes iio ceeet a t tandecacecGtaces ti acddac sais siadcaressdincceazns 118 19 4 0 Beam CUI TG iier ioc
165. be replaced with MARKER directoryljobfile_1 com which causes the jobfile to be run at each die alignment mark instead of a mark locate Jobfile_1 com could look as follows OPEN APPEND SHARE POST EMMA MAILBOX EMMA Set error status code to success Error 1 Locate alignment mark as usual qlocate square O p3 If mark locate fails then define observed position to be expected mark position so that the heightmap is still correct Also set error status code to error if FSTRNLNM ESTATUS NES X00000001 THEN ERROR 2 QMOVE POS 0 0 REL SPO O p3 ENDIF Offset the position for measuring height relative to the alignment mark in order to avoid errors in the height reading due to patterned substrate qmove pos rel 0 300 0 000 qdisplay height If the height measurement fails then do not change height identifier so that last valid height is used If FSTRNLNM ESTATUS EQS X00000001 THEN hgt1 f trninm VB_DHGT_H qset hid h p3 hgt1 ENDIF Pass error code back to layout program exit error In addition the following jobfile jobfile_2 com must be called at each layout cell by defining the following line in the template parameter file vb seq wlvd_dw com directory jobfile_2 com Where jobfile_2 com is as follows OPEN APPEND SHARE POST EMMA MAILBOX EMMA qcal height minimap hid h1 h2 h3 h4 pos 01 02 03 04 qset height map on qexpose pa
166. beam currents 175 17 9 REG With HR final lens coctis etico tie t EI i etd ttr icti HP reae 175 179 2 REG with BHR fmallers uci esiti eit dieto e pete iH eee 176 Is nn n tec wii Part number 878275 Vectorbeam Series Vectorbeam Operator Manual Page vii 19 OVOFVIBV ave cntzes lebe tbe tease cna dis iets ides lena as Vers tans Vests equi pev p Vere teassesvidasvedd musa 177 lee Faleaana E 178 IB aebeSlhoisssicstsud Pere een Modan edd una Med RT UU V aa NT rene re 178 18 4 Interpretation of jobcal fullcal on axis Calibrations ssssssses 178 19 4 1 GornfectoncoelfICIeriis 21d te itae eoe ated th tutae ca hated aas Mnt ete tdeo 178 10 4 2 Resid al coefficient erfore eiiis ituri eet inte antad 179 18 4 3 COMSCUOM enters MN ie i acacuu et deut saute etur see onagni iaeia 179 19 44 Main held Calba TOM ccc tacui iade aar uae a tana aea bac ec Mna eerte ceca 179 19 4 5 Subfield callbfatlOors iece duse thure paeem ta tae c ti tha dese Maret then dentine 179 18 4 6 Beam error feedback calibration sssssssssss 180 18 4 7 Stigmation GallDraliOTi ue iiio eoe ese copre ex reae eda eae inan dad 180 18 5 Interpretation of fullcal deflection field correction calibrations 181 18 5 1 Field focus and stigmiatiol curii cte tare tiscos retine coe ine crates 182 18 5 2 Maii field CISOMIOM seriernes aidais AAS 182 18
167. beam energy is to be 50 kV The pattern generator resolution should be set to 5 nm and the pattern generator grid snapping mode should be set to float to ensure that the gate placement accuracy is maximised By choosing a value of B of 5 equation 19 17 an exposure grid of 20 nm is the result VRU of 4 and the minimum dimension is an integer multiple of this so an exposed gate width of exactly 100 nm can be carried out Actually more process latitude can be obtained using the technique described in Section Negative biasing All the features should have a total bias of 20 nm applied and through an increase in the exposure dose gate widths of 100 nm can be obtained Alternatively if less beam on time is required and the clock frequency is at the maximum 25 MHz a larger beamstep will need to be used The next available VRU value is 8 giving a beamstep size of 40 nm which is not a sub multiple of the 100 nm linewidth One solution is to set the pattern generator grid snapping mode to split and bury in order to maintain the defined 100 nm linewidths Another solution is to apply the negative bias of 20 nm giving a defined width of 80 nm which is a sub multiple of the beamstep size Example 3 The third example pattern is a 4 Gbit memory device The design grid is 1 nm The minimum dimension is 180 nm The beam energy is to be 50 kV If a pattern generator resolution grid of 1 nm is chosen to avoid grid snapping then the maximu
168. ble partly depends on the size of the final aperture selected Lenses for TFE gun The column for a TFE gun has two magnetic lenses and one electrostatic lens The electrostatic lens C1 is made up of the firing unit and further electrodes in the gun The next lens the beam encounters is a magnetic condenser lens C2 A zoom lens function is formed by C1 and C2 The focus point of C2 is constant This enables the beam diameter at the substrate to be adjusted while holding the focus and current density almost constant It also enables the conjugate blanking condition to be held for all currents The final objective lens C3 provides the main focus of the beam on the substrate There are two options for the final lens high resolution HR and ultra high resolution UHR The distance from the principle plane of the HR lens to the substrate is about 42 mm and the distance for the UHR lens is about 35 4 mm The magnetic lenses are temperature controlled by circulating water Beam blanking The beam may be switched on and off by the electrostatic beam blanker system within the column It is driven automatically during pattern writing but may also be operated manually or by job file control Part Number 878275 Vectorbeam Operator Manual 2 2 5 2 2 6 2 3 Page 20 In the on condition the beam is not deflected by the beam blanker In the off condition the beam is deflected away from the optical axis and is stopped on an intermediate ape
169. ble range Check the specified expected positions and the observed positions Check that any specified rotation is within range Possible gross substrate rotation DWXSCAINV X Scale value is out of usable range Check the specified expected positions and the observed positions Check that any specified scale is within range Possible gross substrate error DWYKEYINV Y Keystone value is out of usable range Check the specified expected positions and the observed positions Check that any specified keystone is within range Possible gross substrate error DWYROTINV Y Rotation value is out of usable range Check the specified expected positions and the observed positions Check that any specified rotation is within range Possible gross substrate rotation DWYSCAINV Y Scale value is out of usable range Check the specified expected positions and the observed positions Check that any specified scale is within range Possible gross substrate error EDGEATXTREM Edge too close to extremity of scan Part Number 878275 Vectorbeam Operator Manual Page 276 Increase scan length and check mark size or edge offset EDGEDATANON Edge data does not make sense Check the parameters particularly the bright dark selection EHTDEFERR EHT not correctly defined on the pattern generator EHTDISABLED EHT disabled EHTFAULT EHT hardware fault EHTOFF EHT off EHTON EHT on EHTRAMPDWN EHT ramping down EHTRAMPUP
170. bout half the limb length 4 The cross limb width LMBWID should be set to the actual width For crosses with more than 1 finger in each limb LMBWID should be set to the finger width 5 The cross limb separation LMBSEP is the distance between adjacent edges of adjacent fingers 6 The cross limb lines LMBLNS is the number of fingers in each limb This is 1 for plain crosses 7 The mark tolerance MRKTOL is typically set to 0 1 This can be reduced if mark rejection on the basis of mark size is required Choosing signal parameters 1 The minimum contrast CT is normally set to 0 1 1096 for the datum plate marks and marks with similar contrast The datum plate mark typically has a contrast above 0 2 2096 for normal backoff levels The mark contrast is defined as maximum video level minimum video level 255 where the maximum level will correspond to the mark for a bright mark and the background for a dark mark The mark contrast can be found by locating a mark with the adc qualifier and using the adc plot cpr PVWave program to examine the ADC data see Section Diagnostic output of the mark locate function or Acceptance Test and Operator Jobfiles manual 892777 Figure 11 3 shows an example plot from this program and the contrast is about 0 35 Mark Locate ADC Plot X axis scan 1 coarse scan 1 aso E ee EE ES rrr re Do EH po SET e paw dato Noise RMS 2 66253 ADC units 200 M
171. calls holder_table com directly and therefore the holder number must be specified directly VB_OPER gt H H_0047 The qdisplay air command types up a table showing the occupancy of the airlock positions and on a VB6 with a bar code reader the table contains the numbers of the holders in the various airlock positions However the caspos table positions are not automatically updated when holders are loaded and unloaded but this must be done manually if it desired to keep track of which holder is in which library slot VB5 and VB6 with 10 chuck airlock users may it useful to change the table to keep track of which holder is in which library slot but this is not necessary To view the table type VB_OPER gt caspos Part Number 878275 Vectorbeam Operator Manual Page 156 To enter a holder number into the table at position VB_OPER gt caspos H_ To save the current table to file VB_OPER gt caspos save To load the file into table in memory VB_OPER gt caspos restore To initialise a holder in table position equivalent to H VB OPER caspos Part Number 878275 Vectorbeam Operator Manual 17 17 1 17 1 1 17 1 2 17 1 2 1 Page 157 Job specific machine set ups Gun alignment The gun alignment coils are positioned between lens 1 and lens 2 and are used to adjust both the position the beam from the optic axis and the angle of the beam to the axis Correct gun alignment is when wobbling lens 1
172. cate MRKOFFWIDTH Mark offset width parameter out of range Refer to Emma command manual for valid range MRKPOINTAV Mark point averaging parameter out of range Refer to Emma command manual for valid range MRKPRLL Number of parallel scans parameter out of range Refer to Emma command manual for valid range MRKREDGE Failed to find right edge MRKRT Mark rise time parameter out of range MRKRTT Mark rise time tolerance MRKSFDAC Mark trap field DAC MRKTEDGE Failed to find top edge MRKTFLDP Trap field position out of range MRKTOL Mark tolerance parameter out of range MRKUNKDEF Unknown deflection option Check command syntax and use available options MRKUNKDIR Unknown raster scan direction option Check command syntax and use available options MRKUNKLOC Unknown locate option Check command syntax and use available options MRKUNKPLT Unknown plot option Check command syntax and use available options MRKVRU Mark step resolution parameter out of range MRKWIDPRM Mark width parameter out of range MRKWIDTH Width of feature found does not correspond to param MRKWRITFLD Trap field extends beyond writing field MRKZEXSIZ Zero exel size MSLOPEFAIL Marker Slope Calibration Failed NOCASHOLDER No holder in airlock Check the source position specified Check the occupancy using the qdisplay air read command If it is known that a holder IS present then extreme Par
173. ch of the marks could be reduced to say 30 um to reduce the coarse search time further There should be 3 or 4 such arrays on the wafer Part Number 878275 Vectorbeam Operator Manual 20 3 4 20 3 4 1 20 3 4 2 20 3 4 3 20 4 Page 237 The DCL procedure to locate the target is called locate prealign com The sequence locates any 3 targets on a diagonal This target has the advantages that coarse searching only needs to be done over a small distance short scans are less likely to scan outside the allocated area on mechanically misaligned wafers and the whole target is not destroyed by subsequent processing only 3 targets are destroyed the rest of the targets may be used Die alignment mark layout WT Die Alignment mark Pattern to be exposed Centre of __ alignment marks Centre of pattern die Figure 20 4 Diagram showing die alignment mark layout parameters Die A die is defined for the purposes of this direct write alignment as a rectangular area of the substrate with boundaries which are placed half way between the centre positions of neighbouring patterns The centre of the die corresponds to the centre of the pattern After the die marks have been located and the mapping into die mode has been done the 0 0 point is defined by the offsetx and offsety values The pattern will be centered at the 0 0 point which is not necessarily the centre of the alignment marks OffsetX OffsetY
174. click on it and then click on CONTINUE Wait for qsubstrate transfer SUBX to respond to clicking these buttons and for these buttons to become dimmed Then use control Y C to abort the remainder of the COM file Qsubstrate transfer SUBX responds to control Y C and stops doing whatever it happened to be doing however critical that may be The holder may be left in an unknown position and require the system to be vented to be recovered The Abort button is less drastic and the qsubstrate transfer SUBX code checks the Abort button at uncritical points in the code Control Y C when directed at the COM file allows the current command to complete and then terminates the COM file Holder initialisation sequence When the holder has been loaded onto the stage the holder sequence should be run for that particular holder See Sections Holder parameters Setting up Part Number 878275 Vectorbeam Operator Manual Page 150 the holder parameters and Holder position table caspos first VB_OPER gt H3 The sequence e Defines the positions for the Focus Mark FM the Faraday Cup FC the Gun Target GT the Datum Point DP the Fine Focus Mark FFS the centre of the substrate the substrate size and the substrate height meter table e Initialises the stage and switches to stage mode absolute e Calibrates the height meter on the substrate e Locates the Focus Mark FM and sets the datum offset e Calibrates the height meter on t
175. cm bdue ble dada Eta tds se ncuna tn Ko nae eg 229 19 6 1 Top level layout parameter file 229 19 6 2 Second level layout parameter file ssseeesssss 230 Direct write alignment 2s esc daep ev Ra SE pU M bMS Vot PNIDIRU IE H FI EA MEM I dee T 20 1 20 2 20 3 20 4 20 5 20 6 20 7 20 8 Direct write alignment MEIGS ua sed inccccovcisicaiouaeciiaccasanciteaas aaa 231 20 11 Global alignment eiit tein iain 232 20 1 2 Die bysdie alignimebt ic caccvciae ancients 232 20 1 3 Direct alignment commands on the VB sss 232 A general direct write alignment method for regular rectangular arrays 233 2024 Global aligntl ent sss score io eio aiaa 233 20 2 2 Die by die alignment i neret rene Prts sit te epe d iu rE educa 234 Layout definition for regular rectangular alignment mark arrays 234 20 3 1 Global alignment mark layout sasis sirin sinisa 234 20 3 2 Simple global alignment mark layout ssesssssss 235 20 9 9 Automatic global alignments ioi iuias 236 20 94 Bie aligament mark ByOLL a coire ta sir ah drea so th Feo aas 237 Pacem dala DODGE se cesta p anaa Pes eid nbn Up bx vel cla blog lard dd 237 EVE Propa TO OE DOTT 238 20 5 1 Layout parameter tile Setup niei iit ioter err etr eei 238 Machine preparation for direct write sssssesesessesesenenenennene 241 Substrate prepar
176. command such as ON ERROR THEN EXIT 44 ON ERROR requires the number to be added to EXIT for nested com files This ensures that the error condition is passed up the chain of com files executables etc to the top level The number 44 is used as it combines message number 40 ABORT with severity 4 fatal or severe The number must be put together in binary The bits 0 to 2 signify the severity O2 warning 1 Success 2 error 3 informational 4 fatal or severe The bits 3 to 15 signify the message SET ON is required to get ON to work at all Refer to the appropriate DCL manual for full information 2 After Emma commands any errors can be trapped using IF FSTRNLNM ESTATUS NES X00000001 THEN Such a line is however required after each Emma command Alternatively the symbol Emmaerr can be defined as the name of a jobfile to be run if an Emma error occurs during the execution of a jobfile Emmaerr vb users error recovery jobfile com The text associated with an error number that typed in the DECterm can be found by typing VB_OPER gt write sys output f message NUMBER Part Number 878275 Vectorbeam Operator Manual 27 6 2 Page 272 Subsystem errors 27 6 3 27 6 4 Emma normally interprets warnings and errors received from sub systems and reports them as described above Emma may also receive errors from the sub systems for which it has no interpretation and no F message exists an
177. ctory All standard output and standard error messages will be written to this log file 22 2 2 DECTerm logging using the SET HOST command This is useful for interactive work A new DECTerm is created which is logged If the OPER mode is required in the new DECTerm then OPER status will need to be given up by typing SUPER The new DECTerm with logging can be created with the following command VB SUPER SET HOST 0 LOG lt name of log file gt This will ask you to login as a new user on the same OpenVMS machine All commands typed and their output will be logged into the specified file The VB only has a license for one additional login so this will only work if all other users besides user VB are logged out or the system will report Attempted usage exceeds active license limits 22 2 3 DECTerm logging using the Options menu All output which is typed up in the DECTerm can in addition be written to a logfile by setting the following options in the Options menu of the DECTerm 1 Printing destination to file 2 File to directory name filename 3 Printer to auto print mode To stop the logging switch the Printer in the Options menu to normal print mode Part Number 878275 Vectorbeam Operator Manual 23 Page 249 Creating scanning electron image files The VB has a facility for scanning a rectangular area using the pattern generator and recording the video level at each point in a file
178. current position status and all preset positions can be displayed using VB_OPER gt aperture display all Controlling the automatic aperture adjustment mechanism 1 To initialise set up home and return to last good position type VB OPER aperture init 2 To reset initialisation sequences to factory defaults loses last good position type VB OPER aperture reset 3 To home axes but not return to stored position type VB OPER aperture home 4 To send command direct to controller and view responses type VB OPER aperture com lt command gt Part Number 878275 Vectorbeam Operator Manual 17 4 17 4 1 17 4 1 1 17 4 1 1 1 Page 165 Focusing the beam The focus of the beam on the substrate depends on all the settings of all 3 lenses the fine focus and the height of the substrate However there is only correct focus setup consisting of the following 1 lens 1 is set to give the required beam current 2 lens 2 is set to give conjugate blanking 3 lens 3 is set to focus the beam on a substrate for which height meter reads 0 um 4 the fine focus is set to take any substrate height deviations from 0 um into account All the procedures to focus the beam described in this section deal with obtaining the correct lens 2 C2 value and corresponding fine focus values Focusing is carried out by varying the current in lens 2 C2 and this must give the appropriate fine focus values for the substrate height
179. d and uncertainty in the positions of further marks is lower Alignment of first global mark If the reference feature option is not used then the uncertainty in the position of the substrate after the substrate has been loaded will be present when searching for the first mark If the uncertainty in the position of the first mark is more than the maximum coarse search range then manual global alignment must be carried out in SEM mode If the reference feature option is used then the first global mark will still be located Alignment of further global marks After the reference feature and first global alignment mark or just first global alignment mark have been found the uncertainties in the positions of further marks for global alignment will be much smaller typically 100 um and depend Part Number 878275 Vectorbeam Operator Manual Page 234 mostly on the rotation of the substrate 1 2 or 3 further marks on the substrate as far apart as possible are aligned making a total of 2 3 or 4 global marks If a total of 2 marks are used the direct write mapping will use an X and Y offset and the same scale and rotation in both X and Y If a total of 3 marks are used the direct write mapping will use an X and Y offset and separate X and Y scales and rotations If a total of 4 marks are used rather than 3 the direct write mapping can correct for keystone errors and will be more accurate The mapping coefficients generated are used for eithe
180. d typically give numbers as in the table below Correction Scale Rotation Keystone coefficients um mm um mm um mm m9 vem wn Residual Scale Scale Rotation Rotation Keystone Keystone coefficient errors um mm um mm um mm um mm um mm um mm mean sigma mean sigma mean sigma X or Y 0 01 0 01 0 01 0 01 0 01 0 01 fieldsize fieldsize fieldsize fieldsize fieldsize fieldsize Correction errors Scale Scale Rotation Rotation Keystone Keystone nm nm nm nm nm nm 10 10 10 10 10 10 18 4 5 Subfield calibration An accurate calibration should typically give numbers as in the table below Part Number 878275 Vectorbeam Operator Manual Page 180 Correction Scale Rotation coefficients um mm um mm Ce 9 9 Residual Scale Scale Rotation Rotation coefficient errors um mm um mm um mm um mm X or Y 0 01 0 01 0 01 0 01 sub fieldsize sub fieldsize sub fieldsize sub fieldsize Correction errors Scale nm Scale nm Rotation nm Rotation nm mean sigma mean sigma 10 10 10 10 18 4 6 Beam error feedback calibration An accurate calibration should typically give numbers as in the table below Correction Scale Rotation coefficients um mm um mm 9 ee Residual Scale Scale Rotation Rotation coefficient errors um mm um mm um mm um mm X or Y 0 01 0 01 0 01 0 01 BEF fields
181. d will then report sub system errors directly The text of such errors and warnings will be reported followed by EMMA Emma information messages SUCCESS Command Successful CALNOUPDATE New data discarded because NOUPDATE set DCNEWCURFRQ Current frequency not retained was outside new band EOINFO Successful Message from EO Control in console window EOINIT X Initialising EO communication link EOLNKACT EO communication link active EOLNKNRDY EO communication link is not ready LMAPDEASS Lens Map s deassigned from changed DW transform s LNKALRDYCON Hostis already connected to the sub system LNKESTB Host to sub system communications link established NEWMRK New mark definition NUPPOSDISP Couldn t update stage position display correctly PGINFO Successful Message from PG in console window PGINIT Initialising PG communication link PGLNKACT PG communication link active PGLNKNRDY PG communication link is not ready QUALIGNORE Illegal qualifier s ignored STINIT Initialising Stage communication link STLNKACT Stage communication link active STLNKNRDY Stage communication link is not ready SVINFO Successful Message from Stage in console window Emma error messages All Emma errors are preceeded by EMMA W Normal operating errors are given in the following list ADCFILEERROR File error in file for ADC plot ADCFILEOPENING Could no
182. de by typing VB OPER VID P This command switches the machine into SEM mode and selects the automatic peak level adjustment using the command VB OPER qset video vmode peak mark 0 8 This command is the same as selecting the peak switch in the Set Video Level panel Figure 7 9 Auto peak level will adjust the gain until the peak video level reaches the value set either in the Set Video Level panel or by the mark command option usually about 7096 8096 The final level will be affected by the amount of noise in the signal Alternatively the video gain can be automatically adjusted by switching into SEM mode and selecting the video mean switch in the Set Video Level panel Figure 7 9 or by using the command VB OPER qset video vmode mean bckgrd 0 2 Auto mean level will adjust the gain until the median video level reaches the value set either in the Set Video Level panel or by the bckgrd command option This is usually set to about 20 30 for bright marks when the SEM magnification is such that the background dominates the area on the monitor The final level is not affected so much by noise in the signal but does depend strongly on the Part Number 878275 Vectorbeam Operator Manual 17 2 2 2 17 2 3 Page 162 SEM magnification When running automatic mark searches that also do video adjustments the SEM magnification will need to be set low enough to ensure that marks are still in the field of view even if they are not ex
183. define the order of the shapes is a further option for the Emma command defining the grating VB_OPER gt aset grating shape_meander number_of_lines_per_bundle where n is an integer between 1 and 16384 and specifies the number of sub shapes exposed before switching to the next subfield 16384 2 which is the maximum number of bits in a subfield If this option is set and n 1 then the description at point 1 above applies If the option is not specified then n defaults to 1 If this option is set and n gt 1 then the requirement at point 1 under heading 6 also applies but with the word line replaced by group of lines If this option is set and n gt 1 the order of the sub shapes also complies with point 2 under heading 6 If n is greater than the actual number in a subfield then no message is required and the subfield will simply not be revisited Shape erosion The shapes in patterns are defined as outlines However exposure must be done as a series of exposure points or exels The conversion from outline format to exel format is done by the PG software The upper and right edges are moved VRU exels towards the lower left edges and this is known as shape erosion The remaining points in the eroded outline are exposed Figure 24 3 shows this process for VRU 1 where the grid spacing equals the pattern generator resolution Figure 24 3 Diagram showing a square shape outline and the exposed points at VRU 1 This has
184. details on blocksize The largest blocksize that can be set in the converter for a FRE file is given by largest blocksize in uum Max Fieldsize 32 x PG resolution Equation 19 4 Expressed in terms of exels largest blocksize in exels 25 32 65504 Equation 19 5 The reason not all 2 deflection bits can be used is because the FRE file format uses unsigned short integers to represent the coordinates of the shapes in outline format Such representation is compact but can only define 2 different numbers As the shapes are defined in outline format this limits the number of Part Number 878275 Vectorbeam Operator Manual 19 1 5 2 19 1 5 2 1 19 1 5 2 2 Page 203 exels which can be defined to 2 1 As the user can define a VRU of up to 32 at exposure time the maximum blocksize is limited to 2 32 to avoid possible stitching problems if the blocksize were not an integer number of beamsteps The stitching problems arise if shapes at the edge of the block are fractured by the block boundary to leave shapes smaller than the beamstep VRU x PG resolution as these will be discarded 18 bit pattern generator additional details on blocksize VEP file The largest blocksize that can be set in the converter for a VEP file is the Maximum Fieldsize largest blocksize in um Max Fieldsize Equation 19 6 Expressed in terms of exels largest blocksize in exels 2 262144 Equation 19 7 FRE file
185. diameter can be adjusted manually without using the demagnification table as follows 1 Create a large spot 60 to 100nm to ensure that the spot is definitely larger than the mark slope 2 Carry out jobcal 3 Note the average spot size measured on the focus mark VB_OPER gt QDISPLAY DIA FM SQUARE LINELENGTH 2 EDGE 5 FILT 16 POINT 8 LINESCANS 1 4 Note the current at this large spotsize 172 nA on a 400 um aperture 5 If the required spot size is for example 40 nm divide the large spot size by the smaller one and square the ratio 410 40 10 25 square the ratio 105 06 6 Now divide the large current with the squared spotsize ratio and this will give you approx the current for a 40 nm spot 172 105 06 1 637 nA for a 40 nm spot This formula can be used for all spot size to current values Part Number 878275 Vectorbeam Operator Manual Page 175 17 9 Theoretical tables of on axis beam diameters and beam currents The following tables show the on axis performance of the tool The HR and UHR performance on axis is similar The UHR deflection errors are considerably smaller than for the HR 17 9 1 FEG with HR final lens The source diameter was assumed to be 30 nm the emission intensity was assumed to be 0 5 mA sr 17 9 1 1 20 kV Beam current Beam Beam Beam Beam diameter for diameter for diameter for diameter for 100 um 200 um 300 um 400 um aperture nm aperture nm aperture nm a
186. dth parameter out of range MRKMSRHGT Mark measurement height parameter out of range MRKOFFWIDTH Mark offset width parameter out of range Vectorbeam Operator Manual 08019368 08019370 08019378 08019380 08019388 08019390 08019398 080193A0 080193A8 080193B0 080193B8 080193C0 080193C8 080193D0 080193D8 080193E0 080193E8 080193F0 080193F8 08019400 08019408 08019410 08019418 08019420 08019428 08019430 08019438 08019440 08019448 08019450 08019458 08019460 08019468 08019470 08019478 08019480 08019488 08019490 08019498 080194A0 080194A8 080194B0 080194B8 080194C0 080194C8 080194D0 080194D8 080194E0 080194E8 080194F0 080194F8 08019500 08019508 08019510 08019518 08019520 08019528 08019530 08019538 08019540 08019548 08019550 08019558 08019560 Part Number 878275 Page 286 MRKOFFHGT Mark offset height parameter out of range MRKLMBWID Mark cross limb width parameter out of range MRKTOL Mark tolerance parameter out of range MRKRT Mark rise time parameter out of range MRKCT Mark contrast parameter out of range MRKLOOK Mark look option unknown MRKFILT Mark filter parameter out of range MRKLINEAV Mark line averaging parameter out of range MRKPOINTAV Mark point averaging parameter out of range MRKVRU Mark step resolution parameter out of range MRKPRLL Number of parallel scans parameter out of range MRKCSOFF Mark cross coarse search offset parameter o
187. e fine beam partial database 3 The test beam partial database 12 2 3 2 Calib The coarse beam partial database The calib partial database contains the following 12 2 3 2 1 Parameters None 12 2 3 2 2 Partial databases 1 The size cal partial database 2 The dist cal partial database 3 The stig_cal partial database 4 The mag cal partial database 5 The pos cal partial database 12 2 4 Bottom level partial databases The parameters in each bottom level database are listed below 12 2 4 1 General The general partial database contains Part Number 878275 ONOORWN gt mode SEM FAB blank status subfield settling time acute angle settling time shape to shape settling time 32 actual clock frequencies actual min dose controller frequency actual max dose controller frequency Vectorbeam Operator Manual 10 Page 108 min dose controller setting actual frequency nominal frequency set value for dose set frequency for dose set area dose set line dose set relative dose max dose controller setting actual frequency nominal frequency set value for dose set frequency for dose set area dose set line dose set relative dose 32 clock dose types actual frequency nominal frequency set value for dose set frequency for dose set area dose set line dose set relative dose resist sensitivity nominal current actual current inversion strategy
188. e Couldn t create expose queue Couldn t create Par IP gt Cl queue Couldn t create Cl gt Par OP queue Couldn t spawn vacuum task Couldn t spawn stage task Couldn t spawn expose task Couldn t create PGD Mon semaphore Couldn t spawn PGD Mon task VxWorks Version Number erroneous Couldn t spawn Par IO task Malloc failed in pattern reception Error creating semaphore Error getting semaphore Command Interpreter Errors qEtherlP receive error Wrong no of bytes in long message Stage move table already exists Send to message queue failed Got reply to wrong command Got Abort out of context Invalid command in current state Query of wrong command Parallel I O checksum failure Parallel Input send to Cl failed Parallel Output receive from Cl failed Parallel Input timeout Parallel Output timeout TEST BEAMS Parallel Output timeout TEST BEAMS Parallel Input timeout TEST BEAMS Parallel I P char timeout qErrors receive error qCl receive error Error receiving from qExpose queue Invalid state type Invalid cmnd in current state Illegal number of fields for exposure Vacuum not OK for exposure Stage move timeout Abort arrived too late Invalid semaphore ID PG not removed stage move signal 1 Error receiving from qStage The LOAD position is not defined X destination out of range Y destination out of range Stage move timeout Stage home timeout Stage needs initialising via move home X Interf Measurement
189. e During bake out the hot inside surfaces of the emission chamber release gas molecules much more quickly than at room temperature and they are removed by pumping After cooling outgassing from the emission chamber falls to much lower level than before bake out Three types of pumps are provided to evacuate from atmospheric pressure to ultra high vacuum these are mechanical pre vacuum rotary turbo molecular and ion getter pumping Pre vacuum pumps are efficient between atmospheric pressure and approximately 10 Pascal 0 75 x 10 Torr and evacuation is always initiated with these pumps On the VB5 Series three rotary pumps are provided for the Part Number 878275 Vectorbeam Operator Manual 3 3 3 4 Page 24 airlock the chamber and the emission chamber respectively On VB6 Series without Brooks handling a single rotary pumps is provided for the System and Airlock On VB6 Series with Brooks handling two rotary pumps are provided for the VB6 and Brooks handling The turbo molecular pumps are used for pumping between pressures of about 10 Pascal 0 75 x 10 Torr and about 10 Pascal 0 75 x 10 Torr A substrate chamber operating vacuum of better than 4 x 10 Pascal 3 x 10 Torr is required This is achieved in two stages by first removing most of the gas molecules with the pre vacuum pumps until a pressure of 13 Pascal 9 75 x 10 Torr is reached and then continuing to remove the remaining gas molecules with the turbo
190. e opening the door using the up and down buttons without pressing any interlock buttons In order to move the airlock to the required holder position with the door open hold in the right hand green interlock button and then press the up arrow or down arrow button until the correct holder position is shown on the dial Pressing the up and down buttons for a short time will nudge the airlock library to the next position as long as the right hand interlock button is kept depressed Press button marked door to open the door To unload a holder slide the empty trolley forward until it reaches the mechanical stop Push the lever on the trolley to the right to raise the holder off its feet inside the airlock and with the lever in this position slide the trolley back Release the lever The holder can now be exchanged from the trolley To load a holder select an empty position in the airlock open the door push the lever on the trolley to the right to raise the holder and slide the trolley with the holder forward until it reaches the mechanical stop Release the lever and slide the trolley back Close the airlock door by releasing the door button and pressing both green interlock buttons one on the panel next to the door button and one on the front of the airlock Part Number 878275 Vectorbeam Operator Manual 16 5 Page 148 Holder loading unloading on from stage When the pressure in the airlock is low enough holders can be transferred
191. e 20 3 Figure 20 2 shows an array of marks 20 3 1 2 Global OffsetX Global OffsetY The Global OffsetX and the Global OffsetY parameters are the X and Y distances respectively between the centre of a global cell and the lower left corner of the lower left mark 20 3 1 3 Global CDCX Global CDCY The Global CDCX and Global CDCY parameters are the distances between the global marks in the X and Y directions respectively within a cell 20 3 2 Simple global alignment mark layout 20 3 2 1 Example 1 An example of a commonly used simple global alignment layout is shown in Figure 20 3 Part Number 878275 Vectorbeam Operator Manual 20 3 2 2 20 3 3 Page 236 Substrate Global alignment mark Layout of dies Centre of layout of dies One global cell Global_CDCX Referenceoffset Figure 20 3 Example global alignment mark layout In this case the number of global cells in x and y is 1 The pitch of the global alignment cells is not defined physically but a value should be entered Example 2 It is possible to use the die alignment marks for global alignment and not have any global alignment marks All the parameters for the global alignment are the same as for the die alignment Automatic global alignment Automatic global alignment requires a scheme to deal with the range of shifts and rotations of the substrates in the holders which result in the actual mark positions being up to several hundred mic
192. e PUER YE HR ego Geta REI ed 269 27 39 WVRAX Alpha screen lOCKs UD auos senda ceris rrrnsda cec duc cerra ste ERS 269 Part number 878275 Vectorbeam Series Vectorbeam Operator Manual 27 4 Height meter error IecOVOly isa desit boe Mi vo de dive Ug erases ved asd eria tagen 269 2 5 Gubfield calibration GITOE a det oda ioa cR DU ced ut Un LAE URL RA IM ETE NER LEAN 270 27 6 Errors and warnings message list with suggested recoveries 271 27 6 1 Jobfile automatic recovery sesssseeseeeeeeeeeennnn nnns 271 27 02 JSOUbSyslelmGlTOIS acci peer ba deese asse RO xeu SS 272 27 6 3 EmmainformationmesSsagES issic iseis 272 27 6 4 Emma error messages aa 272 27 655 iEmma etror n mbOets edes hate a 283 27 6 6 Pattern generator error MESSAGES sssessseeeeeee 289 27 6 7 lage eror MESSAGES uec ea seeks sess tu xis red base Ek keen el bas sun 294 27 6 8 EO error messages inpet cobarde iaai 298 Part number 878275 Vectorbeam Series Vectorbeam Operator Manual Page 1 Preface and safety information The purpose of this manual is to provide comprehensive information to enable an operator to use Vectorbeam Series lithography systems in the most efficient manner The manual VB6 with autoloader operator manual part number 893148 covers the Brooks handling option This document is part of a full set of documentation available for users and supplied during installation It sho
193. e UHR final lens around the beam and detect back scattered electron detectors Their outputs can be mixed in any combination although normally they are summed and this is done with the following command VB OPER qset detector BSB1B2B3BA BIAS Where the BIAS qualifier is generally needed above 10nA beam current Vectorbeam Operator Manual Page 128 14 4 Transmission detector A P N junction device is mounted on the superplate of a VB6 at a position of approximately X 132 Y 5 The device detects electrons transmitted through the substrate The detector is a silicon diode type MSXO03 with an active area of 10 x 10 mm The anode of the diode is brought out on connector X6 and the cathode on connector X7 on the chamber lid The diode is connected to a preamplifier mounted on the chamber and the preamplifier output is connected to an input of the video processor card To select the transmission detector signal type VB OPER 2QSET DETECTOR TR This restores the last used preamplifier channel and bias state see below When fully depleted by the application of a reverse bias the time constant for the diode junction is about 4 ns If no bias voltage is applied the time constant could be 1000 times greater In addition with incident currents above 1 nA it is essential to apply reverse bias otherwise the charge density in the detector seriously degrades the transient response It is recommended that the diode is always used with bias
194. e alignment time X stage travel time between marks Y stage travel time between marks Stage settling time Mark locate time Calculate mapping time Load mapping time Displaying and setting the settling times The various settling times can be displayed and set as described in the following sections Mainfield deflector settling times 16 bit pattern generator The current values in us may be displayed by logging onto the DCP and entering gt G_MfSettlingShort gt G_MfSettlingMedium gt G_MfSettlingLong gt G_UseMfSettlingShort must 21 These values can be changed directly by entering gt G_MfSettlingShort value in us gt G_MfSettlingMedium value in us Part Number 878275 Vectorbeam Operator Manual 24 2 1 2 24 2 2 24 2 2 1 Page 254 gt G_MfSettlingLong value in us gt G_UseMfSettlingShort 1 The mainfield settling values will be reset to any values specified in the EMMA CTRL DCP_CONFIG VW file for 16 bit pattern generators or the EMMA CTRL WFDCP_CONFIG VW for 18 bit and 20 bit pattern generators file whenever the DCP is rebooted These values are set by the same statements as above in the relevent file G_MfSettlingShort value in us G_MfSettlingMedium value in us G MfSettlingLong value in us G_UseMfSettlingShort 1 18 bit and 20 bit pattern generators The current values in us may be displayed by logging onto the DCP and entering gt G BlankTimeMin g
195. e background 11 2 2 Mark size 1 The mark locate function will locate rectangles with sides of length 1 to 120 um The mark locate function will locate crosses with any limb length Part Number 878275 Vectorbeam Operator Manual Page 89 as long as all four limbs are within the deflection field by a distance given by CSOFF see later Large crosses with limbs of length 1 mm or square marks with sides of length gt 20 um are used typically for global alignment as their positions will initially be unknown and a larger target is easier to find Square marks with sides of length 4 to 20 um are used typically for die by die alignment The mark sides should be long enough so that several scans can be carried out at different positions along the edge thus reducing the effects of any edge roughness to an acceptable level The mark locate function relies on detecting the edges of a mark and so if point 2 above is met increasing the mark size does not increase the accuracy Alarge mark size may be useful for faster location if the uncertainty in the mark position is large A large area can be scanned more quickly if during the coarse search if the mark is larger Large marks are more likely to be used for global alignment particularly for the first mark as the uncertainty in the first mark s position includes the X and Y substrate insertion error 11 2 3 Mark positioning See also the chapter Direct write alignment for furthe
196. e control electronics rack 1 CER 1 shown in Figure 1 2 contains the plinth interlock control system PICS This diagram is illustrative only it is only accurate for machines with an HR final lens and without Brooks handling MAINS PANEL c l Ss i o ES eo oo em cmm e S OC OO GENMARK CONTROLLER AUX PCB e e o EEEE EE e o o PHOTOMULTIPLIER PSU _ AUX PSU TRAY C cie o o O Cy 03 T hh i o OO o g D o a Tu O0 Pics L i 29 fe c i Lt d SS ae uy W uw wy wy ae weer Ty C ION ION PUMP Zz me PUMP PSR CRATE Fibre optic I F Pcb e f mgl e MAINS DIST N TC TC C aie C Lt o 7 LJ O e ele o o T En A Figure 1 2 Control electronics rack 1 1 3 2 Control electronics rack 2 Part Number 878275 The control electronics rack 2 CER 2 shown in Figure 1 3 contains the pattern generator analogue PGA crate the electron optics EO crate the column control unit CCU and the column control unit power supplies Vectorbeam Operator Manual 1 4 1 4 1 1 4 2 Page 9 Oo re
197. e correct table is selected each time that the beam is moved from the substrate onto the focus mark and vice versa If the height sensor has been calibrated at the centre of a plate and HEIGHT SENSOR AT LIMIT errors are reported when trying to display heights towards the edge of the plate it may be better to calibrate towards the edge of the plate When exposing a pattern if the height meter reports an error when measuring then the last height reading is used Therefore if writing a pattern which starts in a region where the height sensor will not work it is important that the last reading obtained by doing a successful QDISP HEIGHT should be taken as close as possible to the start of the pattern Height meter timing The total time taken for the pattern generator to obtain a reading from height meter on the most substrates is fairly constant at around 80 ms and is dominated by communication times However this time can increase to a few seconds for two reasons 1 The change in the signal gain required from one reading to the next is large The gain adjustment time depends on the table selected but may be up to several seconds Part Number 878275 Vectorbeam Operator Manual 10 5 10 6 10 6 1 Page 76 2 The substrate has a low reflectivity requiring larger integration times In this case the maximum integration time can approach 1 second An error condition such as reading off the substrate surface will cause
198. e fine focus to 0 0 and check image in SEM mode Adjust C2 to give best focus and try again Check the autostig autofocus parameters 0x20830008 DCP Unable to find best stigs Check video levels and image in SEM mode Check autostig parameters or detector selection or column adjustment 0x40830009 DCP Axial stig DAC value out of range Adjust the focus and stigmation manually in SEM mode Check the autostig autofocus parameters and try again 0x4083000A DCP Diagnonal stig DAC value out of range Adjust the focus and stigmation manually in SEM mode Check the autostig autofocus parameters and try again 0x4083001 1 DCP Focus stig zero length scan or dvd by zero Check autostig parameters particularly the scanlength and edgeoffset 0x20830012 DCP Looks like no video Check video levels and image in SEM mode Check autostig parameters or detector selection or column adjustment 0x20830013 DCP No sensible edge data Check video levels and image in SEM mode Check autostig parameters or detector selection or column adjustment 0x20830014 DCP Edge too close to scan limit Check autostig parameters particularly the scanlength and edgeoffset 0x20830015 DCP Bad subfield address Reboot PG 0x20830016 DCP Main field scale out of range Reduce any entered value The main field sensitivity may need adjusting 0x20830017 DCP Main field DAC out of range Reduce any entered value The main field sensitivity may need adjusting 0x20830018 DCP Subfield scale
199. e has been set up correctly Check expected and observed positions Try restarting Emma Contact Vistec if it persists DWDSTMAPINV Destination map mode invalid in ConvertMappedPosns The number that identifies the DW map mode is not in the range O to 9 Internal consistency check has failed Try reloading the current DW mode if not ABS Try setting up the faulty mode again if you know which it is Try restarting Emma Contact Vistec if it persists DWINVPARAMS Invalid number of exp obs points Use 1 3 or 4 points for both expected and observed positions DWMAPORDINV Mapping order invalid if given must be 1 3 or 4 The map order isn t 1 3 or 4 Internal consistency check has failed Try setting up the faulty mode again Try restarting Emma Contact Vistec if it persists DWNACBACK Inaccurate back transform DWSRCMAPINV Source map mode invalid in ConvertMappedPosns The number that identifies the DW map mode is not in the range 0 to 9 Internal consistency check has failed Try reloading the current DW mode if not ABS Try setting up the faulty mode again if you know which it is Try restarting Emma Contact Vistec if it persists DWVECALLOC DW transform matrix allocation failure DWXKEYINV X Keystone value is out of usable range Check the specified expected positions and the observed positions Check that any specified keystone is within range Possible gross substrate error DWXROTINV X Rotation value is out of usa
200. e input data which are not an integer number of beamsteps in size so that the original defined outline is maintained as well This additional data manipulation is one way to ensure the best fidelity in the exposed pattern as when the pattern generator is operating in float mode it rounds down shape dimensions to be an integer number of beamsteps in size and this may change the original defined outline The other way to obtain similar fidelity is to select the split and bury pattern generator mode which causes the pattern generator to apply similar algorithms in real time when the VRU gt 1 The ability to use Cats to process the data for beamstep size means that the results can be viewed and that the amount of work and therefore overhead that the pattern generator has at exposure time is reduced The ability to use the pattern generator to process the data for beamstep size means that the VRU can be varied at exposure time without re converting the pattern VB resolution The VB resolution is only used for the VB20 VB50 and VB100 formats and is calculated by Cats as follows VB Resolution Resolution Rounded up to nearest integer Resolution 2 P eene VB Field This automatically results in the maximum VB Resolution and minimum VB VRU combination because it enables the largest blocksize to be used which results in the shortest exposure time For most patterns there should be no need to define the VB Resolution direct
201. e proximity effect are 1 Assigning doses to the pattern shapes based on their geometry and a knowledge of the inter proximity effect This can be done with the CFA facility in Cats or the Assign command in the Exposure dialogue under the Options menu in Caprox 2 Negative biasing with higher doses to obtain the correct dimensions with higher process latitude 3 Sleeving shapes for exposure with a small spot size Transferring dose distribution to the VB When doses have been assigned to shapes in a pattern either using a proximity correction calculation or by direct assignment the dose distribution can be transferred to the VB in one of the following ways 1 In a clock file 2 In the pattern file itself Using a clock file Tag each shape in the pattern file with a clock number using the facilities in the converter and then use a clock file on the VB to define what dose each clock number stands for The term class used in Caprox is equivalent to clock Note that this method only allows up to 32 clocks due to a 32 clock limit in Emma This method allows both relative doses and absolute doses to be assigned to different clocks Part Number 878275 Vectorbeam Operator Manual 13 3 3 2 13 3 4 13 3 4 1 13 3 4 2 Page 123 Using a pattern file Tag each shape in the pattern file with a relative dose using the facilities in the converter The converter will output a file which has each shape tagged wit
202. e rapid changes to gain during routine operation an automatic calibration procedure must be run initially that measures and saves the backoff settings at each of the 4 coarse gain ranges The automatic video calibration procedure should be done whenever the beam current is changed as follows With the beam current at the required value type VB OPER VID CAL This moves the stage to the Focus Mark and runs through a sequence of video measurements The procedure takes a few minutes and reports when completed Once this has been done a QSET VIDEO command causes the automatic backoff tracking software to find an appropriate BACKOFF setting for the coarse gain range that is in use Part Number 878275 Vectorbeam Operator Manual 17 3 17 3 1 17 3 1 1 17 3 1 2 17 3 2 17 3 2 1 17 3 2 2 Page 163 Final aperture A range of final aperture sizes is fitted to the aperture blade On HR machines these are typically 200 300 400 and 750 um in size Due to the position of the aperture mechanism in the column the aperture sizes fitted to UHR machines are 4 7 times smaller than those fitted to HR machines for the same beam divergence Final aperture selection See Section Theoretical tables of beam diameters and beam currents to choose the best aperture size for the job Manual aperture adjustment mechanism The final aperture size is selected by moving the aperture blade using the aperture positioning knobs The apertu
203. e source drain electrodes which may be defined photo lithographically and then e beam lithography is used to define the gate layer in a mix and match mode More typically on silicon wafers a topographical mark structure is formed by an etching process to yield either an etched trench or pit or a raised mesa by etching a local background area TOPOGRAPHICAL SHALLOW ETCH MATERIAL WPAD lt PIT OR MESA RESIST SILICON VIDEO SIGNAL Figure 11 2 Mark types and their detector signals Figure 11 2 illustrates two types of detector signal which must be analysed to locate marks of two different substrates The majority of marks give a signal as shown on the right The signal shown on the left is best analysed by examining the rate of change of detector level along the scan while the one on the right is best done by examining the detector levels The VB mark locate algorithms allow the user to choose between analysis of detector levels see Section pit type mark or analysis of the rate of change of detector levels see Section edge type mark 11 2 Designing alignment marks for direct write registration During the design of an alignment mark to enable the alignment of subsequent layers the following should be taken into account 11 2 1 Mark type The mark locate function may be applied to rectangular octagonal and circular marks and crosses with limbs of 1 or more fingers All these marks may be brighter or darker than th
204. e stage microprocessor and entering gt useIntelligentSettling g 1 The fixed settling time mode as implemented for the 16 bit pattern generator can be selected by entering gt useIntelligentSettling g 0 The maximum peak to peak variation in microns in BEF signal allowed before the stage is considered settled is set by entering eg gt double SettleWindowXum_g double 0 20 gt double SettleWindow Yum_g double 0 20 Part Number 878275 Vectorbeam Operator Manual 24 3 24 4 Page 256 The sampling rate in ms is set by entering eg gt double SettleInterval_g double 2 5 The number of consecutive samples which must be within the settling requirement is set by entering eg gt SettlePeaks_g 2 The timeout in ms for the settling is set by entering eg gt SettleTimeout_g 2000 All these various parameter values will be reset when the stage is rebooted to program defaults unless they are specified in the EMMA CTRL SV_CONFIG VW file Therefore include the same statements as above in the SV CONFIG VW file For example double SettleWindowXum g double 0 20 Under or overlapping mainfields sparse tiling The stepping pitch between the blocks of a pattern is normally automatically set to the block size in a pattern This can however be overridden to give under or overlapped mainfields both for FRE and VEP format files This option can be used with normally tiled VEP files but cannot b
205. e that the mainfield distortions may need calibrating after adjusting the gun alignment Part Number 878275 Vectorbeam Operator Manual Page 159 17 1 2 2 Fine gun alignment with UHR lens Since the final aperture is between lenses 2 and 3 on UHR systems instead of between the pole pieces of lens 3 the beam moves relative to the aperture by relatively large distances when the gun alignment is changed and may be cut off altogether Therefore the alignment procedure is done initially without any final aperture Another consequence of having the final aperture between lenses 2 and 3 is that the optimum final aperture position can depend on the gun alignment settings The goal of the alignment is to enable the beam current to be changed using appropriate lens 1 and lens 2 settings as usual without requiring the aperture to be realigned The following procedure should be followed 1 2 10 11 12 Part Number 878275 Make sure the FC position is accurate Move to FM Remove the final aperture from the beam On version 1 of the aperture control mechanism this is done by winding in the Right Hand dial fully clockwise This should move to a position where the aperture plate is completely out of the way On version 2 of the aperture control mechanism move to the largest available aperture or to an empty hole if one is available Slowly increase the Beam current to maximum Do this by adjusting lens 1 and lens 2 while all the whi
206. e the rest of the system is vented after a ten minute delay to allow the gun to cool in case the gun is currently switched on If the key switch is in the on vent enable position the entire system is vented after a one hour delay to allow the gun to cool in case the gun is currently switched on To eliminate unecessary delays later versions of PICS VB6 X9 0 VB5 X6 0 and later only wait 10 minutes before venting the gun and the operator must wait 60 minutes after switching off the gun before pressing the vent button 3 7 8 3 7 9 3 7 10 3 7 11 3 7 12 3 7 13 3 7 14 If the gun has been isolated and the rest of the system vented but in spite of this the pressure in the gun has reached atmospheric pressure then setting the ion pump key switch to on vent enable and pressing vent system opens the gun isolation valve Pump system button The function of the pump system button is described in the Section Vacuum system start up Airlock vent button The airlock vent button has been disconnected The airlock full vent must be used The use of the airlock full vent button is described in the Section holder loading unloading in from airlock Airlock full vent button This button requires two presses before the command is accepted The airlock full vent button switches off the pump before venting it and the airlock Airlock pump button The use of the airlock vent button is described i
207. e used to remove the effects of this drift at regular intervals Part Number 878275 Vectorbeam Operator Manual 19 2 1 19 2 2 19 2 3 Page 216 The ontime function may be used to execute a command after a specified time interval This command can be a single machine command or one to execute a command file When the time interval has elapsed the ontime function interrupts any current machine activities in particular pattern exposure and executes the command before returning to the original activity A command is executed only once by the ontime function and therefore to obtain regular ontime events a jobfile must be called which issues another qset ontime command before it finishes Usually the most important parameters that require resetting are the beam position and the beam current However parameters such as the fieldsize calibration and focus may also need to be reset Ontime function See Emma command set Resetting the datum co ordinate system origin The coordinate system datum is normally fixed to the position of a calibration mark The calibration mark position is defined by the operator usually in the holder sequence and this position must be within the mark locate range This position is then not changed as it is the reference The actual mark position is found by mark location and the adjust datum command is used to shift the origin of the coordinate system so that the mark locate position equals the defin
208. e used with VEP files making use of the random field placement The VEP file format allows the stage stepping pitch and the block size to be different and so under or overlapping mainfields can be defined in the pattern file itself without the need to override the stepping pitch The exposure of pattern in this fashion is useful for minimising the number of stage moves during pattern exposure but is only useful for specialised applications where the time saving is extremely important The command for this function consists of an optional qualifier added to the qSet Pattern command namely VB OPER QqSet Pattern filename Blockstep X lt x_pitch gt Y2 y pitch x pitch and y pitch are the required stepping pitches for the Physical Block in millimetres These stepping pitches are not saved in the database and remain valid only until the next qSet Pattern command Blockstep can t be used with qSet Pattern Grating though it s not needed there anyway Random field placement sparse tiling The VEP format has an option which allows the user to select the position of each e beam block rather than use the standard tiling covering the entire pattern area from the lower left corner This is supported by Cats version 10 24 00 onwards and the Cats documentation should be consulted for details One of the uses of random field placement is enabling sparse tiling to be carried out Figure 24 1 shows a simple example of sparse tiling
209. e user wants to locate Standard definitions for the most common mark types present on the datum plate or used for the acceptance tests are however stored in the file Mark_data com These can be used or modified as required Other mark descriptions can be added to the file Each definition is give a mark name to make operation simpler Once defined these are the parameters that are passed to the PG each time a mark locate is invoked by the mark locate command Syntax The syntax of the mark definition function is defined in the Vectorbeam Command Set Manual 878274 Choosing geometry parameters 1 The height and the width H W should be set to the actual height and width of the mark For crosses these parameters are the distances between the ends of the vertical and horizontal limbs For crosses larger than the maximum height and width which can be set simply set the maximum values 2 The measurement height width MH MW is typically set to 1 um The minimum value should normally be MH MW PARA beam diameter so that the adjacent scans do not overlap giving better averaging of any mark edge roughness The maximum value should be less than the maximum edge length available Note that increasing MH MW beyond the edge length 2 MLEN may give spurious measurements on an octagonal mark Part Number 878275 Vectorbeam Operator Manual 11 4 3 Page 92 3 The cross offset height and width OH OW should be set to a
210. e vented independently with dry nitrogen as required Part Number 878275 Vectorbeam Operator Manual 3 5 3 6 3 7 Page 25 VB6 vacuum system The VB6 vacuum system is divided into three main volumes 1 The airlock or on machines with Brooks handling the Brooks transport module TM chamber and load aligner LA airlock 2 The storage pouches substrate chamber and lower electron column 3 The column Each is separated from the other by pneumatically operated valves These volumes may in general be vented independently with dry nitrogen as required Vacuum controller PICS or Brooks control On systems with the single chuck or 10 chuck handling option a control system known as the plinth interlock control system PICS controls the vacuum system On systems with the Brooks handling option the CTC control computer controls the vacuum In both cases the system functions include e The pump down sequence e Monitoring the pressure sensors on the column For invalid conditions an alarm signal is generated and measures are taken to prevent damage to the column e Bakeout sequence e Co ordinating the holder loading unloading Vacuum control panel The vacuum control panel shows a schematic of the vacuum and the holder load unload system and has display LEDs to indicate the status of the various valves vacuum gauges and the load unload system See Figures 3 1 3 2 3 3 and 3 4 Also attached to this panel there i
211. ections will be reset to the reference height i e zeroed whenever a QCAL MAIN LOAD command is used Jobcal and Fullcal automatically set the focus correctly before calibrating the deflection In addition height corrections are recalculated only after each stage move during exposure or after the command QADJUST FIELD is issued On completion of pattern writing the height correction will be left set up for the height of the last field written 9 1 5 Yaw correction 18 bit and 20 bit VB6 only The rotation of stage superplate with stage position is known as the yaw The yaw at any position is mostly repeatable and is measured and compensated for by the magnetic map calibration However any non repeatable component of yaw will directly contribute to the placement errors on the substrate In addition the yaw can vary over distances of only a few millimetres and the magnetic map which is measured at a grid of points separated by larger distances will not contain this information Finally the magnetic map correction is taken as zero around the datum plate for all holders although there might really be an offset in the rotation at the datum marks of different holders due to a difference in the positions of the marks of a few millimetres This offset will lead directly to an offset in the deflection field rotation Part Number 878275 Vectorbeam Operator Manual Page 70 The VB6 has two interferometer axes measuring the Y position
212. ects the reproducibility of zeroing the interferometer The number of parallel fine scans PARASCANS is typically set to 2 This number must be increased for greater noise immunity only if OVSAM and FILT are already set to maximum The number of parallel scans should however be increased if greater immunity to mark edge roughness is required The measurement length MLEN is typically set to 0 5 um for marks with sharp edges such as the calibration marks on the holders The mark edge sharpness is the edge slope multiplied by the thickness of the mark This is the distance between where the edge starts and where it finishes Some marks with less sharply defined edges may require this parameter Part Number 878275 Vectorbeam Operator Manual 11 5 11 6 11 7 Page 95 to be increased The measurement length should be minimised for faster mark location but its value should be at least the sum of the following e The mark edge sharpness usually lt 0 1 um e The spot diameter usually lt 0 1 um Note e MLEN must be smaller than the mark width height or limb width e The rise time RT must be smaller than the measurement length 10 The cross search offset CSOFF is the distance in um by which the coarse search scan is offset from the cross centre 11 The coarse search length for crosses CSLEN should not be set larger than necessary to avoid long mark locate times The CSLEN parameter is required to set the coarse se
213. ed position This may be done by the following sequence VB_OPER gt QMOVE FM VB_OPER gt QLOCATE FM OUTPUT REL VB_OPER gt QDJUST DATUM OUTPUT REL Note Never use QLOCATE FM FM as this overwrites the defined position FM Example jobfile The example job file below adjusts the datum offset based on the expected and observed mark positions and then sets up the ontime function to run itself again after a specified time interval SDP COM Adjusts the datum offset based on the expected and observed mark positions Can use any mark Syntax P1 mark type mandatory P2 mark position mandatory P3 ontime interval optional leave blank if job not to repeat at intervals Logicals reset_datum_interval ontime interval optional this logical can be defined as the time interval format hh mm ss which overrides p3 and allows Part Number 878275 Vectorbeam Operator Manual Part Number 878275 Page 217 the time interval to be changed on the fly Examples l mcjobs sdp FM FM 00 00 05 job is run every 5 secs at fm mcjobs sdp FM FM job is run once only MarkType p1 MarkPos p2 Timelnt p3 IF F TRNLNM reset_datum_interval NES THEN QDISPLAY COMMENT NOTE THAT THE TIME INTERVAL FOR SDP WAS SET BY THE LOGICAL RESET DATUM INTERVAL TimelntZF TRNLNM reset datum interval ENDIF emmaerr Remember
214. ed for by adjusting the doses of the shapes in the pattern The required doses for each shape in the pattern can be found by Part Number 878275 Vectorbeam Operator Manual 13 3 2 1 13 3 2 1 1 13 3 2 2 13 3 2 3 13 3 3 13 3 3 1 Page 122 1 A theoretical calculation 2 Measuring exposures Theoretical proximity correction calculation The inter and intra proximity effects are calculated from the pattern geometry and the backscattered electron intensity distribution Proximity effect calculation is usually done by the pattern converter and both Caprox and Cats will carry out proximity correction if the appropriate licenses have been obtained The relevant user manual should be consulted for details If the pattern contains many shapes the calculation can take many hours even on a state of the art workstation Backscattered electron intensity distribution The backscattered electron intensity profile can be obtained from either 1 A Monte Carlo simulation of beam substrate interactions 2 A series of dot exposures at different doses followed by an evaluation of the size of the resulting disc after development Empirical proximity correction A series of exposures is measured and the optimum dose for the various parts of the pattern is determined This will become impractical for increasing number of shapes Other techniques Less complicated faster but less precise methods of reducing the fidelity loss from th
215. ed mark locations except e The four marks at the extreme corners of the grid e Two adjacent locations horizontally or vertically when a warning about holes will be issued Part Number 878275 Vectorbeam Operator Manual 18 10 3 Part Number 878275 Page 189 More than 5 of locations fail in which case a warning is issued If there are any failures it is advisable to correct the problem The tolerance is intended to allow the cause of the failure to be found The following jobfile shows the use of the stepper lens calibration command for a Jobfile to calibrate stepper lens correction SJOB CAL STEPPER LENS COM ON CONTROL C THEN GOTO FINISH ON CONTROL Y THEN GOTO FINISH Save logfile name and ensure logging is on logfilename FSTRNLNM VB_LOGFILE SLON QSET MODE FAB Create new log file and ensure it s turned on qdisplay comment Redirecting the log for a stepper lens map OPEN WRITE NEWVERSION MCLOG STEPPER LENS LOG CLOSE NEWVERSION define group user nolog VB LOGFILE MCLOG STEPPER LENS LOG SLOF SLON qdisplay comment Started calibrating stepper lens map F TIME INQUIRE START MOVE TO START POSITION LOWER LEFT MARK AND PRESS ENTER MVRL 0 0 SPO START DSPO START Initialise map first QCALIBRATE LENS fm fm 1 1 INIT LOAD Calibrate map QCALIBRATE LENS 10UM SQUARE FINE START 1 5 1 ORDER 5 GRID 15 XORIGIN 0 25 YORIGIN 0 25
216. ed to apply additional non linear corrections The origin of the non linear corrections can be superimposed on the direct write mapping at any stage position This enables a direct write mapping to be set up for the whole wafer Part Number 878275 Vectorbeam Operator Manual 18 10 1 18 10 2 Page 188 and then the stepper lens corrections to placed over each stepper lens field in turn before exposure In this way the direct write mapping and the stepper lens correction remain the same The stepper lens map origin is in the bottom left hand corner of the map The act of placing the lens map origin at a point in the direct write mapping is referred to as assigning the lens map It may help to think of physically fixing the lens map to the wafer with its bottom left hand corner at the position specified and exposing patterns through it as if it were a stencil The assigning and de assigning of lens maps is done with the qmap lens command Stepper lens calibration substrate The patterned substrate used for calibrating should have the following 1 The pattern should be a regularly spaced grid of identical marks The grid must have marks in rectilinear columns and rows but the number of columns and rows can different and the spacings of the columns and rows can be different 2 The grid should at least cover the area in the stepper lens field to be used later for devices Ideally the corners of the grid should be as close as possible t
217. ed to be the mode about to be mapped into It must be the mode about to be mapped into because once the mapping has been done and the new mode loaded it is expected that after typing QMOVE POS expected position of E1 gt the mark E1 is exactly on axis Of course if E1 is defined to be in Absolute mode and then QMOVE POS E1 is typed E1 will not be on axis To be unambiguous the command SSPO DWMAP new mode could be used when setting up expected positions The software does not insist that the mode of expected positions MUST be the mode about to be mapped into although this is the default Part Number 878275 Vectorbeam Operator Manual 20 2 20 2 1 20 2 1 1 20 2 1 2 20 2 1 3 Page 233 If the command QMOVE POS REL SPO is used to set up the expected positions in a mode other than Absolute then this implicitly bases the expected positions on the results of previous mark locations A general direct write alignment method for regular rectangular arrays An alignment scheme and associated jobfiles have been developed for the Vectorbeam enabling global alignment with the option of die by die alignment to any regular rectangular array of marks The jobfile parameters can be chosen to give a variety of different alignment schemes The main steps described below in this general alignment scheme are 1 Optional alignment of reference feature 2 Alignment of first global mark 3 Alignment of a further 1 2 or 3
218. eeping the mouse button pressed down slide the arrow down to Stigmator Balance then release the mouse button A Stigmator Balance adjustment window will appear The size of a window may be increased which will elongate the slider bar ranges so making manual adjustment finer In the bottom section of the Status Window the SEM FAB button may be pressed with the mouse to switch between SEM and FAB mode The Beam On Beam Off button switches between Beam on and Beam off The Abort button may be selected to abort any of the calibrations or to abort a pattern exposure the Pause button may be selected to pause a job file it pauses at the end of the current command To resume after either an Abort or a Pause it is necessary to select the Continue button Emma File Set Display Toolkit Help EO Conditions PG Conditions Stage Conditions EHT 00 kv Resolution 0 0000 nm Mode Absolute Flament 00 W Max Held Size 0 0000 mm X 000 000000 mm wehnlt 0 0 pA vu o Y 000 000000 mm Beam current oo nA Resist Sensitivity Oo pcicm 2 Spot diameter nm Base Clock 0 000 kHz Height Mode RealTime Job WeefRcive Pattem None Selected Sem Beam Off Peart Continue Pause Figure 7 1 The Emma Status Window The second window to appear on starting Emma is a Job Control window where all Emma commands are entered Figure 7 2 Part Number 878275 Vectorbeam Operator Ma
219. eft L shapes 11 10 2 3 2 Shear A line running along the centre of each finger is extrapolated to the midpoint of two L shapes The shear is the distance between these two points The effects of a rotation on the mark are eliminated Four values are output e theshear in Y for the two upper L shapes e the shear in X for the two right L shapes e the shear in Y for the two lower L shapes e the shear in Y for the two left L shapes Part Number 878275 Vectorbeam Operator Manual Page 104 11 11 Actions if the mark locate consistently fails to find the mark If the mark locate has worked previously with a particular mark definition and substrate then the problem is likely to limited to just the video gain and contrast settings Manually adjust the video gain and backoff and retry If the mark locate has worked previously with a particular mark definition and substrate but the beam current has been reduced the problem is likely to be limited to too little filtering and oversampling used to eliminate the effects of noise The filter and ovsam values should be increased and the mark locate retried The situation that the mark locate consistently fails to find the mark is encountered most commonly when trying to locate a new mark type or a mark on a new substrate In this case there are several possible causes and actions to try 1 In SEM mode adjust the video backoff and gain manually to give a g
220. eld may contain a maximum of 16384 exels in the X and Y directions 20 bit The pattern generator consists of a crate of digital electronics and a crate of analogue electronics The digital part of the pattern generator has a 20 bit output i e the maximum main fieldsize is divided into 1048576 exels The 20 bit DAC is actually 2 18 bit DACs in a compound arrangement The area of the mainfield which is covered by the subfields can be selected by the operator in the form of a square which is centred in the mainfield This defines the maximum area which can be used for exposing patterns The number of subfields in the X and Y directions into which this square is divided is fixed at 64 4096 in total Each subfield contains 16384 exels in the X and Y directions Pattern generator digital and analogue crates The pattern generator digital and analogue crates are shown in Figure 4 4 The pattern generator digital crate contains 4 PowerPC microprocessor boards One board functions as the Master Microprocessor MUP another board as the Digital Correction Processor DCP and two boards as Pattern Generator Processors PGPs each of which are a combined Flattening Instancing Sort Processor FISPS and Shape to Line Converter SLC 4 3 1 2 1 3 MUP The MyP Master Microprocessor sequences and controls the operation of the Pattern Generator during pattern writing 4 3 1 2 2 3 PGP The PGPs Pattern Generator Processor each combine FISP and S
221. ended Please be aware of the possible hazards see below and follow the instructions in this manual When operating the Vectorbeam in any way not specifically covered by this manual or in any way that could present a safety hazard then please contact your Vistec service representative This manual is aimed at normal operation of the Vectorbeam l e operation by users based at the operator workstation or machine operations involved with loading substrates wafers It does not cover service operations which would be carried out by trained service personnel For Technical Service advice Contact the local Vistec service engineer Or use the worldwide contact telephone number 800 2255 34 22 Part Number 878275 Vectorbeam Operator Manual Page 2 Additional safety information a Intended use of the vectorbeam and chemical hazards The Vectorbeam is intended for use e Inaclean environment with environmental control e Installed as per the VB6 Pre installation Instructions document 878215 Or for the VB300 see document 893219 or for AWH Automated handling VB6 systems see document 893150 e As a tool for writing electron beam patterns on substrates using the base process covered below The VB is not intended for use with hazardous chemicals and no hazardous chemicals are used in its operation However it is possible that very small amounts of hazardous chemicals vapours could be generated inside the ch
222. ently done with a clockfile Notes for CATS converter users 1 The clocks in both CATS and Emma are numbered starting from 0 2 The logical TE CFA must be defined as the ccfa file before starting writefile if relative doses are to be included in the pattern file For example on VMS systems define te cfa dkal00 base users pattern pattern ccfa On Linux systems first set the TED environment variable to the current working directory before starting Cats TED pwd On Linux systems select the ccfa file after starting Cats Command cfa CFA option TED lt filename gt ccfa On Linux systems before using writefile TE_CFA TED lt filename gt ccfa export TE_CFA Proxeco Define the logical te_cfa as the pec file before running writefile in order to obtain the relative doses in the pattern file CFA facility 1 CATS CFA facility only supports 256 clocks Part Number 878275 Vectorbeam Operator Manual Page 124 2 In the CATS CCFA file remember to list the relative doses first followed by the line VBMINMAX min relative dose max relative dose and then the selection rules e g 1 00 1 06 1 13 1 19 1 25 1 31 1 38 1 46 VBMINMAX 1 00 1 46 Rule 0 WIDTH 0 0 07 7 Rule 1 WIDTH 0 07 0 09 6 Rule 2 WIDTH 0 09 0 10 5 Rule 3 WIDTH 0 10 0 11 4 Rule 4 WIDTH 0 11 0 19 3 Rule 5 WIDTH 0 19 0 4 2 Rule 6 WIDTH 0 4 0 6 1 Rule 7 ALL 0 13 3 5 Notes for Caprox conve
223. et if needed for the above chemicals The Vectorbeam is NOT intended for use in an explosive atmosphere b Potential Hazards and protective measures The following potential hazards exist with the Vectorbeam this is not intended as an exhaustive list but covers the main hazards Part Number 878275 Vectorbeam Operator Manual Page 3 e Laser hazard inside the chamber e X rays from the column and chamber e Electrical Shock any cabinet or cable containing electrical items e Mechanical moving parts especially the airlock loading automation chamber As above there are protections against all of these hazards in normal operation via protective covers and other protections such as interlocks These should not be removed or adjusted except by trained service personnel D Never operate the machine with wet or damp hands Electrical WARNING shock may occur if there is improper grounding or electrical leakage D DO NOT operate the machine if protective covers are missing WARNING Sound levels in normal operation do not exceed 70 dBA no protective measures needed c Master Power disconnect The master power disconnect device for the Vectorbeam is located near the Mains switch unit MSU between the power feed from the Uninterruptable Power Supply UPS and the MSU d Explanation of safety labels and notices See section Meaning of labels found on machine components for an explanatory list of safety label
224. eters can be displayed and set when logged onto the stage microprocessor For the following descriptions of interactions note that the current settings of any parameter can be displayed by simply typing the parameter name at the prompt without assigning a value e g gt double SetDelayX g will display the current fixed X settling time in us and gt double SetDelay Y_g will display the current fixed Y settling time in us 16 bit pattern generator The settling times can be changed directly by entering gt double SetDelayX_g double value in us gt double SetDelayY g double value in us The stage settling values will be reset to any values specified in the EMMA CTRL SV CONFIG VW file whenever the stage is rebooted These values are set by the same statements in the SV CONFIG VW file as above double SetDelayX g double value in us double SetDelay Y_g double value in us 18 bit and 20 bit pattern generators The stage settling times can be configured to be either fixed as for the 16 bit pattern generator or intelligent The operation in the fixed mode is as described above In the intelligent mode the BEF signal is monitored after each stage move and when the peak to peak variation in the BEF signal falls below a specifiable level the stage is flagged as settled This optimises the time for each move and usually results in lower overall settling times The intelligent mode is selected by logging onto th
225. ettings which still give a recognisable octagon Return lens 2 to its starting point 6 If the image moves significantly note the direction of movement while increasing the lens 2 setting Make an adjustment to the gun alignment shift x and shift y settings so as to move the image a few microns in the same direction 7 Adjust the gun alignment tilt x and tilt y settings in order to roughly maximise the brightness of the SEM image 8 Repeat from step 5 until the shift x and shift y settings have been found which give no large movement of the image when lens 2 is varied The shift x and shift y settings should not be changed after this step 9 In order to accurately set up the tilt x and tilt y move to the Faraday cup and use the measure curdis exe program to measure to current distribution over the field VB_OPER gt MEASURE_CURDIS 400 1 00 for VB5 VB OPER MEASURE CURDIS 400 1 75 for VB6 10 Use the calc curdis exe program to display the results VB OPER CALCULATE CURDIS 400 11 Adjust the gun alignment tilt x and tilt y settings Instead of using the sliders use the command dgun to find the current settings and then use the command sgun to apply small changes to the settings 12 Repeat from step 8 until the tilt settings have been found which give the best current distribution 13 These gun alignment settings should be optimum for all beam currents less than the one that has been used for the set up 14 Not
226. eue will take over the oper status and control of the VB Re booting the subsystems remotely If the subsystems require rebooting it may be possible to do this remotely using software commands Pattern generator Log into the MUP microprocessor and type gt reboot 5 which will reboot all the PG microprocessors Alternatively log into each PG microprocessor in turn and type gt reboot 2 Stage Log into the stage microprocessor and type gt reboot 2 Part Number 878275 Vectorbeam Operator Manual Page 245 21 5 3 EO Log into the eo microprocessor and type gt reboot 2 Part Number 878275 Vectorbeam Operator Manual 22 22 1 22 1 1 22 1 2 22 2 Page 247 Logfiles and logging Emma logging File All Emma commands can be recorded in a logfile The logging is enabled by default when Emma is started and can be disabled as follows VB OPER QSET OUT OFF VB OPER QSET OUT OFF The command needs to be repeated due to a bug in the software as does the command to enable the logging VB OPER QSET OUT ON VB OPER QSET OUT ON The default logfile is sys login esprit log dat The logfile name is read from the vb logfile logical when the logging is restarted The logfile can be changed by the following commands VB OPER QSET OUT OFF VB OPER QSET OUT OFF VB OPER define group user vb logfile disk directory logfile log VB OPER QSET OUT ON VB OPER QSET OUT ON Note that the
227. evel in the thermostatic unit is too low LED SYSTEM WATER red is lit when the water level in the thermostatic unit is too low LED UPS is lit when there is fault with the uninterruptable power supply 3 7 6 2 Modifications to FEG vacuum system Part Number 878275 Recent modifications to the column pumping are The gun penning G5 has been removed The valves V6 and V9 and the nitrogen admit have been removed IGP3 is connected directly to the column with no branches The gun and the IGP4 are connected to the chamber with valves V17 and V18 for isolation respectively Recent modifications to the chamber pumping are The vacuum reservoir has been removed Valves V12 V19 V20 and V21 have been removed The turbo bypass pipe has been removed The airlock vent button has been disconnected The airlock full vent must be used Vectorbeam Operator Manual 3 7 7 Page 32 Vent system button The vent system button requires two presses within 5 seconds before the command is accepted When the command has been accepted the LEDs above the two vacuum buttons flash for 35 seconds during which time the pump system button can be pressed to cancel the vent command The vent system command can be used at any point to vent the chamber airlock and optionally the gun If a bakeout is in progress then it resets the bakeout before venting If the ion pump key switch is in the vent disable position the gun is isolated befor
228. f 10 um octagon FM Figure 16 4 Photo showing SEM monitor with image of gun target GT 16 5 7 Datum target layout The datum target layouts for the various holders are shown below Part Number 878275 Vectorbeam Operator Manual Page 153 16 5 7 1 VB5 Figure 16 5 VB5 datum target 16 5 7 2 VB6 excluding 8 inch wafer and scalpel mask holders 20um PERIMETER SCRIBE ALLY LEICA LOGO EN OFF NOTE ORIENTA Big peerage ee TEEEEEEH a i ie f f i H Figure 16 6 VB6 datum target Part Number 878275 Vectorbeam Operator Manual 16 5 7 3 16 5 7 4 16 5 8 Page 154 VB6 8 inch wafer and scalpel mask holders Figure 16 7 VB6 datum target for 8 inch wafer and scalpel holders Detail of datum marks SQUARTS CONTAINING DOT ARRAYS FOR Site Auu ALIGN ur NT FT MARKS Wa SQUARE 1 VCONCDITR CIRCLES EMO Y S Y x mati ONAL Figure 16 8 Detail of datum marks Datum mark contamination When a mark on the datum is scanned the electron beam breaks up molecules of hydrocarbons that are physi sorbed and mobile on the surface of the datum plate to form a solid with a high carbon content In time this deposit affects the backscattered electrons and becomes visible in the SEM image see Figure 16 9 It reduces the apparent edge sharpness making calibration and the adjustment of focus and stigmation less accurate A simple way to quantify and m
229. f Cats resolution by a factor of 32 128 or 512 for 16 bit 18 bit or 20 bit pattern generators respectively Formats VB20 VB50 or VB100 should be chosen to match the EHT setting on the VB of 20 50 or 100 kV respectively The default limits in Cats are in the tables below Format VB VB Minimum Maximum Minimum Maximum Maximum Maximum Field Max field Resolution Resolution VB VB VRU default um um um um Resolution Resolution field um um count VB Not Not used 0 002 0 0625 Not used Not used 32 65504 used 64000 VB20 1024 1024 0 002 0 5 0 002 0 015625 32 65504 64000 VB50 819 2 819 2 0 002 0 4 0 002 0 0125 32 65504 64000 VB100 589 824 589 824 0 002 0 288 0 002 0 009 32 65504 64000 Part Number 878275 Table 19 1 Default limits for 16 bit machines Vectorbeam Operator Manual Page 195 Format VB VB Minimum Maximum Minimum Maximum Maximum Maximum Field Max field Resolution Resolution VB VB VRU default um um um um Resolution Resolution field um um count VB Not Not used 0 0005 0 015625 Not used Not used 128 262144 used 256000 VB20 1024 1024 0 0005 0 5 0 0005 0 00390625 128 262144 256000 VB50 819 2 819 2 0 0005 0 4 0 0005 0 003125 128 262144 256000 VB100 589 824 589 824 0 0005 0 288 0 0005 0 00225 128 262144 256000 Table 19 2 Default limits for
230. f field consistent with sman One exposure field may contain groups of lines exposed with different relative doses The Emma command is VB_OPER gt aset grating n width2 width of line height2 height of line pitch x lt x_pitch gt pitch y lt y_pitch gt start x position y position number of lines number relative dose relative dose shape meander number of shapes exposed together before switching subfields gt block lt blocknumber between 1 and 19 offset X lt gt Y lt gt scale X Y Part Number 878275 Vectorbeam Operator Manual Page 261 Block 0 Grating 1 12 lines peram NNIT X start position X pitch X Centre of block Block 1 Grating 2 10 lines Relative dose 1 2 Block boundary a Figure 24 2 Diagram showing the definition of parameters used in the qset grat command 24 9 1 1 24 9 1 2 24 9 1 3 24 9 1 4 24 9 1 5 Segment nis the segment number which is an integer between 1 and 64 The command allows up to 64 sets of parameter conditions to be defined per block The algorithmic generation function works through each segment in numerical order i e the order defined by n Height width pitch and start The height width pitch and start dimensions are in microns The origin for the start position is the centre of the field Block The block number allows multi field patterns to be defined and can be between 1 and 19 If the block
231. f the mapping coefficients from the design expected positions of the pattern already on the substrate and the same distances as measured observed on the machine e g VB_OPER gt QMAP DWCOORDS WAFER EXPECTED E1 E2 E3 E4 OBSERVED 01 02 03 04 The direct write transformation using four marks does not function if the marks are in a cross pattern Imagine distorting a square into a keystone shape The midpoints of the sides of the square do not move at all Consequently it can be seen that the keystone values are not defined by marks in a pattern If a substrate already has marks defined in a pattern then use three mark alignment 2 The second command is the loading of the coefficients into the stage and deflection systems VB_OPER gt QMAP DWMODE WAFER LOAD lt i Note that if the qmap dwenter command is used it immediately issues a qmap dwmode load command automatically if the mode named in qmap dwenter is the current mapped mode 20 1 3 1 Advanced note on expected positions The expected positions are taken from the CAD data and supplied to the QMAP DWCOORDS command as position identifiers The position identifiers are set up using the QSET SPO command It is expected that under no circumstances are they obtained by mark location because this would give an observed not expected position Consequently it is only the numbers themselves that matter for the expected positions the mode is assum
232. fixed bit pattern not present Failure of Ellink primary 5V supply Elbus addressing error Multiple Ellink errors Failure of Ellink secondary 24 V supply Failure of Ellink secondary 24 V supply Ellink secondary hardware error CCU Operation requested not supported CCU Unable to initialise serial link to SAEHT CCU Actual EHT values do not match expected EHT values CCU Supply ramping CCU SAEHT already initialised CCU Cannot read calibration file AGA Number of tilt steps 0 AGA Number of shift steps 0 AGA EB ADC signal overload AGA Insufficient memory for the number of steps specified AGA Background noise too high to perform reliable AGA AGA EB ADC signal too low AGA Unknown EB ADC signal specified Channel Plate Refresh timed out EB ADC filter value out of range Previous AGA command still running EHT driver not initialised Filament mode out of range Wehnelt mode out of range Gun align bad aligner number argument Gun align bad drive value argument Gun align bad signal number argument Gun align bad kilovoltage value argument Gun align analogue readback out of tolerance Gun align digital readback out of range Gun align settling time delay failed in SetDrive Gun align settling time delay failed in SetEht Gun align kilovoltage value not 0 20 50 or 100 Image Processor Video overload Vectorbeam Operator Manual Vistec Lithography Ltd PO Box 87 515 Coldhams Lane Cambridge CB1 3XE UK Telep
233. ft button the slider bar on the right hand side of the Emma help window keeping the mouse button pressed down move the slider bar to scroll as necessary Part Number 878275 Vectorbeam Operator Manual 8 1 8 2 8 3 8 4 Page 61 Job control The VB user terminal operating system is VMS from HP Compaq The Emma program running on VMS provides a Status Window and a Job Control Window The Job Control Window is similar to a normal DECterm in VMS in that it processes DCL commands In addition the Job Control Window processes Emma commands DCL and Emma commands can be freely intermixed and command sequences are usually saved in files for repeated use A knowledge of both DCL and Emma is therefore required for operating the VB DCL commands Refer to the HP Compaq manuals for information about DCL Emma commands For details of Emma commands refer to the manual Emma Vectorbeam Command Set part number 878274 qset register commands Many variables on the sub systems can be set using the qset register command see manual VB Software Registers part number 893039 Supplied job files A number of job files have been written to carry out all common sequences These are described in the Acceptance Test and Operator Jobfiles User Manual part number 892777 and are referred to in this manual The areas these jobfiles cover are 1 Diagnostics 2 Utilities 3 Calibration 4 Expose layout 5 Acceptance te
234. further function PROVIDED its status is done Logicals Many logicals are set by Emma as a way of making parameters visible and readable by jobfiles They all begin with the prefix vb Useful logicals VB PG TYPE This can take the following values 16 BIT STANDARD 18 BIT WIDEFIELD 20 BIT WIDEFIELD or 00 UNKNOWN This logical is set when an INCM or INPG or INCM SYS PG is done The logical is set only by by versions of Wide Field Emma dated 20 11 00 or later Creating compiled jobfiles Sometimes it can be useful to write jobfiles in C instead of DCL The following are required 1 An ascii editor This is usually the edt or eve editor under VMS 2 AC compiler for VMS Not supplied as standard 3 emmachk h and emmachkj obj Available in the VB software release on the tool in the emma ctrl com directory The C program simply issues standard Emma commands to the mailbox via the emmachk function see emma ctrl com emmachk example c which is installed on each tool for an example The command used in the emmachk function must be the full form no mnemonics without the q at the start Part Number 878275 Vectorbeam Operator Manual Page 65 The other 2 functions in emmachk obj that will be needed are logical_getg and logical_setg to read and write group logicals respectively After compiling my_program it must be linked with emmachk obj A compiled job is run by 1 defining a symbol at the VB_OPER prompt
235. ge in the Y direction to the next feature and observe in the microscope whether the second feature is shifted to the left or right relative the first Or move the stage in the X direction and observe in the microscope whether the second feature is shifted up or down If the relative movement is more than about 3 um per mm distance between the two features then an adjustment to the rotation should be made as described below Adjusting the rotation VB5 Make an adjustment to the table rotation using a screwdriver on the adjusting screw on the side of the holder Measure the rotation of the wafer as described above Repeat the adjustment until the rotation is 3 um mm VB6 Make an adjustment to the table rotation using the thumb screw to the side of the substrate Measure the rotation of the wafer as described above Repeat the adjustment until the rotation is 3 um mm Holder loading unloading in from airlock Use gloves while handling the holder and substrate VB5 Press airlock full vent on vacuum control panel twice in quick succession The green vacuum LED will go out Wait several minutes for the airlock to reach atmospheric pressure Lift up the airlock lid and carefully remove the library so as not to hit the ends of the datum plate mounts on the back of the holders against the side By unhooking the arm across the top of the library only the top half of the library need be removed if desired The holders can then be removed
236. gnetic field between the final lens and the stage and partly by small changes in the mechanical rotation of the stage superplate Unless this effect is cancelled it will cause the misplacement of exposed shapes A calibration is therefore carried out in order to align the deflection field axes to the stage axes over the range of the stage travel The deflection field is first aligned to the stage axes at the datum plate by jobcal and then the relative rotation of the deflection field is measured at each of an array of marks on the autostitch plate These relative rotations are known as the magnetic map and are applied to the deflection field at every stage position except around the datum plate 9 1 4 Height corrections The distance between the substrate surface and the final lens will vary typically by 10 um due to substrate distortion and differences in the mounting in the holders The effect of this on the fieldsize is shown in Figure 9 2 creating a dependence of field scale on height of about 25 nm mm per micron height This effect occurs for both HR and UHR systems but in addition on UHR systems there is a dependence of field rotation on height as the angle of the beam to the substrate has a tangential component as well as the radial component The magnitude of this effect is about 16 nm mm per micron height Part Number 878275 Vectorbeam Operator Manual Page 69 Stage Position Stage Position B n A n h Holder
237. h a clock number and in addition has defined in the pattern file header the maximum and minimum relative doses that the maximum and minimum clock numbers represent For exposure these relative doses must be transformed into absolute doses This is done by defining the resist sensitivity also known as the base dose or the large area clearing dose on the VB as usual and shapes with a relative dose 1 will be exposed with this dose When the pattern file is selected by Emma the dose controller low and high limits are set up automatically based on the range of relative doses and the resist sensitivity A check that the pattern file contains relative doses can be done by verifying that the message Pattern has internal clockband consisting of relative doses from to is typed up when the pattern is selected Note that although in principle all 65536 different doses can be used by this method the converter may limit the number of doses The 32 clock limit in Emma does not limit the number of clocks which can be used with this method Typically it is more convenient for a theoretically calculated proximity correction to use method 2 and for an empirically determined proximity correction to use method 1 This is because a theoretically calculated correction may have a large number of different doses which would be inconvenient to transfer to a clock file and an empirically determined correction may require tweaking at the machine which can be conveni
238. hange tool Slide the holder into the substrate exchange tool Close the Perspex jig down on the holder and load unload substrate Lift up the Perspex jig and remove the holder VB6 D Keep fingers clear when operating the pneumatic switch A pinch WARNING hazard exists when using the loading jig 16 2 2 1 A loading jig is used for exchanging masks in the holders The loading jig is pneumatically operated with two valves one causes the holder to be clamped to the jig and the other releases the substrate support springs The substrate support springs are released by a universal flat plate which is pushed up under the holder Ensure that the switches on the substrate exchange tool are set to holder unclamped and substrate clamped Place the holder into the substrate exchange tool Warning The holder clamp must be energised first and released last Operate switch to clamp the holder Operate switch to unclamp the substrate Load unload substrate Operate switch to clamp the substrate Operate switch to unclamp the holder For mask plates there is an earth clip alignment jig which ensures not only that the earth clips are correctly positioned but also provides a continuity check to ensure that the clips are making contact with the metallisation Piece part holder The piece part holder for the VB6 consists of a metal insert for a 6 inch mask holder into which a sprung plate with dimensions roughly 140 x 90 mm is built The
239. hardware fault Vectorbeam Operator Manual Part Number 878275 Page 299 0x20AA0004 CCU Parameter is out of range 0x40AA0005 CCU Low voltage psu failure Ox40AA0006 CCU Logic psu failure 0x40AA0007 CCU ElLink Primary failure 0x40AA0008 CCU ElLink Secondary failure 0x40AA0009 CCU ElLink Logic psu failure Ox40AA000a CCU Ellink 24V psu failure Ox40AA000b CCU Ellink 24V psu failure Ox40AA000c CCU Mulitiple ElLink faults Ox40AA000d CCU Hardware addressing error Ox40AA000e CCU Lens driver over temperature Ox40AA000f CCU Lens current source failure 0x40AA0010 CCU Gun tilt current source failure 0x40AA0011 CCU Gun shift current source failure 0x40AA0012 CCU Eht is disabled 0x40AA0013 CCU Eht is turned off 0x40AA0014 CCU Filament is disabled 0x40AA0015 CCU Beam is on 0x40AA0016 CCU Generic software error Ox20AAO0F01 SAEHT Unknown command Ox20AAO0F02 SAEHT Unknown parameter Ox20AAOFO03 SAEHT Wrong unit for current mode Ox10AAOF05 SAEHT Busy command queued Ox20AAOF06 SAEHT Busy command rejected Ox40AAOF07 SAEHT Unit badly earthed Ox40AAOF08 SAEHT Bad vacuum Ox20AAOF09 SAEHT Unit not switched on Ox20AAO0FOA SAEHT Interlocks overridden Ox20AA0FOB SAEHT No LaB6 command rejected Ox20AA0FOC SAEHT No FEG command rejected Ox40AA0FOD SAEHT Safety interlocks not OK Ox40AA0FOE SAEHT Filament broken Ox20AAO0FOF SAEHT Undefined Runup profile Ox20AA0F10 SAEHT GSC power off Ox40AA0F12 SAEHT GSC driver not reset O
240. he datum plate 16 5 5 Holder parameters The holder parameters for all holders in use on a machine are kept in the file Holder_table com The file should be edited when using a new holder or when switching to a new focus mark because the old one is contaminated 16 5 5 1 Standard holder parameters for VB5 mm mm mm position mm 3 3 7843 4 3 7843 3 3 7843 3 3 7843 4 3 7843 3 3 7843 3 3 7843 4 3 7843 3 3 7843 Part Number 878275 Vectorbeam Operator Manual Page 151 16 5 5 2 Standard holder parameters for VB6 Holder type FM FC GT Substrate Transmission position position position centre position sample mm mm mm mm 3 inch wafer 78 3 8 2 14941 641 7843 8 2 Low 42 60 13541 6 1 High 118 130 8 inch wafer 742 7 2 149421 641 7 2 7 2 76 5 90 5 3 3inch mask 78 3 8 2 149 1 6 1 7843 8 2 76 5 63 MENN 4inch mask 78 3 8 2 149 1 6 1 7843842 76 5 75 1354 1 641 5inch mask 78 3842 149 1 641 784 3822 76 5 88 1354 1 641 6 inch mask 78 3 8 2 14941 6 1 7843 8 2 76 5 90 5 13541 6 1 16 5 5 3 Brooks holder parameters for VB6 Holder type FM FC GT Substrate Transmission position position position centre position sample mm mm mm mm 5 inch wafer 25 3442 14941 641 25434242 76 5 90 5 2 1 1941 electrostatic 6 inch wafer 78 3842 149 1 6 1 78 3 8 2 76 5 90 5 13541 6 1 6 inch wafer 25 3422 149 1 6 1 25 3 4 2 76 5 90 5 13541 6 1 mechanical
241. he four equations describing the horizontal edges are averaged to give a horizontal line in the centre of the cross The intersection of the two lines is the position which is then returned to the user or calling routine Note This method is very susceptible to error caused by irregular limb edges and MH and MW should be as large as possible Stitch output The stitch qualifier on the mark locate command can be used when the limbs of a cross are made up of more than one finger see Figure 11 14 Vectorbeam Operator Manual Page 103 FELM 120m MILES 3m E H LI Figure 11 14 Fine scans for stitch locate algorithm on a cross with two fingers in each limb The cross can be exposed with each L at the corner of a different subfield or mainfield i e a subfield or mainfield boundary bisects the cross vertically and horizontally By measuring the separation of the L shapes the subfield or mainfield stitching can be measured 11 10 2 3 1 Separation This is half the distance between lines running along the centres of the fingers in one limb This distance is calculated for a point 25 along the cross limb from the inside corner of the L The effects of a rotation on the mark are eliminated Four values are output e the separation in X for the two upper L shapes e the separation in Y for the two right L shapes e the separation in X for the two lower L shapes e the separation in Y for the two l
242. he hysteresis of the pole pieces This takes about 3 seconds Coarse adjustment 1 Move a feature such as the calibration mark under the lens and select SEM mode 2 Set the fine focus value to the value described in the Section Correct fine focus value on datum for conjugate blanking 3 Adjust lens 2 to bring the feature in focus This can be done either Part Number 878275 Vectorbeam Operator Manual 17 4 1 1 2 17 4 1 2 17 4 2 17 4 2 1 17 4 2 2 Page 166 e by using the slider bar in the lens panel under the Set menu of the Emma status window or e by reading the actual lens setting using the command dln2 and setting a modified value using the command sIn2 value Fine adjustment 1 Move a feature such as the calibration mark FM under the lens and select SEM mode 2 Calculate the required fine focus value as described in the Section Correct fine focus value on datum for conjugate blanking 3 Find the fine focus value that actually focuses the beam on the substrate 4 Adjust lens 2 to eliminate the difference in the fine focus values found in steps 2 and 3 above Read the actual lens 2 setting using the command dln2 and set a modified value using the command sln2 value The direction of change of lens 2 is then as follows e To reduce the fine focus value that will be needed to focus the beam increase the current in lens 2 e To increase the fine focus value that is needed to
243. he mini map may be more than compensated for by the elimination of time for height readings during pattern exposure Height identifiers Height identifiers are names defined by the user within Emma to refer to height readings Height identifier names can be anything like position identifier names Part Number 878275 Vectorbeam Operator Manual 10 13 2 Page 84 A height identifier can have the same name as a position identifier without confusion Height identifiers operate in the same way for heights as position identifiers do for positions They can be created in the following ways e VB OPER qset HID lt name gt value e VB OPER qadjust field sethid lt name gt Note that that qadjust field Hid lt name gt forces in the height to be used e VB OPER qmove position sethid lt name gt x pos y pos e VB OPER qmove SPO sethid lt name gt position identifier e VB OPER qdisplay height sethid lt name gt The height sensor must be in real time mode when acquiring the heights for the map or they will just be read heights off the previous map Height identifiers can be displayed with the command VB OPER qdisplay HID lt name gt This also sets the logicals VB_DHGT_H to the height and VB_HEIGHT toa string giving the identifier s name plus its height Creating a mini height map A mini height map can be generated with 1 2 3 or 4 height identifiers and the corresponding position identifiers The height
244. he pattern then this will be known already Otherwise the design grid can be determined by inspecting the shapes making up the pattern As the shapes are defined digitally all the vertices will be an integer number of resolution steps from the origin The design grid is the maximum resolution step size for the grid in order that 10096 of the shape co ordinates lie on the grid The ongrid command in CATS can be used to find the design grid 19 1 11 Pattern generator grid snapping The 16 bit 18 bit and 20 bit pattern generators allow the user to choose the mode of grid snapping using the command Part Number 878275 Vectorbeam Operator Manual Page 212 VB OPER qset pg nofloat float split and bury nosplit As an example the results of exposing the shapes shown in Figure 19 7 with the various options are described below Flying Cat Graphics a Figure 19 7 Converter output for an array of identical squares of size 5 PG resolution points with pitch 11 PG resolution points in both X and Y Nofloat option The nofloat option only works for 16 bit pattern generators and causes 18 bit and 20 bit pattern generators to operate the float option The nofloat option causes all shape vertices to be snapped to the nearest VRU grid point to the lower left of the defined positions in the pattern file The VRU grid is a grid of points with origin at the lower left corner of the mainfield and separation equal to the VRU
245. he point will be selected as being potentially on the mark As soon as a potential point has been found a bisection search is carried out at the point If the bisection search doesn t find a mark then the spiral search is resumed Mark detected here Figure 11 6 Diagram showing spiral search for rectangular marks with pit algorithm 11 8 2 Bisection search The bisection search finds the edges to within the accuracy specified by the MLEN parameter and checks that the mark has the correct size The algorithm assumes it has found a mark if it finds two edges separated by the height or width whichever one is relevant of the mark and which has suitable edges as Part Number 878275 Vectorbeam Operator Manual Page 97 defined by the contrast If the edges are not found or the mark is not of the correct size then the spiral search is resumed If the search is successful the edge positions are passed to the fine search The same search is applied to all horizontal and vertical edges The video level at the point found by the spiral search is measured and at a point one mark width or height away If the video levels are the same for this first comparison the spiral search is resumed Otherwise the point half way between the last two points is measured If the point has a video level within 25 of the average of the mark and background levels then the edge has been found and a fine search is done Otherwise the search contin
246. he surface of the resist 2896 of this beam is reflected internally and the relative intensity of the transmitted beam is then 0 3 This is about the same as the intensity from the surface of the resist However due to refraction in the PMMA the apparent position of the Cr surface is higher and is about 0 6 of the PMMA thickness below the PMMA surface From this it can be seen that the heightmeter will read a height about 0 69 times the PMMA thickness above the surface of the Cr Due to the angle of the beam to the substrate when exposing shapes around the edges of the field the sides of the shapes in the resist after development will be sloped It is the position of the opening at the bottom of the resist which determines where the shapes will be transferred into the substrate The scale must be adjusted for the bottom of the resist for accurate stitching and this is done by applying an offset to the heightmeter table according to resist thickness so that it effectively reads the true height of the substrate surface Part Number 878275 Vectorbeam Operator Manual 10 6 2 Page 77 Height meter offset effect on stitch accuracy The effect on the fieldsize of the offset is typically small for an 800 um field and 0 55 um resist on Cr the fieldsize will be reduced by 11 nm see Figure 10 2 Scale stitch error for a 50 um block size nm 10 Height meter offset um Figure 10 2 Measured dependence of scale stitch error on heigh
247. her for all other holders which have the datum in the centre of an edge On a VB5 a dedicated magnetic map is required for any holders with the datum in the corner and another for all holders with the datum in the centre of an edge Read the following carefully for details An introduction to the magnetic map correction is described in Chapter Corrections The rotation of the deflection field relative to that at the calibration mark is measured for an array of points centred approximately about the centre of the stage This array is typically 10 x 10 points but may be varied by the operator A substrate with an array of alignment marks is used and this is normally the autostitch plate The array of alignment marks will not normally cover the entire range of stage travel as the maximum substrate size will be less than the entire range of the stage For the purposes of applying magnetic map corrections the stage travel is divided into 64 x 64 cells The magnetic map correction for each cell is calculated from a spline fit to the measured data Such extrapolation of corrections to positions outside the measurement array will not be accurate due to the nature of a spline fit This has the following consequences 1 The magnetic map must be calibrated over the area which will be exposed 2 The magnetic map corrections at the calibration mark must be zeroed Part Number 878275 Vectorbeam Operator Manual 15 7 1 Page 137 The ca
248. hines the holders are placed in a library which is in the airlock chamber A pneumatic loader arm under microprocessor control loads or unloads the holder from the library onto the stage VB6 Single chuck and 10 chuck machines There are three sapphire balls on the Zerodur mirror block on the stage which locate into blocks on the underside of the holder forming a kinematic mounting There are also three sapphire balls on the underside of the holder on which the holder sits in the airlock and pouches Care must be taken to avoid contaminating all the mounting blocks and balls The holders are placed directly on a table in the airlock A substrate handler under microprocessor control loads and unloads the holders from the airlock onto the stage The holder may also be loaded into either of the two stabilisation pouches Machines with Brooks handling There are three sapphire e chucks on the Zerodur mirror block on the stage which electrostatically clamp the metal plated underside of the holder There are Part Number 878275 Vectorbeam Operator Manual 1 3 1 3 1 Page 8 also three pins on the underside of the holder on which the holder sits in the airlock CAM and the pouches Care must be taken to avoid contaminating all the surfaces The holders never leave the vacuum during normal operation A robot in the transport module loads and unloads the holders from CAM onto the stage Electronics crates Control electronics rack 1 Th
249. hone 44 0 1223 411123 Fax 44 0 1223 211310 www semiconation com
250. ht beam current and calibration values are displayed Error messages will also appear here A full style VAX style HELP function is available for all commands and command parameters Alternatively refer to the Vectorbeam Command Set document no 878274 Hints 1 CTRL5C e g means hold down control key and press C 2 Note that DCL commands are not case sensitive e g DIR is the same as dir 3 Emma commands like DCL commands are not case sensitive Commands issued directly to subsystem processors such as the pattern generator are case sensitive e g stageNextField is not the same as StageNextField 4 Emma commands can be written in full or short form mnemonic 5 Once a set of qualifiers specifying parameters has been issued with an Emma command they will remain in force if the command is issued without the qualifiers Part Number 878275 Vectorbeam Operator Manual Page 48 7 2 Emma windows overview When Emma is started two windows will be opened The first of these is a Status Window where the current status of the system is displayed this window is continuously updated Figure 7 1 Having started Emma and run the initialisation procedure some information is immediately available from the Status Window in the menus at the top left corner i e calibrations pattern data column conditions PG and EO conditions as well as access to the joystick Select for example with the mouse the Set menu and k
251. ic gun alignment but does not work properly The manual procedures described above should be used instead Setups affected by gun alignment Changing the gun alignment will change e The focus and stigmation corrections and the mainfield distortion corrections which should be recalibrated using fullcal e The shift for fine focus coefficients which should be recalibrated using fullcal or jobcal e The stigmator balance which should be adjusted as described in the Section Stigmator balance setup Video gain and backoff set up The video gain and backoff settings are applied to the signal from the backscattered electron detectors before being passed to the SEM monitor and the mark locate electronics Mark locate beam diameter measurements and automatic focus and stigmation adjustment all require the video gain and backoff to be set up The gain affects the brightness The backoff often called black level on other systems affects the contrast and should be set so that scanning a mark produces a variation in video signal of at least about 20 96 of the total Part Number 878275 Vectorbeam Operator Manual 17 2 1 17 2 2 17 2 2 1 Page 161 range Increasing the contrast further does not usually give significant gains in accuracy The gain and backoff should be set so that all measured points are within the maximum and minimum limits of the video system The video signal levels can be viewed by selecting graph in the Set I
252. ields are processed in a boustrophedon order 4 3 1 1 3 1 SLC The SLC Shape to Line Converters take a part field buffer of shapes from the FISPs and transform each subfield shape into a series of horizontal lines for output to the LineWriter 4 3 1 1 4 1 DCP The DCP Digital Correction Processor applies real time corrections for field and subfield perturbations Pattern generator analogue crate Part Number 878275 Vectorbeam Operator Manual Bh UKE ATER FRIU GGRRECTICIME INTERFACE TO MARK 2l 4 3 1 2 LOCAL ETHERNE EMMA CONTROLS JERNET ETH LOCAL Page 40 PGA GRATE FINAL LENS ASSY Figure 4 3 16 bit pattern generator analogue crate The pattern generator analogue crate is shown in Figure 4 3 The amplifiers driving the sub field deflectors are mounted on the column 18 bit and 20 bit pattern generator hardware PGD Crate BANDWIDTH PGA CRATE COLUMN quzbeccccmeccsxenlrlcscc EEE j PSS Se Sse e mueve m eie SUBFIELD PGP1 PGP2 d PATTERN PATTERN FRACTURE FRACTURE and SORT and SORT SUBFIELD DEFLECTION COILS SHAPEWRITER i Hsec S bo IF HS CORRS M MAIN FIELD 4 i DEFLECTION COILS HSTORRSY gt MAINFIELD H MAIN DAC X DRIVE X eN FINAL LENS C3 MAINFIELD 1 WAINDACY DRIVE Y 1 VME BUS SUBFIELD No
253. ile This job will measure the positions of an n x n grid of points using mark locate and save the positions to a data file The measurements should be done in absolute mode Repeat the data collection with the substrate at 90 180 and 270 degrees rotations anticlockwise from the first rotation The substrate is removed from the holder and physically rotated by 90 degrees anticlockwise for each measurement Analyse the measurements for each rotation separately Calculate the scale rotation keystone bow and scale linearity in both x and y for a best fit of the measurements to the nominal grid positions This can be done with a program such as PVWAVE The following plots are useful for each rotation to aid analysis and the elimination of incorrect data A plot of the data which has been fitted to the nominal grid and has had the X and Y offsets and the average of X and Y rotations mathematically removed This plot shows the fit to the machine grid if substrate insertion offsets are removed A plot of the data which has had the X and Y offsets individual X and Y rotations and individual X and Y scales mathematically removed This plot shows how the fit to the machine grid could be improved if there were no substrate insertion offset scaling or orthogonality errors Calculate the correction coefficients according to the table below Add the correction coefficients to the current ones for the absolute mode and enter into Emma
254. ing Spot table For each beam energy after See Section Spot table selecting a new aperture or after calibration the source emission has changed Deflection field corrections For each beam energy after See Section Fullcal selecting a new maximum field size for which no corrections already exist in a database Height map For each substrate if operating in See Section Height map height map mode calibration Shift for fine focus After changing beam current or at See Section Jobcal suitable intervals in order to Height meter table 7 compensate for drift Focus and stigmation BEF scaling and rotation Main field scaling rotation and keystone Sub field scaling and rotation Stepper lens When exposing wafers which have See Section Stepper lens been exposed on a stepper in direct calibration write mode Part Number 878275 Vectorbeam Operator Manual 18 2 18 3 Page 178 Fullcal Fullcal is used in order to calibrate the deflection field corrections main field distortions sub field distortions focus and stigmation corrections BEF field distortions Fullcal uses the same sequence of calibration routines as jobcal to set up the references except that in addition three field correction calibrations are carried out The BEF distortion corrections are not calibrated The program must be run initially for every new field size to be used Once the calibration has been carried out the results should be saved to a data
255. ing into subfields 1 5 exposure clock frequency Stage movement time Stage movement time the sum for the pattern of Step distance to each successive block Step time The order of the exposed blocks is defined by a meander pattern The step times for distances up to about 1 mm may be estimated from the times measured as part of the acceptance tests The step times for distances 1mm may be estimated from the stage speed specification Stage settling time Total stage settling time Number of X steps X settling time Number of Y steps Y settling time The settling times may be displayed and set as described in the section Displaying and setting the settling times Heightmeter time The total heightmeter measurement time number of blocks heightmeter measurement time Part Number 878275 Vectorbeam Operator Manual 24 1 8 24 1 9 24 2 24 2 1 24 2 1 1 Page 253 Data processing time The data processing time time to transfer pattern data from hard disk to FISP input buffer over ethernet time to fracture and sort shapes into subfields time to transfer data to SLC input buffer time to fracture trapezia into lines time to transfer lines to linewriter input buffer Due to the simultaneous data processing and buffering at several points along the chain the data processing time with two 300 MHz PowerPCs with 128 Mbytes of memory is Os for most patterns Die by die alignment time The die by di
256. invalid for this command DWCOPTOSELF Can t copy current mode to itself DWMAPORDINV Mapping order invalid if given must be 1 3 or 4 DWXSCAINV X Scale value is out of usable range DWYSCAINV Y Scale value is out of usable range DWXROTINV X Rotation value is out of usable range DWYROTINV Y Rotation value is out of usable range DWXKEYINV X Keystone value is out of usable range DWYKEYINV Y Keystone value is out of usable range DWSRCMAPINV Source map mode invalid in ConvertMappedPosns STSRCMAPINV Source map mode invalid in ConvertStagePosns DWDSTMAPINV Destination map mode invalid in ConvertMappedPosns DWDIVZ3DD Division by 0 in DW back transform 3D denom DWDIVZ3DSY Division by 0 in DW back transform 3D ScaleY DWDIVZADSX Division by 0 in DW back transform 4D ScaleX DWDIVZADD Division by 0 in DW back transform 4D denom DWDIVZADCHI Division by 0 in DW back transform 4D Chi DCBANDNSET Couldn t set clock band DCUNKNHW Can t find what PG clock hardware is in use DOSTYPINV Invalid dose type DCBADFREQ Frequency can t be set out of hardware range DCFRQOUTBND Frequency can t be set outside clock band DCBADFRQADJ J Frequency can t be adjusted out of hardware range DCBADDOSE Illegal dose value zero or negative DCBADEXEL Illegal exel size zero or negative DCCMDFAIL Clock command failed DCCMDNOGOOD Clock command irrelevant to installed hardware D
257. ion points in size and which have a pitch of 11 PG resolution points in pitch The pitch and the 19 1 12 size are correct The split and bury operation can only be switched off by either selecting the Nofloat option or by using the command VB OPER qset set pg nosplit Choosing the pattern limits The first priority when choosing the pattern limits is obviously to set the limits so that the entire required pattern is included However the choice of the pattern limits is important for the correct positioning of the pattern on the substrate especially in direct write This is because it is usual to locate either the centre or the corner of a pattern at a particular stage position using the qmove pos centre Part Number 878275 Vectorbeam Operator Manual 19 1 13 19 1 14 19 1 14 1 19 1 14 1 1 19 1 14 1 2 Page 214 or qmove pos corner commands prior to exposure The limits may be set to larger values than those required just to include the pattern so that the centre of the pattern is at particular position in relation to the shapes In the direct write scheme described in the manual using the layout command it is necessary to note the distance of the centre of the pattern to the lower left corner of the lower left alignment mark This will depend on the pattern limits chosen Negative biasing Negative biasing is a technique that can be applied to improve the fidelity of the lithography The technique is t
258. irectly to the VB pattern generator resolution For the formats VB20 VB50 and VB100 the resolution is decoupled from the VB pattern generator resolution in order to allow larger values In many cases however the Resolution and VB Resolution will be the same The VB pattern generator resolution is given by the VB Resolution parameter in Cats For all formats the limits placed on resolution can be overridden using the force option with the resolution command as shown below but this increases the risk of trying to operate the machine beyond the limits of the hardware Resolution 0 0025 force Part Number 878275 Vectorbeam Operator Manual 19 1 1 1 7 19 1 1 1 8 19 1 1 1 9 Page 197 When using the formats VB20 VB50 or VB100 the VB Field parameter should be increased rather than using the force option to override the maximum limit Beamstep The beamstep defines for the purposes of pattern conversion only the exel size which will be used for exposure The beamstep size is given by VRU PG resolution The beamstep parameter can as an option be set by the user to an integer multiple of the Cats resolution When the user does not define the beamstep the default is for it to be the same as the Cats resolution All shapes will be forced by Cats to be an integer number of beamsteps in height and width The shapes are still placed on the nearest resolution grid point The split and bury technique will be used on shapes in th
259. is command finds the optimum fine focus and stigmation settings by scanning the beam across the mark edges and measuring mark edge slope for a range of settings The maximum mark edge slope corresponds to the best focus By measuring 4 edges at 0 45 90 and 135 the stigmator settings are also found When the adjustment has finished the results and some statistics are typed on screen A description of how the statistics are generated is in the chapter Calibration The parameters filter lines and points to be supplied with the command set the amount of signal averaging The optimum amount of averaging for accurate results is that which results in a repeatability within about 0 0015 fine focus units The optimum amount of averaging will vary with beam current beam Part Number 878275 Vectorbeam Operator Manual 17 4 3 17 4 4 Page 168 energy and mark contrast This automatic fine focus and stigmation adjustment is carried out when jobcal or fullcal are run If the fine focus setting after automatic adjustment is outside the range described in the Section Correct fine focus value on datum for conjugate blanking then lens 2 must be adjusted Debugging problems with automatic focus stigmation adjustment To check and setup the correct operation and parameters for the focus and stig routines login to the DCP Type gt G_Focus 3 When an automatic focus and stigmation adjustment is run a plot of Focus score v
260. is not available LNKTIMOUT Link timeout for receive Ethernet packet LNKHSTNCON Host is not connected LNKWNGHST Subsystem is already connected to another host UNSOLINP Unsolicited input on Ethernet link PACKORERR Ethernet packet order error FNCNTAVAIL Function is not implemented EHTRAMPUP EHT ramping up EHTRAMPDWN EHT ramping down EHTDISABLED EHT disabled EHTOFF EHT off EHTON EHT on EHTFAULT EHT hardware fault FILRAMPON Filament ramping on FILRAMPUP Filament ramping up FILRAMPDWN Filament ramping down FILDISABLED Filament disabled FILOFF Filament off FILBLOWN Filament blown FILFAULT Filament hardware fault FLTCNVVME VAX to VME floating point conversion number too small FLTCNVVAX VME to VAX floating point conversion number too big FPE Floating point arithmetic error INVALIDRES Invalid communication response ILLPRONUM Illegal profile number ILLMAXMIN Illegal max and min values ILLVALUE Illegal parameter value entered INVALIDFMT The command was not in a valid format PARAMRNG Parameter passed was out of range HWFLT Hardware fault RBOUTTOL Readback out of tolerance HSTCOMM Host communication not initialised UNRECCMND Sub system received but did not recognise command UABLTODB Unable to open machine database file UABLTOUPD Unable to update machine database file DBFILEHEAD Bad database file header BADDBS
261. isable in this state as a key symbol If this window is iconised then double click with the left mouse button to bring it up on the screen Click on the Session menu heading and then click on Work in Progress Typically there will be two or more active tasks in the window for example e Window Manager this should always be present e Calendar this will depend on the users selection but is normally included Part Number 878275 Vectorbeam Operator Manual Page 64 8 8 2 8 9 8 9 1 8 9 1 1 8 10 Then if the user has opened further applications e g Calculator or CVIEW then these will also appear in this window If there is no CVIEW window or icon on the screen but the CVIEW user has quit the program after an X Toolkit warning then CVIEW remains in the Work in Progress box with the status Done This means it is still using system resources This situation arises when the user clicks on Close in the X Toolkit To remove a task in the Work in Progress window with the Done status use the following operations 1 Open the Work in Progress window 2 Click once on the done task which will highlight the task line 3 The boxes at the bottom of the screen now change status 4 Click on the Remove Task button Recommendation It is advisable to regularly check the Work in Progress window and remove any tasks with the done status This applies to Emma as this task is left over from starting Emma and has no
262. ive to the fine focus coil is shown schematically in figure 17 1 and is the sum of 1 The fine focus setting shown in Emma 2 The value required to compensate for any height difference between the current position and the reference height recorded at the last calibration of the main field Figure 17 1 Schematic of fine focus control Notice that the fine focus setting shown in Emma never changes after afld The fine focus setting is only changed when an automatic focus or stigmation adjustment is carried out Unlike DLN3 read DFOC read does not show the actual fine focus current therefore it is not possible to observe the above summation Part Number 878275 Vectorbeam Operator Manual Page 170 If it is necessary to expose a focus matrix all that is required is to issue the command qset focus before each exposure The fine focus value should be varied over a suitable range around the value found at the calibration mark 17 5 17 6 17 6 1 17 6 2 Beam current measurement To measure the beam current move the stage to the Faraday cup select Beam on in the Status Window and enter in the Job Control Window VB_OPER gt QDISPLAY CURRENT The beam current will be displayed in the Job Control Window A program has been written to carry out these operations Enter in the Job Control Window VB_OPER gt MEA C Beam diameter measurement Background The beam diameter full width half
263. ize BEF fieldsize BEF fieldsize BEF fieldsize Correction errors Scale nm Scale nm Rotation nm Rotation nm mean sigma mean sigma 10 10 10 10 18 4 7 Stigmation calibration The results and statistics of the stigmation calibration are compiled as follows 1 When the autostig completes successfully the final stig values are entered into the stig mean value entry in the returned variable The last four focus values returned from the 45degree separated scans are averaged to give the final focus Part Number 878275 Vectorbeam Operator Manual Page 181 value which is also entered into the mean focus entry in the returned variable The sigma of the four scans is entered into the sigma focus entry The remaining sigma entries are cleared to 0 0 The only way to calculate a sigma for the stigmation values would be to perform the autostig more than once which isn t practical 2 If stig statistics have been requested the DCP performs another five sets of four focus scans with 45degrees separation The mean axial and diagonal errors are calculated from the differences between the orthoganol focus scans likewise with the sigmas The mean and sigma of focus error is calculated from the difference between the mean focus value and the result of each focus scan 3 The final accuracy of the stig calibration is calculated using accuracy 1 0 mean_axial mean_diag 2 0 100 0 4 The stig and focus erro
264. ize for the 20 bit pattern generator The subfields span a square area known as the Maximum Blocksize whose size can be set by the operator and which is centred on the field centre The default size of the Maximum Blocksize is the Maximum Fieldsize but the Maximum Blocksize can be set to be smaller in order to adjust the subfield size The Maximum Blocksize must be larger of either the block height or width of any pattern files to be exposed If the Maximum Blocksize is less than the default the total number of bits spanned by all the subfields will be less than 2 The number of subfields that span the Maximum Blocksize can be set by the operator The maximum number is 64 and this is the default The smallest number of subfields that can be set is limited by the maximum subfield deflection and the Maximum Blocksize The maximum number of bits in a subfield is 2 16384 and with 64 subfields and a maximum blocksize equal to the maximum fieldsize all bits are used The command to set for example 63 subfields with a Maximum Blocksize of 516 096 um is VB OPER qset block subfields 63 maximum 0 516096 The blocksize must be within the limits of the hardware The range of physical blocksizes is given in the Sales Specification for the particular machine Choosing the subfields 16 bit pattern generator The operator does not have to define any parameters relating to the subfields 18 bit pattern generator The operator must define the
265. k frequency The total pattern area can be found from the converter or alternatively the total number of exposed exels is typed up after exposure Mainfield deflector settling time Mainfield deflector settling time the sum for each block of the pattern of Part Number 878275 Vectorbeam Operator Manual 24 1 3 24 1 4 24 1 5 24 1 6 24 1 7 Page 252 Number of adjacent occupied subfields maximum 64 64 short delay Number of successive occupied subfields with 2 to 6 subfields separation medium delay Number of successive occupied subfields with gt 6 subfields separation long delay The information about the number of short medium and long subfield jumps is not easily obtained An estimate of the average number of subfield changes per block can be made based on the pattern density The settling times may be displayed and set as described in the section Displaying and setting the settling times Subfield deflector settling time Subfield deflector settling time Number of shapes in pattern after fracturing into subfields Subfield settling delay Shape to shape delay The total number of shapes after subfield fracturing is typed up in the job control window after exposure The settling times may be displayed and set as described in the section Displaying and setting the settling times Shape synchronisation time Shape synchronisation time Number of shapes in pattern after fractur
266. k centre location The overall XY mark position is found by adding the XY stage position BACKSCATTERED ELECTRONS SS Figure 11 1 Collector geometry Several forms of mark geometry and image formation are used which generally can be made compatible with normal semiconductor processing for example square metal areas viewed with the backscattered detector Marks are normally pre defined lithographically in direct write applications as part of the previous or earlier layers whilst for calibration each substrate holder is provided with a small array of target marks on the holder and at the height of the nominal writing plane The marks can be fabricated to yield two forms of backscattered electron contrast 1 Material elemental atomic number contrast 2 Topographical contrast Part Number 878275 Vectorbeam Operator Manual Page 88 Material contrast occurs due to the difference in backscattered electron coefficients between different elemental materials The higher the element is in the periodic table the larger its backscatter coefficient will be It is for this reason that the high contrast holder mounted target marks are fabricated in thin gold on a silicon wafer chip On process wafers it is essential that compatible materials are used which in the case of silicon cannot be gold but high Z materials from the refractory elements such as tungsten can be considered On GaAs wafers gold is typically used for th
267. k positions outside stage limits FILMAPSIZ File mapping error MAPSIZERR Map size error OUTVBNL Outside virtual block limits BLKTRNERR Block transfer failed MRKDBFULL Mark definition data base full NOMRKFND No mark found in the data base TRAPDEFLIM Trapezium deflection limit exceeded MAINDEFLIM Main field deflection limit exceeded MRKDEFFAIL Mark definition failed LOCFAIL Locate mark failed CRTDWLERR Correction down load error EHTDEFERR EHT not correctly defined on the pattern generator FLDSIZERR Field sizing error LHSPOOR Laser height sensor reading poor LHSATLIM Laser height sensor brightness at limit LHSOVRRNG Laser height sensor over range LHSOFF Laser height sensor off LHSFAULT Laser height sensor fault LHSPSUFAULT Laser height sensor PSU fault LHSSYSFAULT Laser height sensor SYS fault LHSCOMFAULT Laser height sensor COM fault LHSTBL Unable to set the Laser Height sensor table MRKUNKDEF Unknown deflection option MRKUNKDIR Unknown raster scan direction option MRKUNKLOC Unknown locate option MRKUNKPLT Unknown plot option MRKZEXSIZ Zero exel size MRKMFLDP Main field position out of range MRKTFLDP Trap field position out of range MRKWRITFLD Trap field extends beyond writing field MRKWIDPRM Mark width parameter out of range MRKHGTPRM Mark height parameter out of range MRKMSRWIDTH Mark measurement wi
268. ke the position measurement more noisy Increasing the rise time above the sum a b will ensure the best repeatability However the measurement length MLEN must be set larger than the rise time and so an excessively large rise time will increasing the locate time unnecessarily The rise time can be found by locating a mark with the adc qualifier and using the adc plot cpr PVWave program to examine the ADC data see Section Diagnostic output of the mark locate function or Acceptance Test and Operator Jobfiles manual 892777 Figure 11 4 shows an example plot from this program and the rise time is about 0 06 um Mark Locate ADC Plot X axis scan 1 coarse scan 250 ET TT TT TTTTTTTTTTTTTTTTTTT TTTTTTTTTTTTTTTTTTTTT ptr Sa pow garn Noise RMS 2 66253 ADC units Start of edge at 2 52 um 200 4 150 Y 100 4 End of edge at 2 46 d e Era drca cb p E c pac gen aea p c Dp cce s Ee parie 2 80 2 70 2 60 2 50 2 40 2 30 2 20 Video Irtensity ADC units Deflection X um File VESDAT ADC DUMP OB 54 13 MAY 2005 dat Figure 11 4 Example plot of mark locate ADC data 3 The type BRIGHT DARK should be set for the contrast obtained from the mark 4 The filter FILTER is typically set to 16 or 32 for low beam currents giving noisy signals but 8 for relatively high currents This parameter along with OVSAM and LINES is used to reduce the effects of noise on the detec
269. ks relative to the calibration grid origin should be as close as possible to the positions of the four corner marks of the grid on the calibration substrate This is to avoid any component of the stepper lens correction finding its way into the die mode 3 Alens map must always be explicitly assigned to a direct write mapping after it has been set up Setting up a direct write mode will automatically de assign any lens map that was assigned to it 4 When the lens map has been assigned all the area covered by the map can be exposed The following jobfile called STEPPER_LENS COM can be used together with the die by die direct alignment scheme described in the Chapter Direct write alignment by using the line VB SEQ WLVD_DW COM STEPPER_LENS COM in the layout parameters file OPEN APPEND SHARE POST EMMA MAILBOX EMMA Create position identifier for current centre of die in die mode SSPO DIECENTRE 0 0 Load stepper lens map 1 into die mode qmap lens 1 DIECENTRE MAP DIE XORIGIN 10 875 YORIGIN 10 875 Move to centre of cell in new coordinate system QMOVE POS CENTRE 0 0 Expose pattern QEXPOSE PAT Deassign stepper lens map QMAP LENS INIT 18 10 5 Further notes on using the stepper lens correction 1 The lens maps are conceptually not treated as transforms in their own right but rather as small additional corrections to a mapping Therefore if the current direct write m
270. l Lens Y Grid size too small order 1 is minimum Increase Y grid points LHSATLIM Laser height sensor brightness at limit Check that the reading is not being taken off the substrate that the selected height table is correct that the height calibration has been done on the substrate LHSCOMFAULT Laser height sensor COM fault LHSFAULT Laser height sensor fault Check that the reading is not being taken off the substrate that the selected height table is correct that the height calibration has been done on the substrate LHSOFF Laser height sensor off Switch on LHSOVRRNG Laser height sensor over range Substrate not present or incorrect stage position or incorrect height meter table specified or height table calibration not valid Rectify and retry LHSPOOR Laser height sensor reading poor Substrate not present or incorrect stage position or incorrect height meter table specified or height table calibration not valid Rectify and retry LHSPSUFAULT Laser height sensor PSU fault LHSSYSFAULT Laser height sensor SYS fault LHSTBL Unable to set the Laser Height sensor table LIMITEDACC Beam diameter of limited accuracy LMP2FEWDATA Can t fit polynomial to data too few good points LMPBADDWMAP The locates at all 4 corners must succeed but didn t LMPBADORDER Bad lens map order found when calculating polynomial LMPCOLHOLE Data has consecutive failed locates looking down columns LM
271. l systems e Main deflection double lever deflection e Subfield deflection single lever deflection with HR lens and double lever deflection with UHR lens This unit is sometimes referred to as the trapezium or trap deflection It is possible to adjust automatically calibrate the size of the main and subfield deflection in order to optimise the resolution of the main and subfield deflection system according to different application requirements The range of the subfield deflector is calibrated to be 1 64 of the main field range The main field pattern is therefore divided into an array of 64 x 64 subfields this is performed in real time by the pattern generator hardware Part Number 878275 Vectorbeam Operator Manual 2 4 Page 21 During pattern writing pattern information is read from the Operator Terminal disk and processed by the pattern generator unit The pattern generator generates the analogue drive signals which drive the main and subfield deflectors Correction systems compensate for lens and deflection errors and assure compatibility of the main and subfield deflection with the stage co ordinate system The main deflection system positions the beam to the centre of a subfield and each subfield is addressed sequentially as the pattern is written Main deflection is specified by the main deflection addresses provided by the pattern generator In order to meet the accuracy requirements of the system this also generates
272. l to ithography manuals Vistec semi com Giving full details of the changes required Vistec Lithography Ltd PO Box 87 515 Coldhams Lane Cambridge CB1 3XE UK 2 Tel 44 0 1223 411123 VU S iF e C Fax 44 0 1223 211310 PARE r hodraphy Lithography Table of Contents Preface and safety information e sssocccssooccesssooccssssocecssssooecsssoseceessosesssse L 1 Part number 878275 Vectorbeam Series systems esssossssesssooccesssooecssosccecssooecesssosecesssesee D 1 1 1 2 1 3 1 4 1 5 1 6 1 7 1 8 ieee 1 10 Vectorbeam system OVeP VIEW raso idi rei di e greek Veo d eta dg ERE M e ERR 5 pir MR OT D E DEOR 6 Wes TII EU 6 12 2 Substrate load Syste Messiera ee tei nup Scr EE Rad De Pedo iu RE 7 Ekeroos GOS Sonst di tcu ad at c d mta ead 8 1 3 1 Control electromes TACK Wi cess seis elis cssc eee i edk yr taco ddha cans diaaceudecha vh kx yR wk un 8 13 2 Control electronics TACK 2 iiec ta eeu ced xi eee ree Del eexr esa ches expl De imr 8 dis we rsinengesbe dizi TT TE EN 9 kae XOSDIDGE A se seseielis ion pee lir eio res ee toes dg eresuvtistteri aes S letus Et 9 WA SOSBIDebBuneseecion pee iro e ries edie dio ieri dpeteeuefle ene es E Me PEE 9 SAT CH et sie EAT CADE uaque qnie Eoo isla dugavslt cid aeo da ble Quern Lora 10 Power SUPPO i si Perna ceases rein iie EE eer Proceed bera a eb f ek adag 11 Operaio BO SONS resinen ai 11 Temperature control water Dallis os noU
273. le aligning the column ONLY USING THE TILTS do not touch the shifts at this time A maximum beam current of between 300nA 500nA should be possible It is possible at this point to move to the Faraday cup position identifier FC and measure the beam current to find the optimum TILT positions that give maximum current Experience has shown that a lens 2 value of 0 3 gives a fairly focused beam at the maximum current so start by setting this and then adjusting lens 1 and the tilts Once happy the tilts are set lower the beam current to around 5nA Reduce the SEM magnification so the image is small on the monitor by using about X500 Slowly reduce the lens 2 current to 0 while keeping the image on the SEM monitor The image will become blurred This time ONLY USING THE SHIFTS do not touch the tilts adjust the sliders from minimum to maximum and note the slider values at which point the image starts to disappear This is the point at which the beam clips the lens 2 aperture Set the slider to the centre of these clipped points this should be the centre of the aperture Do this for both X and Y SHIFT sliders The image should now be in the centre of the monitor Increase the lens 2 until the image again becomes focused The image SHOULD NOT MOVE If it does the SHIFT sliders have not been set correctly and will need adjusting until image movement is minimal less than 10nm in X and Y Now the SHIFTS have been set DO NOT TOUCH THEM
274. ler The following steps which are described in more detail below are required 1 Assign doses to shapes using either a proximity correction calculation or direct assignment Both these methods are available in both Cats and Caprox 2 Choose either the clockfile or the pattern as the method to transfer the dose distribution to the machine and convert the pattern accordingly 3 Copy the pattern to the machine and expose as normal A clockfile is required if the relative doses are not included in the pattern file Dose controller operation with proximity corrected patterns Any one of the 65536 frequencies supplied by the dose controller will automatically be selected for each shape during pattern exposure according to the clock number in the pattern file associated with the shape The dose is thus changed from shape to shape This facility enables compensations to be applied to correct the proximity effect caused by electron scattering within the resist and substrate The large number 65536 of available frequencies spread between the maximum and minimum frequencies means that the dose interval between successive frequencies is small compared with the variation required for proximity correction giving a high accuracy capability Proximity correction The proximity effect degrades the fidelity of the lithography The effect has been studied by many people and the literature should be consulted for details The effect can be compensat
275. les above meet this requirement but 48 subfields with a maximum blocksize of 0 8192 mm and a maximum fieldsize of 0 8192 mm would give 5461 33 bits per subfield which is not allowed In summary the following should be taken into account when choosing the subfields and the maximum blocksize 1 There must be an integer number of bits in each subfield 2 The maximum subfield deflection must not be exceeded 3 The maximum blocksize must be larger or equal to the physical blocksize required by the pattern 4 A larger number of subfields e g 64 may give higher accuracy but will be slower due to the larger number of subfield changes 5 The periodicity of the subfields may be matched to periodic sensitive areas of the pattern to avoid subfield stitch errors occurring in these areas 20 bit pattern generator The operator must define the number of subfields and the maximum blocksize The number of subfields and the maximum blocksize are the same in both X and Y The number of the subfields that span the maximum blocksize must be defined by the operator The size of a subfield bit is always set to be the same size as a mainfield bit The maximum number of subfields is 64 and the default number is also 64 The minimum number of subfields possible is limited by the maximum subfield size and the maximum blocksize that must be spanned The maximum subfield size has two limits 3 A bits limit of 2 or 16384 bits 4 Asize limit of 20 um
276. libration mark is usually outside the measured array but has by definition zero magnetic map correction Correction cells in an area around the calibration mark are zeroed by using a command qualifier such as VB_OPER gt QCALIBRATE STAGE DATUM RECT 0 0 125 12 9 etc 3 See the Vectorbeam Command Set Manual 878274 for details of these qualifiers The area should include the entire datum plate as it is likely that at a later time a fresh calibration mark on the datum plate will be chosen when the first has become contaminated The calibration mark can be moved in such a way without recalibrating the magnetic map because the distance between the marks is small and the difference in the corrections is negligible 4 Separate magnetic maps are required for using holders on the same machine which have the datum plate in different positions This is important when using 8 inch wafer holders on the VB6 and NIST type X ray holders on the VB5 These holders have datum plates in different positions to that used by all other holders including the standard autostitch plate mask holders normally used for calibrating the magnetic map The difference in the calibration mark positions is large several cm and the difference in the corrections may be significant In this case the user must generate a suitable autostitch target substrate An appropriate substrate an 8 inch wafer or X ray mask must be patterned such that an array of marks with suit
277. lly only one holder has a knife edge loaded It does however allow spot sizes below 10 nm to be measured A fresh knife edge can be made by cleaving a piece of GaAs wafer so as to give a smooth face normal to the wafer surface Care must be taken to ensure the face is smooth has no damage to the top edge and has little debris on the edge The piece must be small enough to mount on the knife edge holder The edge is nominally in the same plane as the substrate surface in the holder but because the knife edge is not front face referenced the edge may be a few hundred microns above or below the plane 1 Load holder containing the knife edge on the stage 2 With a 16 bit pattern generator in order to be able to scan the beam in small steps to measure the smallest spot sizes more accurately the pattern generator resolution should be set to a small value Set the maximum fieldsize to e g 0 16384 mm VB OPER FLD 0 16384 3 For the 18 bit and 20 bit pattern generators this is not necessary Part Number 878275 Vectorbeam Operator Manual 17 6 3 Page 172 Zero the field corrections VB_OPER gt QCAL MAIN DIST INIT LOAD VB_OPER gt QCAL BEF DIST INIT LOAD VB_OPER gt QCAL STIG DIST INIT LOAD Calibrate the main field and BEF for this fieldsize VB_OPER gt QCAL MAIN FM NOALIGN COVER F TRNLMN VB_COVER ITER 2 DIAG VB_OPER gt QCAL BEF FM NOALIGN RANGE 15 ITER 2 DIAG Move the stage to the knife edge posi
278. lowing commands for X and Y respectively VB OPER qset register 3 31 value in microns VB OPER qset register 3 34 value in microns The effect of a height offset on scale can be calculated from the usual rotation for height dependence of about 16 nm um at the edge of a 1 mm field These offsets must be removed before mainfield calibration as otherwise their effect will be calibrated out and then re applied This is done automatically by jobcal Subfield scaling offsets No subfield scaling offsets can be applied 18 bit and 20 bit machines Calibration offsets for the mainfield BEF subfield focus and stigmation are supported by Emma commands making them easy to enter control apply at the correct time and monitor These offsets are e Enabled only if they have been switched on using the qset corrections on offsets command e Automatically never applied during any of the calibrations e Not applied when a qadjust field command is executed unless the qualifier offset is supplied e Applied by default during pattern exposure Pattern exposure is done effectively using the qadjust field offset command at each stage position To display all set calibration offsets type VB OPER qdisplay offsets Part Number 878275 Vectorbeam Operator Manual 18 9 2 1 18 9 2 2 18 10 Page 187 Mainfield The mainfield scale rotation and keystone calibrations can be offset For example to define a 10
279. lts Part Number 878275 Vectorbeam Operator Manual 19 19 1 19 1 1 Page 193 Exposing a substrate Pattern data preparation The pattern to be exposed contains shapes that are defined during the device design stage The position and size of the shapes are thus usually defined with no regard to the lithography tool The shapes are defined in some file format such as Calma GDSII or SPD or CTXT Pattern data preparation is the process of transforming the design tool output file format into a file format that the machine can expose There are two Vistec proprietary formats for Vectorbeams FRE which was designed for machines with a 16 bit pattern generator and VEP which was designed for machines with 18 bit and 20 bit pattern generators The 18 bit and 20 bit pattern generators can expose the FRE file format in addition to the usual VEP format for backwards compatibility The conversion process also contains essential elements such as taking into account the fracturing needed to expose the whole pattern using a field by field exposure sequence and breaking up the polygon description of the pattern supplied by the CAD system into small elementary trapezia In addition the process must place the vertices of all shapes on the pattern generator resolution grid Parameters such as the pattern generator resolution grid and fieldsize are chosen in order to obtain the desired best compromise between speed and fidelity given various limitati
280. lude in each jobfile the command SUPER at the end so that when the jobfile has finished operation can be continued remotely If the oper status is assigned to a DECTerm which cannot be accessed it is possible to get rid of that DECTerm make sure that you do not interfere with someone else s machine operation by typing VB SUPER stop vb oper Part Number 878275 Vectorbeam Operator Manual 21 3 21 4 21 5 21 5 1 21 5 2 Page 244 The oper status can then be obtained Type VB OPERsstart in order to connect to the subsystems and set up the environment for the new terminal The machine can now be operated in interactive mode Login without using second license In order not to use the second login license which prevents the system manager logging remotely the menu item called New Login in the Applications menu of the Session Manager should used This should be selected whenever it is required to login to another account rather than SET HOST 0 The New Login facility uses the command CREATE TERM DETACH NOLOGIN Which creates a new login WITHOUT using up the System Managers license NOTE The command above and the New Login menu can only be used when sitting at the machine They cannot be used remotely Batch queue operation In order to avoid having to be connected for the duration of the job jobs can be submitted to the batch queue Type VB OPER submit jobfilename com The batch qu
281. lution steps from the origin The design grid is the maximum resolution step size for the grid in order that 100 of the shape co ordinates lie on the grid The ongrid command in CATS can be used to find the design grid 2 Obtain a maximum fieldsize larger than the required blocksize See the examples section later in this chapter for more information Maximum fieldsize pattern generator resolution ranges The official guaranteed range of maximum fieldsize and pattern generator resolution is given in the Sales Specification for the particular machine The tables below show typical values 16 bit pattern generator Part Number 878275 Vectorbeam Operator Manual Page 200 Max fieldsize PG 20 kV 50 kV 100 kV resolution Minimum 0 131072 mm 2 nm 0 131072 mm 2 nm 0 181072 mm 2 nm mm 2nm 1 024 mm 15 625 0 8192 mm 12 5 nm 0 589824 mm 9 nm nm Depending on the hardware it may be possible to use a larger range as shown below but this is not guaranteed Max fieldsize PG 20 kV 50 kV 100 kV resolution Minimum 0 131072 mm 2 nm 0 131072 mm 2 nm 0 181072 mm 2 nm mm 2 nm 1 150000 mm 17 5 0 930000 mm 14 1 0 65536 mm 10 nm nm nm In order to obtain the minimum value of maximum fieldsize it may be necessary to change the position of links on the X and Y mainfield DAC cards Link 7 and link 9 must be fitted and the following needs to be added to the emma ctrl dcp config vw file G bS
282. ly as long as VB Field is set correctly For example using format VB20 a Resolution of 0 015625 will give VB Resolutionz0 015625 and VB VRU 1 but a Resolution of 0 016 will give VB Resolutionz 0 008 and VB VRU 2 For some patterns it may be desirable to use a lower value of VB Resolution for higher accuracy and resolution and the value can be changed using the command VB Resolution value The Resolution does not need to be a power of 2 multiplied by the VB Resolution but rather any integer multiple VB VRU The VB VRU parameter is only used for the VB20 VB50 and VB100 formats and Part Number 878275 Vectorbeam Operator Manual 19 1 1 1 10 19 1 2 19 1 2 1 19 1 2 2 Page 198 is calculated as follows VB VRU Beamstep VB Resolution There is no need to set the VB VRU parameter directly although for some patterns it may be desirable to use a higher value of VB VRU for higher accuracy and resolution and the value can be changed using the command VB VRU lt value gt The VB VRU can take the following values e 16 bit machines 1 2 4 8 16 32 e 18 bit machines 1 2 4 8 16 32 64 128 e 20 bit machines 1 2 4 8 16 32 64 128 256 512 The Vectorbeam pattern generator can expose the pattern with any VRU set on the Vectorbeam however best results are obtained when the VRU on the Vectorbeam is set to VB VRU Maximum number of shapes in a block A feature was introduced in version 14 02 of Cats so
283. m fieldsize for a 16 bit pattern generator will be only 65 536 um This would mean a relatively large number of stage moves and an impractically large job time Therefore it is usual to set the pattern generator resolution grid to be larger and accept some grid snapping By inspection of the pattern using for example the ongrid command in the CATS converter a grid can be found in the range 5 to 12 5 nm for which the amount of grid snapping will be relatively small This may well be some integer such as 5 or 10 nm In this example 5 nm will be assumed to be optimal The minimum dimension of 180 nm with B 4 Equation 19 14 implies an exposure grid of 45 nm If the technique of negative biasing is applied then biasing the pattern by 20 nm gives a minimum dimension of 160 nm and using a VRU of 8 means that the exposure grid size is 40 nm The minimum dimension is an exact multiple of the beamstep size which results in these features being the correct size Other larger features may be not be an exact multiple of the beamstep size so either the pattern generator grid snapping mode should be set to split and bury or the beamstep parameter should be defined at conversion time Drift removal using ontime The mechanical and electrical components of the VB are affected by changing temperature and even 1 C change can have a large effect This affects the pattern placement accuracy for exposures that require several hours The ontime function can b
284. mage panel Figure 7 4 and for bright marks the background should be at about 10 of the range above zero or higher and the mark level should be about 1096 of the range below the maximum or lower For dark marks the mark level will low and the background will be high Marks with low inherent contrast may not produce the 1096 and 9096 video levels for any gain and backoff settings The video signal may saturate at lower backoff levels Saturation is characterised by the lack of noise in the video signal even with the minimum amount of signal averaging Set the gain and backoff to give the maximum contrast without saturation When executing a jobcal or a fullcal both the gain and the backoff are set automatically to give a contrast level of about 70 on the calibration mark In other cases and especially when aligning to marks on substrates as required for direct write or measuring stitch acceptance tests the gain level can be adjusted automatically but the backoff level needs to adjusted manually Manual gain and backoff adjustment Move the gain and backoff sliders in the Set Video Level panel Figure 7 9 or define values directly in the command VB OPER qset video gain 0 3 backoff 0 8 Automatic gain and backoff adjustment HR machines There are two functions that automatically adjust the gain called peak and mean There is no function to automatically adjust the backoff The video gain can be switched into automatic adjustment mo
285. mallFldSzInUse 1 19 1 4 1 2 18 bit pattern generator standard fieldsize with HR or UHR lens Max fieldsize PG 20 kV 50 kV 100 kV resolution Minimum 0 131072 mm 0 5 0 131072 mm 0 5 0 131072 mm 0 5 nm nm nm Maximum 1 024 mm 3 90625 0 8192 mm 3 125 0 589824 mm 2 25 nm nm nm Depending on the hardware it may be possible to use a larger range as shown below but this is not guaranteed Max fieldsize PG 20 kV 50 kV 100 kV resolution Minimum 0 131072 mm 0 5 0 131072 mm 0 5 0 131072 mm 0 5 nm nm nm 1 150000 mm 4 387 0 930000 mm 3 548 0 65536 mm 2 5 nm nm nm 19 1 4 1 3 18 bit pattern generator large fieldsize with HR lens Part Number 878275 Vectorbeam Operator Manual Page 201 Max fieldsize PG 20 kV 50 kV 100 kV resolution Minimum 0 16384 mm 0 625 0 16384 mm 0 625 0 131072 mm 0 5 nm nm nm 1 024 mm 3 90625 1 310720 mm 5 nm 0 589824 mm 2 25 nm nm Depending on the hardware it may be possible to use a larger range as shown below but this is not guaranteed Max fieldsize PG 20 kV 50 kV 100 kV resolution Minimum 0 16384 mm 0 625 0 16384 mm 0 625 0 131072 mm 0 5 nm nm nm 1 310720 mm 5 nm 1 310720 mm 5 nm 0 8192 mm 3 125 nm 19 1 4 1 4 18 bit pattern generator large fieldsize with UHR lens Max fieldsize PG 20 kV 50 kV 100 kV resolution Minimum 0 16384 mm 0 625 0 16384 mm 0 625 0 131072 mm 0 5 nm nm nm 1 310720 mm 5 nm 1 310720 mm 5
286. modem to an internet service provider Checking status remotely The Emma interface program itself does not have any command that provides current status information to a user logged in remotely however it is possible to read the contents of the group logicals Many group logicals are set by Emma as a way of outputting data or showing the status They can all be shown by typing vb super show logical vb Some of these logicals will always show the current status but some will only be updated when the appropriate qdisplay command is issued by the jobfile For example vb pat file is updated with the selected pattern file whenever the qselect pattern command is issued but vb dcpo abs x is only updated with the current stage position when the command qdisplay position command is issued All the logicals beginning with vb lay are set by the layout program In particular vb lay cur cell x and vb lay cur cell y are updated by layout on starting each new cell and so it is possible to see remotely which cell the machine is currently working on Commands can be included in jobfiles to set new logicals at various points which can then read remotely For example define group user my logical Some string or other Obtaining the oper control prompt If possible the command super should be entered at the operator terminal to allow the remote user to obtain the oper prompt by typing VB SUPER OPER It might be useful to inc
287. molecular pumps until the required vacuum is obtained Both pumps are switched ON and OFF together When the turbo molecular pump reaches its maximum speed the pre vacuum pump functions as a backing pump High capacity turbo molecular pumps are provided for the airlock and the chamber lon getter pumps are used for pumping between pressures of about 1 x 10 3 Pascal 0 75 x 10 Torr and about 1 x 10 Pascal 0 75 x 10 Torr These pumps are used on the gun emission chamber The reason for using separate volumes is that total evacuation of the system requires much longer time than the evacuation of individual volumes Therefore it is better to vent only that chamber to which access is required and the low pressure in the rest of the system is retained TFE column vacuum system Gun emission chamber as measured 1 x 10 Torr 2 uA with ion pumps 1 and 2 Mid column as measured with ion 1x 107 Torr pump 3 Lower column as measured with ion 4 x 10 Torr pump 4 A similar bake out procedure is used for TFE source installation and run up but the temperature used is limited to 180 C The procedure is somewhat more detailed due to the UHV requirements of the TFE type sources VB5 vacuum system The VB5 vacuum system is divided into three main volumes 1 The airlock 2 The substrate chamber and lower electron column 3 The column Each is separated from the other by pneumatically operated valves These volumes may in general b
288. n Global cell 4 POS 999 999 for no stored global position These symbols are optional and the defaults are the expected positions calculated from the layout parameters They may be used to define more accurate X Y positions for the global marks than the expected positions calculated from the layout parameters Such positions can come from substrate measurements before exposure and allow more easily a batch of substrates to be exposed reliably with the same jobfile especially if automatic global alignment is used The search for the global marks will be carried out at these positions instead of the normal expected positions It is not necessary to define these positions and they may all be set to 999 DirectWriteStrategy DW2 Defines the method used DW1 means that no die by die alignment is carried out but just the previous global alignment is used DW2 means die by die alignment is carried out and all four marks of each die are located before exposing each die When the machine only finds 0 1 or 2 alignment marks a warning message is generated and the die is not processed DW3 means die by die alignment is carried out by measuring only those marks which have not already been measured If neighbouring dies use the same alignment marks then the mark is only located once the first time it is encountered and then stored in memory When the machine only finds 0 1 or 2 alignment marks a warning message is generated and the die is no
289. n file different to current value Set current value to that saved in file if known and then retry Alternatively use the override qualifier but note that all other mismatches will be overriden Part Number 878275 Vectorbeam Operator Manual Page 274 DBMFSENS Main sensitivity in file different to current value Set current value to that saved in file if known and then retry Alternatively use the override qualifier but note that all other mismatches will be overriden DBMVPIVOT Main pivot point in file different to current value Set current value to that saved in file if known and then retry Alternatively use the override qualifier but note that all other mismatches will be overriden DBPARSE Failed to parse file specification DBSEARCH Could not decode file specification or no file DBTFSENS Trap sensitivity in file different to current value Set current value to that saved in file if known and then retry Alternatively use the override qualifier but note that all other mismatches will be overriden DBTVPIVOT Trap pivot point in file different to current value Set current value to that saved in file if known and then retry Alternatively use the override qualifier but note that all other mismatches will be overriden DCBADDOSE Illegal dose value zero or negative Set valid dose DCBADEXEL Illegal exel size zero or negative DCBADFREQ Frequency can t be set out of hardware range Set frequency less than 2
290. n the Section holder loading unloading in from airlock Bakeout button This button requires two presses before the command is accepted The use of the bakeout button is described in the Section bakeout Bakeout reset button The use of the bakeout reset button is described in the Section bakeout Reset button The reset button acknowledges any errors or warnings which may be shown by the vacuum control panel The buzzer is deactivated by pressing the button once The buzzer is deactivated by pressing the button twice PICS returns to the Part Number 878275 Vectorbeam Operator Manual Page 33 point where it was when the error occurred 3 8 PICS and Brooks power up and reboot actions The PICS microprocessor will boot automatically when powered on or if the red or black reset button on the front of the PICS processor card is pressed PICS should not normally reset as long as the machine is powered on as this causes PICS to set the vacuum system to a safe condition from which it may take several hours to reach operating conditions For instructions on starting the Brooks vacuum control see the VB6 PMC Operator Manual 893157 When PICS or Brooks has booted the safe condition is set up by taking the following actions 3 8 1 VB5 The valve V11 requires compressed air to remain closed V11 gun roughing isolation valve Undefined Undefined Penning gauges Note that V11 requires compressed air press
291. n the centre of the mark The intersection of the two lines is the position which is then returned to the user or calling routine This behaviour is illustrated in the following diagrams Although the illustration is for a square mark this can easily be adapted to fit an octagon by simply ensuring that the fine scans are confined to a region that does not encompass the diagonal edges Part Number 878275 Vectorbeam Operator Manual Page 98 FSLIM 8um MW tum MLEN 1 5um AAS P Fine Vertical Scans E l x Fine Horizontal Scans o W 4um Figure 11 7 Diagram showing fine search scans for rectangular marks with pit algorithm 11 9 11 9 1 Edge locate algorithm The edge locate algorithm is useful for marks which produce detector signals as shown on the left side of Figure 11 2 The edge locate algorithm is separated into a raster search and a fine search locate fine L 3 1 e Y anne locate raster 9 fine success il 4 fail Figure 11 8 Edge locate algorithm search flow diagram Coarse search The coarse search scans are in the form of a series of lines of length CSLIM The first coarse search scan is horizontal and centred on the expected position of the mark If two edges are not found another parallel scan is performed above the first If two edges are still not found another parallel scan is performed below the first The horizontal coarse scans expand o
292. nal limits of validity X Y positions of DW mapping at measurement lens map coefficients 5 Lens map names 12 2 4 12 Mark The mark partial database contains the mark locate parameters for mark definitions 1 to 20 1 The number of defined marks max 20 2 20 sets of mark locate parameters mark name mark description width height measurement width measurement height offset width offset height limb width dimension tolerance limb separation number of limbs rise time rise time tolerance contrast locate method mark polarity filter fine scan resolution number of lines number of samples cross coarse search length cross coarse search offset fine scan VRU mark coarse search limit mark fine search limit coarse scan resolution number of parallel scans coarse scan VRU mark measurement length Part Number 878275 Vectorbeam Operator Manual Page 111 3 9 sets of standard mark parameters X Y position mode mark name 12 2 4 13 Holders 12 3 Database saving The database may contain the total machine settings or selected parts Optional text max 80 characters may be specified such as the field size and the beam energy and is stored with the file VB OPER dfile save file spec dbase gt lt text gt total The facility to save and load selected parts of the database can be very useful It can be used to combine constant machine set ups such as the magnetic map with job specific set ups such as
293. nates a spot on surface of substrate 10 1 10 1 1 Page 73 Height meter The height meter subsystem measures the distance between the final lens and the surface of the substrate as shown in Figure 10 1 This distance is known as the height Height measurements are necessary for high accuracy lithography since changes in height due to bow or other substrate non flatness can introduce errors which affect the deflection field size and beam focus see Chapter Corrections During pattern exposure height meter measurements are taken after each stage movement and the required compensations are applied automatically to maintain high accuracy deflection see Chapter Corrections The subsystem comprises an infrared laser beam directed at the substrate surface and a CCD array detector which allows the position of the reflected beam to be measured An increase in the reading corresponds to an increase in the separation between the final lens and the substrate Final lens di EUN Video Substrate a Laser spot movement as substrate height changes Spot position on CCD moves as substrate height changes System converts change in spot position to change in height Figure 10 1 Diagram showing the operation of the heightmeter Height meter tables The height meter system must cope with the various reflectivities of different substrates When a height reading is taken the laser drive and the detector gain will be varied a
294. nd off axis calibrations are large then there is little range left for additional direct write mappings The exact limits vary from machine to machine as the subfield rotation values depend on the mechanical arrangement of the deflection coil assembly On D660 the X and the Y rotation corrections are balanced so rotating the coil will not Part Number 878275 Vectorbeam Operator Manual 24 8 2 24 8 3 24 8 3 1 24 9 Page 259 improve the situation 18 bit pattern generator The mainfield fine scaling DAC which is 18 bit has a total range of about 6 25 of the field size This means that for a field size of 0 32768 mm it has a range of about 20 5 um The subfield has separate scaling and rotation correction DACS The scaling DAC covers 3 125 and the rotation DAC covers 6 25 The DACS are both 14 bit As the subfield deflection DAC is 14 bits 1 Isb on the subfield scaling DAC has the same weighting as 2 Isb on the mainfield scaling DAC and the software sets the values appropriately This means that the smallest change of the maximum fieldsize is about 0 15 nm or 0 5 ppm 20 bit pattern generator The mainfield fine scaling DAC which is 18 bit has a total range of about 6 25 of the field size This means that for a field size of 0 32768 mm it has a range of about 20 5 um The subfield has separate scaling and rotation correction DACS The scaling DAC covers 3 125 and the rotation DAC cover
295. ne format this limits the number of exels which can be defined in a logical block to 2 6 1 The maximum logical blocksize is limited to 25 32 to avoid possible stitching problems on the 16 bit pattern generator if the blocksize were not an integer number of beamsteps As 4 x 4 logical blocks are combined into a physical block the physical blocksize is therefore limited to 2 128 20 bit pattern generator additional details on blocksize VEP file The largest blocksize that can be set in the converter for a VEP file is the Maximum Fieldsize largest blocksize in um Max Fieldsize Equation 19 10 Expressed in terms of exels largest blocksize in exels 2 1048576 Equation 19 11 FRE file When the 20 bit pattern generator exposes a FRE file a 20 bit physical block is produced by combining 16 x 16 16 bit logical blocks The physical blocksize is the size of the rectangle centred on the centre of the deflection field within which the beam may deflected during exposure see Figure 19 3 The logical blocks correspond to the blocks of the standard 16 bit pattern file Emma takes care of combining the appropriate blocks at exposure time Logical blocksize Physical blocksize Maximum fieldsize Figure 19 3 Diagram showing the arrangement of logical blocks and the physical block The logical blocksize is defined in the pattern converter and contained in each Part Number 878275 Vectorbeam Operator Manual 19
296. ng qset reg command check parameters Reboot PG 0x400b0506 Illegal data type If using qset reg command check parameters Reboot PG 0x200b0600 Operator caused errors Command aborted by operator Press the continue button and then continue 0x400beeee Exceptions usually bus errors Task suspended because of exception Possible hardware error Reboot PG 0x400c0001 0x400c0002 0x400c0003 0x400c0004 0x400c0005 0x400c0006 Master Micro Command Interpreter Errors MupCl receive from qEtherlP queue failed Reboot PG MupCl send to SuP Manager queue failed Reboot PG MupCl receive from SuP Manager queue failed Reboot PG MupCl received a reply to the wrong command Reboot PG Receive from qErrors queue failed Reboot PG Slave doesn t exist Possible PG configuration file error Reboot PG 0x400c0007 0x400c0008 0x400c0009 0x400c000a 0x400c000b 0x300c000c 0x400c000d Wrong no of bytes in long message Reboot PG MupCl send to qPatternIP queue failed MupCl receive from qPatternOP queue failed Reboot PG MupCl receive from qLHS Root queue failed Reboot PG MupCl send to qRootLHS queue failed Reboot PG Reply queue not empty before queueing command Reboot PG Register command received by wrong subsystem Reboot PG If using qset reg command check parameters Reboot PG 0x400c000e Pattern Generator s w configuration inconsistent Possible PG configuration file error Reboot PG
297. ngs In addition the user may save the same database or part of the database to a user defined filename and this file may be loaded later to restore the settings of the machine to a known state 12 1 Database structure There are two levels of partial databases 1 Top level partial databases combining individual parameters and bottom level partial databases 2 Bottom level partial databases which only contain individual parameters The arrangement is shown in Figure 12 1 Total Calib N ANN Coarse Fine Test Stepper lens Mark Spot General Pos Dist Stig Size Mag cal cal cal cal cal Figure 12 1 Diagram of database structure 12 2 Database parameters The individual parameters in the databases are listed in the following sections Note 1 Some of the parameters in the bottom level partial databases appear in other bottom level partial databases e g EHT in coarse test and fine 2 Many parameters in the total database are not in any partial database 12 2 1 Header for all databases All databases contain the following in the header 1 Type of database 2 Database version class X Y or V Part Number 878275 Vectorbeam Operator Manual Page 106 Database version number Filename at time of saving Date and time of savingMachine ID Emma version class Emma version number Total length Optional user text Any user defined comment can be saved with each partial database 10 Spare space
298. no temperature monitors VB6 with single or 10 holder airlock There are temperature monitors on the superplate on the X stage table on the Y stage table on the Z stage table in the crane pouch in the front pouch in the airlock on the ring around the bottom of the column on the x and y peltiers on the x and y motors and in the air space over the plinth VB6 with Brooks handling There are temperature monitors on the superplate on the X stage table on the Y stage table on the Z stage table on the ring around the bottom of the column on the x and y peltiers on the x and y motors in the air space over the plinth in the air circulating in the PGA electronics crate and in the air circulating in the CCU electronics crate Obtaining temperature readings Direct temperature measurement The temperatures from all the sensors may be displayed directly on screen using the command VB OPER DTMP or QDISP TEMPERATURE Background temperature monitoring The temperatures from all the sensors are read at regular intervals using the background temperature monitoring facility which also writes the data to a file This monitoring process is started by selecting Temperature monitoring in the applications menu of the windows manager or by running the jobfile Part Number 878275 Vectorbeam Operator Manual 15 2 3 15 2 3 1 15 2 3 2 15 3 15 3 1 15 3 1 1 15 3 1 2 Page 132 VB OPER QGEMMASCTRL COM TEMPMON COM In
299. nt current analogue reading unstable SAEHT Filament power analogue reading unstable SAEHT Analogue reading unstable Command paused by operator Command aborted by operator Conversion parameter out of range Standby value greater than maximum Minimum value greater than maximum Maximum value less than minimum CCU Extractor setting less than focus setting CCU Focus setting greater than extractor setting Value out of min max range wehnelt Value out of min max range emission Analogue readback error wehnelt Filament disabled Analogue readback error filament EHT logic PSU fault Power failure of gun 15 V supply EHT disabled EHT turned off Analogue readback error EHT Signal selected for AD conversion is unknown Wehnelt value out of min max range Filament value out of min max range EHT value out of min max range EHT value out of DAC range EHT hardware fault Power failure of aligner driver 15 V supply Failure of tilt current source s Failure of shift current source s Aligner driver interrupt status register hardware error CO power amplifier temp exceeds 80 C Failure of CO lens driver current source Power failure of CO lens driver 15 V supply CO lens driver board power failure C1 power amplifier temp exceeds 80 C Failure of C1 lens driver current source Vectorbeam Operator Manual Part Number 878275 0x40AACO26 0x40AAC027 0x40AAC028 0x40AACO29 0x40AACO2A 0x40AAC02B 0x40AACO2C 0x40AA
300. nual Page 49 Figure 7 2 The Emma Job Control Window Further terminals may be opened by selecting DECterm from the Session Manager Applications menu Figure 7 3 Figure 7 3 Supervisor DECTerm Part Number 878275 Vectorbeam Operator Manual Page 50 7 2 1 Set menu The panels called up under the Set menu can all be cancelled by selecting CANCEL at the bottom The Set Image panel Figure 7 4 may be used to set the SEM image processing functions Image help Monitor pO 000 000 pi 000 000 p2 000 000 Measure Scan None Normal Cross Hairs 4 Spot 0 Graticule EE RiRaster Graph al Undraw Exposure 4 Fab Display W Freeze TA 1 4 Rotation 2 4 Graph Period pp 1 2 3 z Graph Average Misc o 4 Detector L4 _ Front Face Bias Frame Average _ Data Zone Cancel Figure 7 4 Set image 7 2 1 1 Cross hairs In order to display two vertical and two horizontal lines on the SEM screen select the crosshairs option The position of the cross hairs can be changed by clicking with mouse on the lines which appear in this panel The left mouse button moves both cross hairs together and the right mouse button allows them to be moved independently Part Number 878275 Vectorbeam Operator Manual Page 51 The EO Service Setup panel Figure 7 5 should only be used by Vistec Engineers during the set up of the VB Figure 7 5 Set EO service set up The Set Stigmat
301. nuously updated Pgmonitor File Enquire Help Master Processor DCP Processor LW Processor Status UNKNOVAY UNKNOWN Fisp Status Table Sic Status Table Figure 7 19 Tool kit PGMonitor Part Number 878275 Vectorbeam Operator Manual Page 59 When the Joystick is selected the Joystick Control Panel is displayed Figure 7 20 or Figure 7 21 this enables the stage to be driven manually the boxes labelled Inc X and Inc Y and Mag may be edited and the size of the stage steps set By selecting with the mouse left button one of the bars not the actual slider bars the stage will move by the size of the step set in the axis selected The joystick is used in SEM scanning mode am X mm Y mm Mag OEE ee Current Position 79 623212 5 744368 x 14431 Saved Position 000 000000 000 000000 x 1000 000 Cancel Restore Save Relative Figure 7 20 Tool kit Joystick Panel Absolute Mode By selecting with the mouse left button the Relative icon a second Joystick Control Panel will be displayed instead Figure 7 21 By selecting with the mouse left button the slider bar on the side of the grid keeping the mouse button pressed down move the slider to change the magnification of the grid and also the SEM magnification relative to the grid Select the grid with the mouse and slide the pointer in the direction that the stage is to be moved The Stage position is continually updated in the Pos X and Pos Y boxes When
302. nute intervals using the above jobfile is VB OPER2QSET ONTIME 00 30 00 MCJOBS SDP COM FM FM 00 30 00 19 3 Job preparation The recommended way for inexperienced users to expose a substrate with a prepared pattern is to follow the sequence below Typically the user will want to expose an array of patterns and this can be done efficiently using the layout command The layout command allows the definition of a rectangular array of cells see Figure 19 11 and will carry out a defined set of operations on each cell in turn The jobfiles described below make use of the layout command Experienced users may want to adapt the routine and files for their own applications Substrate Layout of dies Centre of layout of dies A layout cell Figure 19 11 A layout of cells as used by the layout command 19 3 1 Layout parameter file set up Copy the template file vb jobs layout parameters template com to a suitable file name for the exposure The file defines various parameters for the exposure of a layout as symbols in the Job Control Window Edit the file and change the parameters to those required These symbols are used by a second file WLVD COM which issues all the required commands to expose the defined layout KKK KKK KKK KKK KKREKRK eee eee Header for all layout parameter files eek KKK KKK ee ke ke ke KERR ERE ON ERROR THEN GOTO FINISH ALL Part Number 878275 Vectorbeam Operator Manual Page 219 ON CONTROL
303. o deflect the beam so that it tracks the stage see Figure 9 1 The beam error feedback BEF range is 20 um for both 2 512 and 2 1024 interferometers The main beam error feedback BEF correction is applied continuously by the hardware but the fine BEF scaling and rotation corrections are ONLY applied when pattern writing or mark locating The scaling and rotation corrections will not therefore be updated if the BEF changes at other times due to e g a stage move To update the BEF scaling and rotation correction simply execute an on axis mark locate either main or sub field The BEF corrections will be updated to the currently required value regardless of whether the mark locate is successful N B The POSM qualifier on a mark locate will move the stage AFTER the mark locate and this will change the BEF value This may cause the SEM image to show the mark as not being exactly on axis However to get around this and update the BEF corrections carry out another mark locate without a postmove Part Number 878275 Vectorbeam Operator Manual Page 68 Position Devlotloarn Signaty 7 ASource Image Plane In ter feromete Comparator tensor Figure 9 1 Image plane deviation compensation scheme 9 1 3 Magnetic map correction The rotation of the deflection field with respect to the stage axes depends to a small extent on the stage position This behaviour is repeatable and is partly caused by local variations in the ma
304. o reduce the exposed feature size and to increase the exposure dose so that the final size of the features in the resist after development remains the same Care must be taken that small shapes are not lost when applying the bias When this technique is applied to gratings the defined linewidth may be reduced to the point where single pass lines are exposed The advantages are better process control and a reduced risk of residue in the centre of shapes due to the larger dose The disadvantages are greater pattern processing and exposure times Examples Three examples are described The first is a case where the design grid is much larger than the pattern generator resolution grid The second is a case where the design grid is about the same as the pattern generator resolution grid and the third is a case where the design grid is much smaller than the pattern generator resolution grid Example 1 The first example pattern is a grating of size 4 0 x 0 3 mm consisting of lines across the width with a pitch of 400 nm and a linewidth of 200 nm As the line length is a multiple of 200 nm the design grid is 200 nm The minimum dimension is also 200 nm The beam energy is to be 100 kV The uniform pattern and good development control allow the use of a beamstep size only half of the minimum dimension i e 100 nm B 2 No negative bias The pattern generator grid snapping mode should be set to float to enable the placement to nearest resolution
305. o the die by die alignment marks which will be used for the devices 3 The maximum number of rows or columns is 17 The minimum number must be at least 1 more than the mathematical order of the map to calibrated The orders available are 3 4 or 5 so a sensible minimum is 6 In practise it is convenient to use a 17 x 17 grid since 9 x 9 and 5 x 5 grids can be selected from the 17 x 17 grid by specifying a mark spacing twice or four times the real value 4 The reticle with the array of marks should be manufactured by the same route as the device reticles If the manufacturing process introduces distortions they must be the same as on the devices 5 The geometry of the marks must suitable for location by the VB and the marks should have good contrast when viewed with the VB Stepper lens map calibration Note 1 The machine should be fully calibrated 2 The desired position of the bottom left hand mark in the grid relative to the lens map origin can be set up using the xorigin and yorigin qualifiers The lens map origin is the extreme bottom left hand corner of the optical exposure If the qualifiers are not given qcal lens will assume that the origin coincides with the centre of the bottom left hand mark This is a slightly dubious assumption since it implies that three quarters of the mark are beyond the boundary of the reticle are therefore could not have been exposed optically 3 The qcal lens command will tolerate a few fail
306. o toe st reta ioo oo ea sed en e Seen ee ade ee iaiia 118 915 EXDOSUIGGEG isti iiie iii teorie oie EE dete Hr pies 119 Part number 878275 Vectorbeam Series Vectorbeam Operator Manual 19 9 OSS CONN as vies eet a eb edn asa IDE venie us 119 19 234 JDose controller bandiSettip 3 ene ct ite nce 119 13 2 2 Update of pattern generator frequencies on exposure 120 19 29 VMS leglCals 3 ecrire ic ONERE 120 19 3 Exposing proximity corrected pattes inses siot roni ae dava aiana 121 13 3 1 Dose controller operation with proximity corrected patterns 121 19 3 2 Proximity COITOCIOF iuit aieo tepore aiai 121 13 3 3 Transferring dose distribution to the VB seessssusss 122 19 3 4 Notes for CATS converter users sssessssseeeenene 123 13 3 5 Notes for Caprox converter users ssssseeeeenes 124 19 9 6 Exatlipleclockfile sin cerit GE EN IO ene ena 125 13 3 7 Exposing proximity corrected patterns seessesssssss 125 0 e v ti Beamciirent mole RES DSTI 127 14 2 Photomultiplier backscatter detector sese 127 14 3 4 quadrant P N junction backscatter detector ssssssesss 127 14 4 Transmission DELEON iuueni cies csko dett deti bg a icd Roda iiL a Res uk eund du neb de aM dad 128 1s Machine Seb DE Lo aida rene BAT ISi Temperature COON ae uero as va pd Tos
307. oad filename test test col 4 Restore the settings VB OPER sfab restore test 5 If necessary load the magnetic map for the holder to be used from file If this is not available see Section Magnetic map database generation VB OPER dgfile load lt filename mag_cal gt mag_cal 6 Carry out a jobcal This scheme will ensure that the correct values are always used and is particularly useful as Emma does not show what the last loaded database was Deflection field corrections confidence check As a precaution the deflection field corrections database which has been loaded should be checked See Section Checking deflection field corrections in Chapter Calibrations Calib com Many machine set ups are stored in the file calib com This allows easy display and editing of various parameters The file vo seq calib com is run during the Emma start up sequence and loads these parameters into the machine However this file duplicates many of the set ups which are saved in the databases and care should be taken that both are in synchronisation The only parameters in calib com that are not stored in the machine database are Part Number 878275 Vectorbeam Operator Manual Part Number 878275 Page 115 The height meter scaling coefficient and offset The scaling for height coefficients x and y offset scaling and squared terms The temperature sensor coefficients Information about the cathode operating point no parameter
308. obal cells see Figure 20 2 Global CellsInX Defines the number of global cells in the x direction see Figure 20 2 Global CellsInY Defines the number of global cells in the y direction see Figure 20 2 Global offsetx 0 25 The offset in mm in the x direction of the lower left corner of the lower left alignment mark from the centre of the global cell see Figure 20 2 Global offsety 0 25 The offset in mm in the y direction of the lower left corner of the lower left alignment mark from the centre of the global cell see Figure 20 2 Global marksizeX 0 01 The X dimension in mm of the global alignment mark This can be given as 0 00 if it is desired to set the global offset to be the distance to the centre of the mark instead of the corner Global marksizeY 0 01 The Y dimension in mm of the global alignment mark This can be given as 0 00 if itis desired to set the global offset to be the distance to the centre of the mark instead of the corner Global CDCX 2 5 The X pitch in mm of the global alignment marks within cell see Figure 20 2 Global_CDCY 2 5 The Y pitch in mm of the global alignment marks within cell see Figure 20 2 Global MARKER Mark 10 Part Number 878275 Vectorbeam Operator Manual Page 239 The name of the mark definition for the global alignment marks Global_Mark_Locate AUTO This is optional and defaults to MAN Defines mark locate procedure
309. obfile type CTRL C 2 Enter stop at vb oper prompt 3 Enter spell at vb oper prompt During Emma command execution If an Emma command is being executed such as a pattern exposure or calibration 1 Press abort button on Emma status window 2 In the DECTerm running the jobfile type CTRL C 3 Press continue button on Emma status window 4 Enter stop at vb oper prompt 5 Enter spell at vb oper prompt Notes on CTRL C and CTRL Y Emma is prone to crash on repeated CTRL C or CTRL Y Therefore only press CTL C once not CTRL Y and wait for the current command to complete its execution This could be several minutes if the machine is for example doing a mainfield calibration The abort button can be used to stop the current Emma command If the jobfile does not terminate after an appropriate time try CTRL C again The behaviour of DCL jobfiles when CTRL C is pressed is also determined by the current settings in the DCL environment see DCL documentation or help on and help set This should be explicitly set within each jobfile for example Part Number 878275 Vectorbeam Operator Manual 8 8 8 8 1 Page 63 SETON _ enables specified action using ON command ON CONTROL_C THEN GOTO CTRL_EXIT CTRL_EXIT Commands to tidy up before exit EXIT Once the jobfile has been stopped another Emma bug needs to be dealt with The mailboxes need to be resynchronised by running emma ctrl release ve
310. occur during qcal main when updating scale keystone and rotation Retry command check ethernet link reboot PG stop Emma and run Emma environment exit Emma and session CURRFAIL Failed to set requested current Check the validity of the current spot and demg tables Check the current column adjustment Check the current holder positions Check the current autofocus and autostig parameters DBBADMC This file was created on a different machine The values such as the substrate load position in the database are unlikely to be correct Use only with caution DBBEFSENS Bef sensitivity in file different to current value Set current value to that saved in file if known and then retry Alternatively use the override qualifier but note that all other mismatches will be overriden DBEHT EHT in file different to current value Database file specified may be incorrect for current EHT Use appropriate database DBFILEHEAD Bad database file header Use alternative database DBFLD Field size in file different to current value Set current value to that saved in file if known and then retry Alternatively use the override qualifier but note that all other mismatches will be overriden DBFOCSENS Height focus factor in file different to current value Set current value to that saved in file if known and then retry Alternatively use the override qualifier but note that all other mismatches will be overriden DBHRSENS Height rotation factor i
311. ometer in Cabinet B The beam is focused on the substrate by adjusting lens 3 C3 so that the fine focus setting is appropriate for the substrate height This will normally be done by the Installation Engineer and the setting of lens 3 C3 is fixed from this The procedure is described under the section Lens C3 set up below Part Number 878275 Vectorbeam Operator Manual 15 6 1 Page 135 After this fundamental conjugate blanking setup when the beam current is changed by adjusting lens 1 C1 the focus of lens 2 C2 is restored to the conjugate point by simply ensuring that the beam is accurately re focused on the substrate Since the lens 3 C3 value is fixed and the fine focus setting only takes the substrate height into account an accurate focus can only be obtained if the focus of lens 2 C2 is at the same point as after the fundamental setup ie the conjugate blanking point The procedure to set lens C2 to focus the machine in this way procedure is described under the section Focusing the beam in Chapter Job specific machine setups Lens C3 setup The lens 3 C3 setting is a fundamental setup that is done once at installation or after mechanical work on the column It needs to be determined for each beam energy Once it has been determined it remains fixed and it should be saved in the databases for that beam energy This is necessary to ensure that the mainfield distortions subfield distortions field focus and
312. on the stage Part Number 878275 Vectorbeam Operator Manual 10 3 10 4 Page 75 Height meter readings In order to obtain a height meter reading it should be ensured that the correct table has been selected see Section Table selection in Emma When a height meter reading is taken The automatic gain adjustment is turned on The laser is permanently on The signal level is adjusted automatically in small steps A height reading is taken after an integration time together with a quality assessment see Section Height meter warnings and error messages The time required for the reading depends on the amount of integration required to obtain the correct signal level and varies from about 20 ms for the brightest substrates upwards If the quality assessment shows a signal level outside the correct range steps 2 and 3 will be repeated several times The final reading will be reported together with the quality assessment The automatic gain adjustment is turned off The calibration values are restored whenever the Emma command qdisplay height table is issued Otherwise the last laser drive and gain values found for a height meter readings will be used initially For this reason after selecting the correct table use only the command qdisplay height as gradual reflectivity variations across the substrate will be followed with less time overhead When using the real time height mode it is important that th
313. onal defaults to cell and can be cell column or row If the first part of the operator is the dose for a cell is calculated as startdose n dosestep dosestep is in uC cm where n depends on the secord part If the first part of the operator is the dose for cell is calculated as startdose dosestep dosestep is a number where n depends on the second part If the second part is cell n is incremented by 1 starting at 0 for each successive cell in the order in which they are exposed If the second part is row n is the row number 1 so that the dose depends on the row number with the startdose being in row 1 If the second part is column n is the column number 1 so that the dose depends on the column number with the startdose being in column 1 BEAM 100kv 10nA 300um Defines the name of the beam to be restored before starting The database file with the appropriate column settings must have been saved previously in the directory mcdb with the extension test mcdb 100kv_10na_300um test If 999 is given then no beam is loaded PITCHINX RI Part Number 878275 Vectorbeam Operator Manual Page 220 Defines the pitch in mm in the x direction between dies PITCHINY ie Defines the pitch in mm in the y direction between dies CELLSINX 10 Defines the number of dies in the layout in the x direction CELLSINY Defines the number of dies in the layout in the y direction LAYOUTCENTRE 70 00
314. one using the corner measurements and used to eliminate the linear terms From the measurements in the private DW mapped coordinate system a pair of two dimensional polynomials up to 5 order are computed which represent the best approximation to the higher order deviations Up to five stepper lens maps can be loaded at any one time but the appropriate one must be selected for exposure using the qmap lens command The success of the technique relies on the ability to measure the true stepper lens distortions unaffected by other distortions such as the deviations from true grid of the reticle pattern and the deviations from true grid of the VB Normally the second order deviations from true grid of the VB stage are small enough so that the absolute stage mapping mode does not have to be set up first However one of the limits of the stepper lens correction is the ability of the VB stage to maintain any distortion it may have over the area in which the stepper lens was calibrated over its entire range If the stepper lens correction is applied to a VB exposure overlaying a stepper exposure which consists of four separate chips inside a single stepper field the chips will each have different distortions The stepper lens correction is designed to apply a correction to the pattern position in addition to the usual direct write mapping correction The direct write mapping is relied on to apply the linear corrections and the stepper lens map is us
315. onitor this effect is to run the beam diameter measurements with a small enough beam current so that the result is dominated by the edge sharpness rather than the spot size The edge sharpness for all edges should be lt 100 nm If this is not the case then another mark should be found and defined in the holder table Since the horizontal and vertical edges are used for mainfield calibrations they contaminate more quickly and another indication of contaminated FM mark is a larger measured diameter on the horizontal and vertical edges than on the diagonal edges Part Number 878275 Vectorbeam Operator Manual Page 155 Figure 16 9 Contaminated datum mark that should not be used for calibration any longer 16 5 9 Holder position table caspos The jobfile caspos com maintains a 10 position table of holder numbers to give users a shortcut way of calling the holder initialisation program using the symbol H For example for a VB5 the holder number in caspos table position 4 is initialised by VB_OPER gt H 4 For a VB6 VB_OPER gt H C4 The symbol H is defined as VB SEQ holder_init This command file calls the file caspos com which in turn calls the file holder_table com with the holder name specified in the caspos table Therefore holder_table com must contain the necessary information about the particular holder to be initialised Note that for a VB6 H with no C prefix before the holder number calls holder_init com which
316. ons carry out another iteration of the calibration rather than run fullcal to ensure the required accuracy in a shorter time This should only be done using exactly the same number of grid points and the same cover as were used for the original fullcal If the fine tuning is done using different grid or cover than for the original calibrations the original errors will not be properly corrected If the original grid and cover are not known it is necessary to initialise the corrections and calibrate from scratch 18 8 1 Field focus and stigmation This should be done before the main field distortions are fine tuned 1 Align final aperture 2 Run jobcal VB_OPER gt JOBCAL 3 Adjust the on axis focus and stigmation as described in the Section Automatic focus and stigmation adjustment in order to define the parameters which will be used by the next step 4 Calibrate the field focus and stigmation VB_OPER gt QCAL STIG FM DIST NOALIGN GRID 5 COVER F TRNLMN VB_COVER DAC POS 3 ACC 0 005 ITER 1 DIAG 5 Check the corrections as described in the section Checking the deflection field corrections above 6 When satisfactory results have been obtained save the field focus and stigmation corrections to the appropriate database see Chapter Databases 18 8 2 Mainfield distortion This should be done after the field focus and stigmation has been fine tuned 1 Align final aperture 2 Switch the main field di
317. ons of the machine These limitations include the coarseness of the pattern generator resolution grid affecting the amount of grid snapping field and subfield stitching errors the maximum current which can be focused into a given spot size and the minimum spot size The converters typically allow the pattern file to be manipulated using various operators such as overlap removal scaling tone reversal biasing and mirroring These manipulations are extremely convenient for microfabrication regarding respectively double exposure device scaling positive negative resist process bias and mask manufacture Pattern data conversion and processing software In Vistec Vectorbeam Series Systems the converter program is provided by third party vendors typically either the CATS software from Synopsys Inc or the CAPROX software from Sigma C GmbH These vendors typically provide training for the use of their programs and the appropriate manuals should be referred to when using the programs Some basic concepts and examples are however described below CATS CAPROX runs on any DEC VAX or Alpha computer system running OpenVMS and provides facilities for converting pattern data from various CAD formats typically GDSII to Vectorbeam binary pattern format Patterns can be displayed on the high resolution screen of the user control computer system and such features as pan zoom and measure allow inspection of critical dimensions Many plotters are sup
318. ood contrast and signal The bright part of the mark should be fairly close to the maximum brightness and the dark part of the mark should be fairly close to completely dark Retry the mark locate 2 Check each of the mark locate parameters used and ensure they are reasonable as described earlier in this chapter Retry the mark locate 3 Centre the mark on the SEM screen using the joystick function 4 Reduce the minimum contrast level to 5 96 temporarily Retry the mark locate 5 Use the diagnostic output of the mark locate function as described earlier in this chapter and plot the sampled video data using the PVWave routine Examine the plots and see if the edges of the mark can at least be recognised manually 6 Increase the amount of filtering and oversampling to reduce the effects of noise in the video signal 7 If the pit mark locate is being used with no success try switching to the edge routine 8 Login to the DCP see Section Logging in to the subsystem controllers in Chapter Computer system retry the mark locate and check for more detailed error messages Part Number 878275 Vectorbeam Operator Manual Page 105 12 Databases Emma automatically and immediately saves every update to any machine parameter e g calibrations or settings in a reserved filename esp03 ebeam dbase in the directory vb db This file is called a database and is loaded whenever Emma is restarted thereby restoring the last setti
319. or Balance Panel Figure 7 6 is used to set up the stigmator Part Number 878275 Vectorbeam Operator Manual Page 52 drive electronics to eliminate beam shift Emma help Balance Axial X Axial Y Diag X Diag Y Done Reset Cancel Figure 7 6 Set Stigmator Balance The stigmation and the focus may be manually adjusted using the Set Stigmation Focus Panel Figure 7 7 by keeping the left mouse button pressed down and moving the slider bar until the stigmation is satisfactory Repeat with the other slider bar Emma help Focus Stig Mark description Focus Auto Focus Done Reset Cancel Figure 7 7 Set Stigmator Focus Part Number 878275 Vectorbeam Operator Manual Page 53 The Set Gun Aligner panel Figure 7 8 is used to set the gun alignment manually so that the beam passes down the column The emission image is used for finding the beam when it has a large misalignment Emma help Aligninent Tilt X Tilt Shift X Shift Y Emission image Figure 7 8 Set gun aligner The Set Video Level panel can be used to set the video levels manually via the mouse or select the automatic adjustment Emma Help Manual Gain Backoff Filter Ho H H Done Reset Cancel Figure 7 9 Set video level Part Number 878275 Vectorbeam Operator Manual Page 54 The Set Lens Adjustment panel Figure 7 10 may be used to set the lenses manually via the mouse
320. or replaced by hand from the library slots VB6 single holder airlock Press airlock full vent on the vacuum control panel twice in quick succession The green vacuum LED will go out Wait several minutes for the airlock to reach atmospheric pressure Unscrew airlock door catch pull out the trolley until it reaches the mechanical stop and remove or replace the holder Figure 16 1 shows the orientation of the holder in the airlock Part Number 878275 Vectorbeam Operator Manual Page 147 Airlock Door Figure 16 1 Diagram showing orientation of holder in single chuck airlock where a indicates the reference pins b indicates the Faraday cup c indicates the datum plate d indicates the molybdenum block e indicates the origin of the stage coordinate system and f indicates 16 4 3 the wafer flat although the wafer orientation can be varied VB6 ten holder airlock Press airlock full vent on the vacuum control panel twice in quick succession The green Pe1 LED will go out Wait several minutes for the airlock to reach atmospheric pressure Make sure the holder trolley is pulled fully back from the door against the stop as otherwise the door will not operate The door up arrow and down arrow buttons will light up when atmospheric pressure has been reached in the airlock The door will only open when atmospheric pressure has been reached The airlock can be moved to the required holder position befor
321. ot used On a VB5 using a combination of X ray mask holders and any other holders requires separate magnetic maps to be created and saved as partial databases see Section Magnetic map calibration and below On a VB6 using a combination of 8 inch wafer holders and any other holders requires separate magnetic maps to be created and saved as partial database see Section Magnetic map calibration and below Use the following command to save the database VB_OPER gt dqfile save lt filename dbase gt total 20kV basic database A suggestion for the filename to make it self explanatory is VB OPERo qfile save 20kv_setup dbase total Field corrections database generation To generate the field corrections database for a particular fieldsize the fullcal routine should be used When a fullcal has been carried out only the field corrections should be saved by using the commands VB OPERo dfile save filename dist dist cal 20 kV distortion corrections VB OPER qfile save lt filename stig gt stig_cal 20 kV focus stig corrections Beam database generation Set up the beam as described in the Section Beam current adjustment or Beam diameter adjustment When a beam has been set up the current column conditions should be loaded into one of the three complete column set ups coarse test or fine which are available in Emma using for example VB OPERo sfab update test This column set up can then be saved using for e
322. out of range Reduce any entered value The subfield sensitivity may need adjusting 0x20830019 DCP Subfield DAC out of range Reduce any entered value The subfield sensitivity may need adjusting 0x2083001A DCP BEF DAC out of range Reduce any entered value The BEF sensitivity may need adjusting 0x2083001B DCP Channel plate lost refresh 0x2083001C DCP Bad video too little contrast etc Check the SEM image on a suitable mark 0x2083001D DCP Focus not found video not good enough Check video levels and image in SEM mode Check autostig parameters or detector selection or column adjustment Vectorbeam Operator Manual 27 6 6 18 27 6 7 Part Number 878275 Page 294 0x4083001E DCP Unable to initialise clock Clock hardware not present or error Reboot PG 0x4083001F DCP Readback failed in clock Clock hardware not present or error Reboot PG 0x20830020 DCP Function not valid for current clock H W Possible PG configuration file error 0x20830021 DCP Clock band too narrow Widen band 0x30830022 DCP Clock band wide you may lose accuracy Reduce band 0x30830023 DCP Clock freq outside band nearest poss was set 0x20830024 DCP Autostig zero exel size Increase scanlength 0x20830025 DCP Autofocus stig Curve fit has no maximum Adjust the focus and stigmation manually in SEM mode Check the autostig autofocus parameters and try again 0x20830026 DCP Autofocus stig Curve fit is out of h w range Adjust the foc
323. pattern generator resolution Figure 19 8 shows the results of using this option with a VRU of 2 to expose the pattern shown in Figure 19 7 Figure 19 8 The exposed exels shown as circles give squares which alternate between 4 and 6 PG resolution points in size and which alternate between 10 and 12 PG resolution points in pitch Both the pitch and size are not correct Float option The float option causes the lower left shape vertex to be left on the nearest pattern generator resolution point with the upper right vertex being an integer number of beamsteps away Figure 19 9 shows the results of using this option with a VRU of 2 to expose the pattern shown in Figure 19 6 Part Number 878275 Vectorbeam Operator Manual Page 213 Figure 19 9 The exposed exels shown as circles give squares which are 4 PG resolution points in size and which have a pitch of 11 PG resolution points in pitch The pitch is correct but the size is too small Split and bury option The split and bury option causes the lower left and upper right shape vertices to be left placed on the nearest pattern generator resolution points However since the shapes are not an integer number of beamsteps in both width and height they are split into 4 shapes that are Figure 19 10 shows the results of using this option with a VRU of 2 to expose the pattern shown in Figure 19 6 Figure 19 10 The exposed exels shown as circles give squares which are 5 PG resolut
324. pe Specify valid dose type DVECALLOC dvector allocation failure DWABSMODINV Absolute mode is invalid for this command DWBACKSOLN Invalid DW back transform Part Number 878275 Vectorbeam Operator Manual Page 275 DWCOPTOSELF Can t copy current mode to itself Specify two different modes DWCURMODINV Current mode Absolute is invalid for this command DWDIVZ3DD Division by 0 in DW back transform 3D denom Failed internal consistency check Make sure the guilty mode has been set up correctly Check expected and observed positions Try restarting Emma Contact Vistec if it persists DWDIVZ3DSY Division by 0 in DW back transform 3D ScaleY Failed internal consistency check Make sure the guilty mode has been set up correctly Check expected and observed positions Try restarting Emma Contact Vistec if it persists DWDIVZ4DCHI Division by 0 in DW back transform 4D Chi Failed internal consistency check Make sure the guilty mode has been set up correctly Check expected and observed positions Try restarting Emma Contact Vistec if it persists DWDIVZ4DD Division by 0 in DW back transform 4D denom Failed internal consistency check Make sure the guilty mode has been set up correctly Check expected and observed positions Try restarting Emma Contact Vistec if it persists DWDIVZADSX Division by 0 in DW back transform 4D ScaleX Failed internal consistency check Make sure the guilty mod
325. perture nm 17 9 1 2 50 kV Beam current diameter for diameter for diameter for diameter for 100 um 200 um 300 um 400 um aperture nm aperture nm aperture nm aperture nm Part Number 878275 Vectorbeam Operator Manual Page 176 17 9 1 3 100 kV Beam current Beam Beam Beam Beam nA diameter for diameter for diameter for diameter for 100 um 200 um 300 um 400 um aperture nm aperture nm aperture nm aperture nm 17 9 2 FEG with UHR final lens The source diameter was assumed to be 30 nm the emission intensity was assumed to be 0 5 mA sr 17 9 2 1 100 kV Beam current Beam Beam Beam nA diameter for diameter for diameter for 50 um 65 um 100 um aperture nm aperture nm aperture nm Part Number 878275 Vectorbeam Operator Manual Page 177 18 Calibration 18 1 Overview The following table lists the calibrations in the order in which they are performed Calibration When is the calibration performed Height meter For each substrate See Section Height meter calibration Magnetic map For each beam energy See Section Magnetic map calibration After the final lens or stage has been assembled after servicing Separate magnetic maps are required for holders with datum plates in the corner those with datum plates centred along an edge Demagnification table For each beam energy each time Vistec engineer the column not gun has been assembled after servic
326. point The pattern generator resolution must be a sub multiple of 100 nm and the largest value within the range for standard fieldsize machines see table of maximum fieldsize pattern generator resolution is 8 333 nm 100 12 8 333 This gives a maximum fieldsize of 0 546133333 and enables a block height and width of 500 um to be used The block height and width are integer multiple of lines and so the block boundary will occur at the same position relative to the lines at each boundary along the grating In addition there will be no boundary across the width because the width of 300 um fits inside the 400 um block Using negative bias A negative bias of 100 nm is applied to the width of the 200 nm lines This variation uses the same setup as above Using an exposure grid of 100 nm the lines become single pass lines The dose should be increased to achieve the 200 nm linewidth required Note The linewidth and pitch of gratings are often not integer number of Part Number 878275 Vectorbeam Operator Manual 19 1 14 2 19 1 14 3 19 2 Page 215 nanometres and this can give a pattern generator resolution grid size with many decimal places In addition if the grating is defined using the ctxt facility in CATS resolve may need to be set appropriately Example 2 The second example pattern is an array of gates that has been designed on a grid of 5 nm The gate width is 100 nm and is the minimum dimension of the pattern The
327. ported to generate a hard copy of the graphical pattern data The converter program will produce either a FRE or a VEP format output file to be used directly on the Vectorbeam Part Number 878275 Vectorbeam Operator Manual Page 194 It is usual to prepare pattern data on a separate computer to avoid using machine time and then copy the files to the Vectorbeam user terminal The pattern data are downloaded from the Vectorbeam operator terminal via the Ethernet link transfer speed approximately 200 Kbytes sec to the pattern generator during exposure The user manuals for these programs should be referred to for details of operation The application of the programs to the Vectorbeam is given below 19 1 1 1 Notes for Cats users 19 1 1 1 1 Format VB VB20 VB50 VB100 To generate files for the Vectorbeam the format must be set to VB VB20 VB50 or VB100 The format VB is the basic option which can generate files for use at any kV and for which the Cats resolution parameter corresponds directly to the Vectorbeam pattern generator resolution The formats VB20 VB50 and VB100 decouple the Cats resolution from the Vectorbeam pattern generator resolution and place limits on resolution beamstep and the Cats VB resolution parameter in order to guide the user to choose the correct parameters to match the VB hardware The decoupling of the Cats resolution from the Vectorbeam pattern generator resolution extends the maximum value o
328. ppm scale offset on the X mainfield and a 15 ppm offset on the Y scale type the following VB OPER qcalibrate offset main xscale 0 000010 yscale 0 000015 To clear all mainfield offsets type VB OPER qcalibrate offset main init load Subfield The subfield scale and rotation calibrations can be offset However any offset on the mainfield is automatically also applied to the subfield Therefore only additional subfield calibration offsets are required to be defined explicitly Stepper lens calibration The stepper lens calibration is intended to allow more accurate direct write alignment when using the VB to expose wafers which have other layers exposed on a stepper The lens being mapped is the lens of the optical stepper and not the VB electron lens A modern stepper lens has typically 100 nm or less of distortion in the image which covers typically an area of up to 22x22 mm The distortions are non linear and so cannot be corrected by the standard direct write mapping which only deals with 1 order terms Characteristic stepper field distortions lens map are measured and then reproduced during e beam exposure The stepper lens map is calibrated from an array of marks on a substrate exposed by the stepper within a single stepper field The array must cover the stepper lens field and will have the characteristic stepper lens distortions The array is measured in the VB using the qcal lens command A private DW mapping is d
329. q tud Dabei ce gd ea ad ees 187 18 10 1 Stepper lens calibration substrate eesessssssss 188 18 10 2 Stepper lens map calibratiO siisii aisada 188 16 10 9 Checking the CalibrathON xi scstesecsscenedespesadacassetedensstadadeesetaaachsacneacnssets 189 16 10 4 WSC OFICASINADS vases siccsiarectececsccatsnzahatececeesdiay a 190 18 10 5 Further notes on using the stepper lens correction 190 19 Exposing a supstrate Liebe ccc vcccccesnsasccccnesosantansocciocemuncinns LII 1941 Patern data PROC ARAN e irri brit reb HE Tae abt Pere ia leges e pieta 193 19 1 1 Pattern data conversion and processing software 193 19 1 2 Setting pattern file attributes on DEC compulter 198 19 1 9 Pattern generator FesolUtlOn saccos ooo roi bete etes oreet clescenvectes 199 19 1 4 Choosing the pattern generator resolution maximum fieldsize 199 19 1 5 Blook height and Widi 3 cecacia coo ri eo eoe tt their T eren ite tortis 202 TING Ghoosing tle DIOCKSIZO ocaeca etos tto iscritti etes ro eror toris 205 19 1 7 Sublleld TFactiifiQgsuciuuductaa ccce rita ont sace etat rote diac eet tnc toe dd ie eeddes 207 19 1 9 CHOOSING TNE subflelds oca dent sut td tures tapa ceret tec te ettet itae 208 19 1 9 Beamstep size and VU sciret ate te tero sabes eet tat voe rh renes as 210 19 1 10 Choosing the spot size beamstep size 211 19 1
330. r due to the larger number of subfield changes The periodicity of the subfields may be matched to periodic sensitive areas of the pattern to avoid subfield stitch errors occurring in these areas Beamstep size and VRU The beamstep size is related to the pattern generator resolution by Beamstep size Pattern generator resolution x VRU Equation 19 15 where VRU can take the following values e 16 bit machines 1 2 4 8 16 32 e 18 bit machines running V2004 02 or earlier 1 2 4 8 16 32 64 128 e 18 bit machines running V2005 01 or later Any integer between 1 and 128 e 20 bit machines running V2004 02 or earlier 1 2 4 8 16 32 64 128 256 512 e 20 bit machines running V2005 01 or later Any integer between 1 and 512 Setting the VRU parameter enables the exposure of shapes with spot sizes that are large compared with pattern generator resolution Using a large spot size without the VRU shapes would be oversized and the required pattern generator stepping frequency would be high Part Number 878275 Vectorbeam Operator Manual Page 211 Figure 19 6 Effect of setting VRU 2 The black and white discs represent the individual pattern generator grid points The black discs represent the points which are exposed separated by the beamstep size In addition 18 and 20 bit machines running V2005 01 or later can define a VRU to be associated with any clock and override the default VRU set using the qset VRU comm
331. r explanation of the layout of direct write alignment marks 1 Marks must be isolated from other marks and other features including dirt by at least twice the uncertainty in their position if automatic mark locate is to be used The search range of the mark locate must take this uncertainty into account For a direct write registration using global alignment followed by die by die alignment the following applies e The global alignment marks should ideally be around the perimeter of the area to be exposed for best accuracy e The uncertainty in the position of the first mark to be found after a substrate has been loaded will be several tens of microns at best and may be as much as a few mm If the uncertainty in the position of the first mark is more than the maximum coarse search range 200 um then manual global alignment is carried out in SEM mode It may be advantageous therefore to place such a mark a relatively large distance from any sensitive device areas e The uncertainty in the next 2 or 3 marks positions for global alignment will be much smaller typically lt 100 um and depends mostly on the rotation of the substrate e After a3 or 4 mark global alignment has been carried out the machine can move accurately to other marks for die by die alignment The uncertainty in the position of these marks will be determined as much by the accuracy in their manufacture as by the machine operation machine and can be less than 1 um 2
332. r for moving accurately to the die by die alignment marks or for exposing the dies directly 20 2 2 Die by die alignment A 2 3 or 4 mark global alignment is carried out first as described above The global alignment enables alignment marks associated with each die to be placed accurately within the automatic mark locate search range The uncertainty in the position of these marks will be determined as much by the accuracy in their manufacture as by the machine operation and can be less than 1 um The die by die alignment is usually fully automatic and consists of aligning to each die in turn and exposing 20 3 Layout definition for regular rectangular alignment mark arrays The layout program utility is used to access any regular 2 D rectangular array of positions on any substrate by referring to indices rather than coordinates 20 3 1 Global alignment mark layout An example substrate with global alignment marks is shown in Figure 20 2 along with the chosen layout of 2 x 2 global cells The 2 x 2 layout allows all the global marks to be accessed other layouts might also work The global cell layout is only used to carry out the global alignment and then the die layout is selected The centre of the global cell layout is always taken to be at the same position as the centre of the die layout and this is what relates the positions If GlobalPitchX is positive GlobalPitchY is positive Global OffsetX and Global OffsetY are negative
333. r rectangular marks 11 9 2 Fine search When the fine scan routine is invoked it will already have been supplied by the coarse search with a position which is a close approximation to the centre of the mark It can therefore confine its scans to regions in the immediate vicinity of the expected edge positions The fine scans do not scan over the central part of the mark to save time as no additional information is gained The fine scans consist of a series of short line scans the centres of which are placed on each of the four expected edges The scans are horizontal or vertical for vertical and horizontal edges respectively The number of parallel scans is given by the parameter parascans These scans are in the region defined by MH or MW and of length MLEN If any of the scans fail to find an edge they are discarded unless more than half fail in any particular grid in which case an error is returned Once a satisfactory set of scans on an edge has been gathered a least squares fit is carried out on the edge data to provide an equation describing the edge The resulting equations two describing the vertical edges and two describing the horizontal edges are then used to determine the marks centre This centre position is then returned to the user or calling routine This behaviour is illustrated in the following diagrams Although the illustration is for a square mark this can easily be adapted to fit an octagon by simply ensuring tha
334. r which the scans are spaced out is given by the MH and MW parameters If any of the scans fail to find an edge they are discarded unless more than half fail in any particular grid in which case an error is returned The fine search scans are shown in Figure 11 13 Part Number 878275 Vectorbeam Operator Manual 11 10 2 1 11 10 2 2 11 10 2 3 Part Number 878275 Page 102 MH OH MW OW Figure 11 13 Diagram showing fine search scans in red for cross mark Default output For each limb one point is calculated by averaging all the points along both outside edges Two lines are calculated one line running between the two points of the vertical limbs and one line running between the two points of the horizontal limbs The output position is the intersection of two lines Linefit output Least squares fitting is carried out on the edge data which gives eight lines describing the left and right edges of the four limbs In the case of a cross with more than one finger in each limb Figure 11 14 only the outer edges of the outer fingers are used The coordinates of the points where the top right lines the bottom right lines the bottom left lines and the top left lines intersect are calculated giving four corners of a box These corners are used to check that the limb widths are within tolerance The four equations describing the vertical edges are averaged to give a vertical line in the centre of the cross and t
335. ration and contain the pattern data fractured and sorted into fields but retaining structure The pattern generator transforms these data into individual shapes and fractures and sorts them into subfields Exposure occurs by scanning all the shapes within each subfield with the sub field deflection coils The subfields are positioned sequentially with the main field deflection coil Stage Controller This microprocessor operates the stage motors the interferometer and the temperature measurement It also links to the bar code reader non Brooks systems On axis controller This 68020 microprocessor is mounted in the IP crate The IP crate also contains the video processing boards The processor operates the video processing for the monitor to provide an SEM image the stand alone EHT supply via an RS232 link from the IP crate the lenses and the gun alignment via the electron optic EO crate in CER1 PICs controller The PICs controller operates the vacuum pumps and gauges and the Genmark robot Communication to this controller is done via the stage processor and a parallel link Logging in to the subsystem controllers To login to the MUP DCP or FISP SLC type Vb super rlogin dcp And provide the username pgdiag and password espritpg To login to the stage processor type Vb super rlogin stage And provide the username svdiag and password espritsv To login to the eo processor type Vb super rlogin oa And provide the
336. ration accuracy of last mainfield calibration accuracy of last subfield calibration accuracy of last BEF calibration accuracy of last focus calibration accuracy of last stigmation calibration current demagnification table spot table 3 stored demagnification tables spot tables selected spot size selected beam current mark slope EHT 12 2 4 8 Coarse 1 2 3 Part Number 878275 EHT filament drive filament mode Vectorbeam Operator Manual Page 110 4 filament standby 5 wehnelt drive 6 wehnelt mode 7 extractor drive 8 runup profile number 9 runup time 10 gun alignment X 11 gun alignment Y 12 gun tilt X 13 gun tilt Y 14 lens 0 15 lens 1 16 lens 2 17 lens 3 18 mid column alignment X 19 mid column alignment Y 20 axial stigmator 21 diagonal stigmator 22 video gain 23 video backoff 24 video target 25 video background 26 video mode 27 video filter 28 SEM magnification 29 SEM scan rate 30 fine focus 31 spare 12 2 4 9 fine The fine partial database has the same parameters as the coarse partial database 12 2 4 10 test The test partial database has the same parameters as the coarse partial database 12 2 4 11 Stepper lens The stepper lens partial database contains 1 Lens map number associated with mode 2 XY coordinates of lens map origin in the DW mode 3 XY positional limits of validity 4 Allthe 5 stepper lens maps mapping order X Y positio
337. re 0801AF22 ETHIPQFAIL Subsystem couldn t queue Ethernet message for input 0801AF2A ETHIPQOVF Subsystem rejected command its i o queue would overflow 0801AF32 ETHEMMASUPCMD A slave processor received an Emma command 0801AF3A ETHDEADCI Subsystem s Command Interpreter process suspended 0801AF42 PGERROR Pattern Generator error message in console window 0801AF4A SVERROR Stage error message in console window 0801AF52 EOERROR EO Control error message in console window 0801AF5A PGFAULT Pattern Generator fault message in console window 0801AF62 SVFAULT Stage fault message in console window 0801AF6A EOFAULT EO Control fault message in console window 0801AF72 PGVXERR Pattern Generator VxWorks error message in console window 0801AF7A SVVXERR Stage VxWorks error message in console window 0801AF82 EOVXERR EO Control VxWorks error message in console window 0801AF8A UNKNSUBSYS Message from unknown subsystem in console window 0801AF92 SBERR Structure Block transfer failed 0801AF9A QUALBADVAL Bad value supplied for qualifier 0801AFA2 SVD2FEWROWS Can t solve SVD matrix has fewer rows than columns 0801AFAA SVDNOCONV SVD matrix arithmetic didn t converge 0801AFB2 SVDDIVZH Division by zero in SVD matrix arithmetic h 0801AFBA SVDDIVZAIL Division by zero in SVD matrix arithmetic Ail 0801AFC2 SVDDIVZAII Division by zero in SVD matrix arithmetic Aii
338. re positioning knobs are on the side of the chamber on HR machines and on the side of the column on UHR machines There are four aperture positions on a VB5 and six aperture positions on a VB6 both HR and UHR Automatic aperture adjustment mechanism The final aperture size may be selected using the command VB OPER aperture move num lt number between 1 and 16 gt Where the number is a stored preset position Final aperture alignment The final aperture must be aligned to the optical axis of the Final Lens lens 3 to obtain the best resolution When the aperture is not aligned the smallest spot size will not be obtained and the position of the beam on the stage will vary with the Final Lens current Aligning the aperture is done by varying the current in the Final Lens around its nominal value and observing and minimising the movement of a feature on the stage Do NOT use QSET WOBBLE ON FFOCUS as this will align the aperture to the fine focus coil Manual aperture adjustment mechanism A program has been written to automate this procedure Move to an alignment mark and type VB OPER G VBSSEQ ALIGN APERTURE COM 15 FM FM 0 005 The program measures the mark position at two Final Lens currents and displays the difference in the positions Observe the output on the screen and make adjustments to the aperture positioning knobs until the difference is smaller than the limit defined on the command line 15 nm Automatic aperture adju
339. reg 3 48 When carrying out a fine focus and stigmation adjustment on the datum to determine whether the fine focus setting is correct for conjugate blanking any focus offset due to the last qadjust field command must be zeroed See following sections for further understanding of this The focus offset due to any previous qadjust field command is done by issuing the following command VB OPER qcal main fm load Part Number 878275 Vectorbeam Operator Manual 17 4 5 17 4 6 Page 169 This command does not carry out any calibration but simply loads the existing values and zeroes the focus offset due to the qadjust field command qadjust field The machine will set the focus based on the height meter reading and the focus for height coefficient when the following command is issued VB_OPER gt QADJUST FIELD The amount of fine focus applied due to height difference current height height recorded at last qcal main focus for height coefficient Equation 17 2 It is important that this command is performed before mark locates are to be performed at a particular position If the stage position is one where the height sensor will not operate then the command should either be performed at a location known to be at a similar height or a the QADJUST FIELD HEIGHT n form should be used to set the height to the required value How the fine focus setting and adjust field combine to drive the fine focus The dr
340. rift is significant Alignment will most likely be necessary after mechanical work on the column If the gun alignment is accurately set up the same settings can be used for a large range of beam currents Fine gun alignment with HR lens The following manual routine for aligning the gun should be used on machines with an HR final lens rather than the automatic routine Part Number 878275 Vectorbeam Operator Manual Page 158 1 Align the final aperture using an intermediate beam current 2 Set up either the largest current which gives a useable image of the calibration mark or just the largest current at which the machine will be operated This is because the gun alignment is more sensitive at high currents and the high current settings can be used at low currents but not vice versa The largest current is typically gt 200 nA for a 400 um aperture Set up the focus and stigmation manually Use any gun alignment settings to maximise the current approximately 3 Observe the SEM image of the datum mark and adjust the magnification so that the mark covers more than half of the screen Locate the mark and move to its position so it is centred on screen 4 Bring up both the lens adjustment panel and the gun aligner panel in Emma under set menu 5 Select the lens 2 setting and with the mouse button held down wobble the lens 2 setting and observe the image The image should remain at the same position within a few microns for lens 2 s
341. roducing an image to be displayed on the video monitor The detectors fitted to a Vectorbeam depend on the particular machine configuration The following table shows the options Machine Beam Photomultiplier 4 quadrant P Transmission type current backscatter N junction detector meter detector backscatter detector VB5 Yes Yes No No VB6 HR Yes Yes No Yes VB6 UHR Yes No Yes Yes The detectors apart from the beam current meter are connected to the video input and may be selected by the control software Emma using the command qset detector Beam current meter The beam current meter consists of a total beam current collection cup known as a Faraday Cup which is mounted on the stage on VB6 Systems and on the holder on VB5 Systems It has pre amplifier and range selection electronics that enable measurement of the beam current from less than 50 pA to more than 200 nA Photomultiplier backscatter detector Four scintillators on the ends of light pipes leading to photomultiplier tubes are mounted at 90 degrees with respect to each other close to point of impact of the beam on the substrate They detect back scattered electrons Their outputs can be mixed in any combination although normally they are summed and this is done with the following command VB OPER qset detector PMP1P2P3P4 4 quadrant P N junction backscatter detector 4 P N junction devices are mounted on the bottom of th
342. rons away from the expected mark positions Reinserting a particular substrate in a particular holder will typically be only within 25 uim With different holders and substrates the shifts are larger Since the coarse search of the mark locate is limited to the fieldsize which may be less than the substrate shift the scheme should ideally involve searching over a larger distance in order to ensure capturing the marks A suggested scheme is to use an array of square marks centered at the required global mark position The pitch of the marks increases from the centre of the array outwards The mark locate is configured to locate the squares Three adjacent marks on a diagonal are located and the measured positions allow the position of the centre mark of the array to be found This scheme permits fast automated global alignment because it minimizes the coarse search area while allowing substrate shifts of up to half the size of the array If the pit mark locate algorithm is used only the first mark will require a coarse search over more than about 1 um For example an array of 8 um by 8 um squares can be used The squares are placed at varying pitch starting at 75 um at the centre and growing in 1 um increments towards the edges of the mark area If there are 27 by 27 squares this would occupy a total area of about 2 mm by 2 mm The size of the array can be adjusted for the area allocated on the substrate eg 500 um by 500 um The starting pit
343. rrection sissid 67 94 9 Magnete map Corector 5 reet erre ne ro t tea 68 9 4 Height Cottecllollo 2 tete op eo EE Es 68 9 1 5 Yaw correction 18 bit and 20 bit VB6 only sssssss 69 9 1 6 Deflection field COrfeclorls oc ord tir tti e edi eed e tide 70 9 1 7 Shift scale and rotation corrections for fine focus 71 9 2 Corrections tor ires 71 9 2 1 Direct write Correcto Nessi tin rin rere ee Se kodak o eve pad 71 9 2 2 Stage mapping machine mode ssssssssseeeee 71 9 2 8 Stepper lens Corect Niipperi aa ens 71 10 Height eter acai XE 105 Peon once 2 afc 2 occ Cu ene ee eet a o eer res eS ran terre 73 TOi TaD SiCMOIG SS DER 73 101 2 Table selectiomin EMME sss a 74 102 Heh meter UCN a ica eoe Hir Ced aaa Uva eire E atero RU D dta buds 74 IOS Height metar reddings ciis eerte vein eodd Ye eta end pecus odi vei 75 TO d Height meter timilbigo saca ui ccpa cus sn tu edu nA Ci EYE CE Pelr A AG OUT UN EUR C detas 75 10 5 Height meter warnings and error messages sssseeeeene 76 10 5 HelOht metar OS BIB dad sunt Lodel t da dus pls FOL Tk Dipl on addo an dub dox ERA Md 76 10 6 1 Hesist thickness combpenhsallOTi cios oiu cicer hte recedant 76 10 6 2 Height meter offset effect on stitch accuracy ssssusss 77 10 63 SOM ol OMS Gl rudi error fossero chi nde Section ie tc edet tue weardtansiesed tees
344. rrections calibration 11 Accuracy of last mainfield field corrections calibration 12 Accuracy of last subfield field corrections calibration 13 Accuracy of last BEF field corrections calibration 12 2 4 5 Stig cal oxooROM PG resolution EHT Max fieldsize aperture fine focus axial stigmation diagonal stigmation fine focus field corrections as a function of subfield index x subfield index y axial and diagonal stigmation field corrections as a function of subfield index x subfield index y Time of last focus field corrections calibration Time of last stigmation field corrections calibration Accuracy of last focus field corrections calibration Accuracy of last stigmation field corrections calibration 12 2 4 6 Size cal 12 2 4 7 Spot The spot partial database contains OARON PG resolution EHT max fieldsize calibration height EHT at which sensitivities are calibrated mainfield sensitivity subfield sensitivity BEF sensitivity mainfield pivot point subfield pivot point focus for height coefficient X Y calibration position rotation for height coefficient mainfield scaling rotation and keystone subfield scaling and rotation BEF scaling and rotation date and time of last mainfield calibration date and time of last subfield calibration date and time of last BEF calibration date and time of last focus calibration date and time of last stigmation calib
345. rrently no support for defining point doses in Emma and so the user must set the frequencies based on Equation 13 3 if point doses are to be specified Resist sensitivity parameter The resist sensitivity parameter is the exposure dose which is used for calculating the clock frequencies by executing qadjust clock command when the clocks are defined as using a relative dose qset dose rel command This is a commonly used operating set up for exposing proximity corrected patterns For this operating set up it is usual that the resist sensitivity parameter corresponds to the large area clearing dose The large area clearing dose refers to the dose required for an area large enough to receive the maximum exposure from backscattered electrons The large area clearing dose is always smaller than the dose required for smaller features By convention the large area clearing dose is assigned to clock 0 and for convenience is assigned a relative Part Number 878275 Vectorbeam Operator Manual 13 1 5 13 1 5 1 13 1 5 2 13 1 5 3 13 1 6 Page 118 dose of 1 0 Smaller features can be assigned to higher clock numbers with relative doses greater than 1 0 by the proximity correction software Usually the higher the clock number the higher the relative dose The same set up can be used for exposing non proximity corrected patterns It is conventional for the pattern converters to assign clock 0 to all shapes when they are no
346. rror should be small but could be introduced by Variation of magnetic field with stage position or a change in temperature of the substrate as the plate is written 5 Scale Linearity error This error should be very small but could be introduced by Stage pitch roll or yaw 6 Bow error Uncorrected Bow error is due to the bow of the zerodur mirrors It should however be very consistent Entering stage map coefficients The command qMAP STENTER allows the user to manually enter mapping coefficients and the mode to which they are to apply The following jobfile is an example Scale and rotation MPST ABS XS 0 0000018 Y R 0 0000323 Keystone MPST ABS X2XY 0 0000000037 Y2XY 0 000000003 Scale linearity MPST ABS X2XSQ 0 00000001 19 Y2YSQ 0 000000001 Bow MPST ABS X2YSQ 0 0000000392 Y2XSQ 0 0000000087 Load the coefficients MVHM Display stage map coefficients The command qDISP MAP enables the user to display the coefficients for an optionally specified mode or the current mode if no mode is supplied VB OPER 4qdisp map abs Current stage mode is NATURAL Coefficients for ABSOLUTE mode are s X Y SCALE 0 00001 4213917000 0 000014529809000 ROTATION 0 001296107700000 0 001379073780000 X SQUARED 0 000000007655390 0 000000007075946 XY 0 00000000281 3361 0 000000015719697 Y SQUARED 0 000000072221504 0 000000020973950 Machine stage map mode set up The machine stage map mode allows the
347. rs are converted by EMMA into units of nm by calculating how much the beam diameter changes if the focus is adjusted by the error value It does this by comparing the ratio of the aperture radius against the virtual main pivot to the additional height required to put the beam back into best focus An accurate calibration should typically give numbers as in the table below Focus See Section Focus Sampled coefficient Mean Sigma errors Sampled coefficient Mean Sigma errors nm lt 0 3 x spot size lt 0 3 x spot size 18 5 Interpretation of fullcal deflection field correction calibrations In contrast to the on axis calibrations the field correction calibrations are not followed by further measurements to assess the accuracy Fullcal carries out two iterations for each field correction and the adjustment made on the second iteration give a measure of how good the calibrations are Part Number 878275 Vectorbeam Operator Manual 18 5 1 18 5 2 18 5 3 18 5 4 18 6 18 6 1 18 6 2 Page 182 Field focus and stigmation Largest observed error for the final iteration of the stigmation should be lt 0 01 Main field distortion The largest observed error for the second iteration should lt 20 nm Sub field distortion The largest observed error for the second iteration should lt 20 nm Beam error feedback distortion This is calculated from the main field distortion and so relies on an accurate main field
348. rs is 1 and for 20 bit pattern generators is 4 Alternatively the scan resolution can also be set directly in nanometres using the RESN parameter The scan resolution RESN sets the distance in nanometres between adjacent points during the fine scan but only to the nearest PG bit This ensures similar signal averaging and mark locate times for different maximum fieldsizes The scan resolution CSRES sets the number of PG deflection bits between adjacent points during the coarse scan A value of 8 means every eighth bit is used and this is a typical value for 16 bit and 18 bit pattern generators A value of 32 is typical for 20 bit pattern generators A value of 1 means every bit is used and gives maximum resolution but maybe unnecessarily slow Alternatively the scan resolution can also be set directly in nanometres using the RESN parameter The scan resolution CRESN sets the distance in nanometres between adjacent points during the coarse scan but only to the nearest PG bit This ensures similar signal averaging and mark locate times for different maximum fieldsizes The coarse search limit CSLIM directly limits the coarse search length CSLEN for rectangular marks but defines the limit of CSOFF CSLEN for crosses This should not be set larger than necessary to avoid long mark locate times For example a coarse search limit of 50 um is recommended for the mark locate to initialise the holder on a VB5 which refl
349. rsion com resync com There is usually a symbol SPELL defined to do this VB OPER spell OpenVMS tips OpenVMS is designed as a multitasking multi user operating system capable of supporting thousands of simultaneous users reliably Each user is allocated a portion of the systems resources as decided by the account set up by the System Manager For Vectorbeam use an entry level desktop workstation with limited but normally adequate resources is used which is licensed for a single user However to allow for System Management use two users are normally allowed on the system at any one time However the user on the local terminal can log in to as many accounts as he requires This system works well provided applications are allowed to complete their execution cycles There are always cases when a program can crash When this happens there is no clean exit and system resources will not always be returned to the account holder Typical situations when a program can crash are attempts to access a file opened for write from a second window For example CVIEW opening a file which is already in use for writing a wafer Good programming should avoid Such errors but inevitably errors slip through the net for example when the access to a file used in several program codes is set to write which is the default Recommended tasks to avoid user hang ups The users environment always includes a Session Manager window This may be iconised and is recogn
350. rter users 1 In order to transfer the dose distribution from the converter to the VB via the pattern file it is necessary to set the number of dose classes to 1 select automatic classification and leave reserve default class unchecked before starting the formatter These settings are in the Dose class settings dialogue under process dose classification change 2 In order to transfer the dose distribution from the converter to the VB via a clock file set the number of dose classes to the required number of dose classes i e greater than 1 Create a clockfile by looking up the doses of each class in Caprox 3 The clocks in Caprox are numbered starting from 1 but the clocks in Emma are numbered starting from 0 This needs to be taken into account when transferring the dose of each class as displayed in Caprox into a clockfile for the VB 4 The maximum number of dose classes is 256 Part Number 878275 Vectorbeam Operator Manual 13 3 6 13 3 7 13 3 7 1 13 3 7 2 Page 125 Example clockfile A typical clockfile to set up the clocks for a particular pattern is given below In order for the file to be executed automatically by WLVD COM see chapter Exposing a substrate it should have the same name as the pattern but with the extension clk and be in the same directory as the pattern QSET DOSE REL 0 1 2 QSET DOSE REL 1 1 23 QSET DOSE REL 2 1 23 QSET DOSE REL 3 1 24 QSET DOSE REL
351. rture A lens before the beam blanking plates creates a focal point cross over of the electron beam in the electrical centre of the plates With this configuration the position of the spot in the substrate plane is independent of the voltage between the beam blanker plates up to the point when the beam is turned off This prevents the occurrence of blanking tails in the resist image as the beam is turned off or on This technique is known as conjugate plane blanking Fine focus coil Within the final lens a fine focus coil is used to e accurately focus the beam on the substrate static e correct focus for height variations static e enable high speed focus corrections for the deflected beam dynamic Stigmator coil A stigmator coil assembly double quadruple is used to compensate any astigmatism of the beam The astigmatism of the beam is compensated by diagonal and axial stigmator coils These coils are used for e astatic correction optimum for non deflected beam e acorrection for the deflected beam dynamic stigmator correction Beam deflection The deflection unit assembly is located between the condenser lenses and the final lens This consists of sets of magnetic coils which create deflection fields transverse to the beam axis in orthogonal pairs X and Y axis deflection In order to cover a sizeable area of the substrate at high speed whilst maintaining high accuracy the assembly has two separate magnetic coi
352. s Check pattern limits are less than the stage travel range and reposition exposure so that it all falls within the stage range BADTIME Unable to get system time date BLKTRNERR Block transfer failed Check ethernet link to PG Reboot PG CALNOUPFORCE Not in Absolute mode NOUPDATE DIAG options forced Switch to absolute mode if update required CANTREEXP_NOPAT Cannot reexpose as no pattern yet exposed Don t use reexpose qualifier CANTREEXP_PRTFLD Cannot reexpose as pattern has part fields Don t use reexpose qualifier CASSUNDEF Cassette undefined Issue qdisplay air read command CLKADJFAIL Clock adjustment failed Check VRU current clocks doses and resist sensitivity parameters Retry CLKADJNEG Invalid unset nominal dose Set valid nominal dose CMDABORT Command Aborted Correct for first error reported CMDNCOMPLETE Command not completed Correct for first error reported COLRESTORE Unable to restore column settings Retry COLUPDATE Unable to update column settings Retry Check lens drivers Check gun C1 operating conditions for FEG CORBIG Calculated corrections within 1096 of maximum Investigate why corrections are so large Maybe no problem CORTOOBIG Calculated corrections greater than maximum Possible error in specified positions or observed positions or the substrate rotation is too large Possible gross substrate errors CRTDWLERR Correction down load error May
353. s 6 25 The DACS are both 14 bit As the subfield deflection DAC is 14 bits 1 Isb on the subfield scaling DAC has the same weighting as 1 Isb on the mainfield scaling DAC and the software sets the values appropriately This means that the smallest change of the maximum fieldsize is about 0 15 nm or 0 5 ppm For a 20 bit mainfield the resolution of the DAC is however only 0 25 that of the PG resolution Limits The maximum range of subfield scaling around the nominal that the hardware can apply is 3 125 However due to the hardware arrangement this is the total scaling which includes contributions from calibrations as well as direct write scalings The hardware uses separate corrections DACs for the scaling and for the rotation for both the X direction and the Y direction Corrections are updated on a point by point basis during exposure of a shape within a subfield Scaling in X is carried out by adjusting the X shift All scaling corrections are fed through digital multipliers and summed before being applied to the scaling correction DAC ona point by point basis This means the subfield scaling corrections DAC must apply both the on axis and off axis scaling and any additional scaling required through the direct write mapping If the on and off axis scaling values are large then there is little range left for additional direct write mappings Grating generator A grating generator has been implemented in the 18 bit and 20 bi
354. s Focus Value is drawn as shown below Focus waveform centred on 0 103699 FS Focus Score FV Focus value Scanning over focus range 0 303699 to 0 096301 FS 24346 000 FV 0 304 FS 45807 000 FV 0 104 FS 23443 000 FV 0 096 Best focus 0 105759 Iteration 1 Prev val 0 103699 New val 0 105759 Focus score is used to determine best focus The higher the focus score value the better the focus The idea is to achieve a good looking peak through selection of marks and parameters Setting G_Debug 1 will show the accuracy of focus and stigmation values achieved during each iteration This can be used after setting the parameters and tweaking them for optimum repeatability and speed NOTE When finished make certain all debug modes are disabled by setting them back to zero e g G_Focus 0 G_Debug 0 Correct fine focus value on datum for conjugate blanking The fine focus value after focusing on the datum mark must be correct to maintain conjugate blanking and should be Height meter reading x focus for height coefficient 0 012 Equation 17 1 The focus for height coefficient is a negative number and can be read from the Sensitivity panel option in the Display menu in the Emma status window On 18 bit and 20 bit systems it can also be displayed by typing VB_OPER gt adisplay
355. s a warning buzzer If an error has occurred then PICS will indicate where the error took place by flashing the relevant LED while sounding the buzzer If a warning e g it is taking longer than expected to pump to required pressure has occurred then the LED will flash but the buzzer will not sound Part Number 878275 Vectorbeam Operator Manual Page 26 3 7 1 VB5 FEG vacuum control panel Figure 3 1 VB5 TFE vacuum control panel 3 7 2 VB5 mimic panel key e LEDs above the buttons are lit when the button has been activated e LED DOOR is lit when the loading chamber door is properly closed e LED GUN is lit when the emission chamber is properly closed e V1 to V18 are lit when the associated valve is open as seen by the valve detector e LEDs for rotary pumps PVP are lit when the associated PVP is working e LEDs for turbo pumps are lit when the associated pump is running Part Number 878275 Vectorbeam Operator Manual Page 27 LEDs for ion pumps IGP1 4 are lit when the associated pump is working G1 G3 G4 G5 are lit when the vacuum in the related areas is better than the required level LEDs LOADER are lit to show whether the loader arm is extended or retracted LEDs COUPL are lit to show whether the loader arm is coupled Y or uncoupled N LED UPS is lit when there is fault with the uninterruptable power supply LEDs UP DOWN and STOP are lit to show whether the cassette is moving up or down
356. s and positions must always be supplied in pairs 1 pair will fill the entire height map with that one height If 2 pairs are given a plane will be fitted to them in such a way that it has zero slope perpendicular to the line joining the points If 3 pairs are given a plane is fitted If 4 pairs are given a 4 term surface is fitted Height readings taken from the mini map far away from the area defined by the position identifiers will have been produced by extrapolation and are likely to be inaccurate For example given four position identifiers p1 p2 p3 p4 which define the area within which a mini height map is to be used qset height real MUST be in real time mode mvsp p1 seth h1 mvsp p2 seth h2 mvsp p3 seth h3 mvsp p4 seth h4 qcal height minimap hid h1 h2 h3 h4 pos p1 p2 p3 p4 qset height map on Select mapped mode However it would be more efficient to acquire the heights whilst acquiring the Part Number 878275 Vectorbeam Operator Manual Page 85 mark positions as that will save driving the stage around the four marks a second time 10 13 3 Mini height map use with standard layout parameter file template The standard layout parameter file template approach to exposing substrates described in the chapters Exposing a substrate and Direct write alignment supports the use of mini height maps The name of the mark definition for the die by die alignment marks symbol MARKER should
357. s are changed on the machine Vectorbeam Operator Manual 13 13 1 1 13 1 2 13 1 3 13 1 4 Page 117 Exposure dose The exposure dose is controlled by defining the stepping frequency of the pattern generator between exposure grid points The pattern generator contains a dose controller to control the scanning rate of exposure Usually however the operator wishes to define an appropriate exposure dose for the resist being used rather than the frequency Therefore the required frequency is calculated by Emma when the qadjust clock command is issued using the formulae given below Area dose The exposure dose is most often specified is an area dose This is when the pattern contains shapes with X and Y dimensions greater than the spot size and in this case the dose and the frequency are related by Dose Current __4 Equation 13 1 Frequency x ExposureGrid Line dose A line dose is used when the pattern contains only shapes with either the X or Y dimension greater than the spot size and the other dimension equal to the spot size e g single pass lines In this case the dose and the frequency are related by Dose Current Equation 13 2 Frequency x ExposureGrid Point dose A point dose is used when the pattern contains only shapes with both the X and Y dimensions equal to the spot size e g dots The dose and the frequency are related by Dose Current EUN Equation 13 3 There is cu
358. s of the stage movement are 153 mm x 165 mm for VB6 systems and 127 mm x 127 mm for VB5 systems A laser interferometer system measures the X Y position of the stage accurately and provides the reference X Y co ordinate for the system This information together with the desired position supplied by the job control is used in a servo loop to bring the stage accurately to the desired position As the substrate position cannot be controlled to the required precision by stage movements alone the difference between final mechanical position and the desired position is corrected via signal feedback to the main beam deflection Part Number 878275 Vectorbeam Operator Manual 1 2 2 1 2 2 1 1 2 2 2 Page 7 system This is known as Beam Error Feedback BEF This function ensures that the beam is constantly tracking the required substrate location even in the presence of minor vibration or mechanical drifts The stage is driven by air cooled motors within the main vacuum chamber and these are protected by several interlocks including loss of cooling air and end of travel range Substrate load system Loading of the substrate onto the stage takes place using an intermediate carrier known as a holder or chuck For systems with the single chuck or 10 chuck airlock handling option the substrate is first placed manually in the holder a range of different sized holders can be provided to suit standard size wafers and photo plates as well as
359. s or even the lifetime of the source A spot table calibration is carried out when a new electron source is fitted and at regular intervals if the existing one becomes inaccurate because the source characteristics change with age A spot table calibration is required for each final aperture size The spot table is used as a look up table for QSET SPOT and QSET BEAM CURRENT commands The C1 and C2 lens settings that correspond to the spot size or the beam current requested are then set See Sections Beam diameter adjustment and Beam current adjustment Demagnification table Nominal tables of values for C1 and C2 are given in the document Default demagnification Tables for LaBg and Schottky Columns part number 892894 These should be used as a starting point for manual trimming The document Manual procedures for obtaining demagnification tables on the Vectorbeam part number 892878 should be referred to Spot table The makespot utility described in document Manual procedures for obtaining demagnification tables on the Vectorbeam part number 892878 can be used to generate a manual spot table Alternatively the qcal spot command can be used Before running the QCAL SPOT command 1 The current demagnification table must be valid 2 Make sure that the mark position FM 10um octagon and the Faraday cup position FC are defined before running the jobfile This is usually done by the holder initialisation sequence 3 The mark
360. s that may be found on the Vectorbeam In this manual the following kinds of notations are used for safety related warning statements D If this hazard is not avoided there is a possibility of death or injury WARNING resulting moderate injury may result It may also indicate that a possibility of damage to the equipment may result If this hazard is not avoided there is a possibility that light or CAUTION If any warning labels become difficult to read contact your service representative to obtain replacements Labels must be replaced if they become damaged peeled or illegible See section Labels on all Vectorbeam machines in Chapter Vectorbeam series systems for further explanatory information on safety labels Part Number 878275 Vectorbeam Operator Manual 1 1 Page 5 Vectorbeam Series systems The Vectorbeam VB is typically used to expose integrated circuit pattern data on a substrate The substrates generally fall into two categories 1 A silicon or GaAs wafer The circuit pattern may be written on an unpatterned wafer or on a wafer with previously exposed and processed layers Direct Writing Operation The circuit pattern is transferred directly on to the wafer surface 2 A mask or reticle After the mask making operation the mask or reticle is used in a second exposure tool e g in a wafer stepper to replicate the circuit pattern on wafers using light or X rays Substrates are coated with
361. s the main field scaling and the holder initialisation sequence must be run This scheme allows any field corrections column set up and magnetic map to be used with the same basic machine set up Thus it is only necessary to determine each set up once rather than several times in all the different combinations Also there should not be any confusion as to which database contains valid machine set ups By saving only partial databases with the relevant information as described below it will not be possible to overwrite any other database parameters when these partial databases are reloaded independent of whether the qualifiers are used or not Basic database generation The set ups that must be performed to obtain a valid basic database include 1 the stage load position vital to avoid possible damage during holder transfer 2 the height correction coefficient calibrations 3 the detector calibrations 4 the main field and sub field sensitivities 5 the pivot points 6 the shape to shape settling delay 7 the stage mapping coefficients for absolute mode 8 the stigmator balance 9 the mark locate parameters 10 the demagnification table optional 11 the spot table optional 12 the stepper lens correction optional Vectorbeam Operator Manual 12 5 3 12 5 4 12 5 5 Page 113 13 the magnetic map The magnetic map calibration may be part of the basic database only if certain combinations of holders are n
362. saved beams to be used and measure beam current as described in the Section Beam current measurement Measure the beam current uniformity Jobcal check 1 Run jobcal and note any warnings given by the program 2 Check the noise in the calibration values as described in the Section Jobcal Main field distortions check In order to verify that the main field distortions are valid use the main field distortion calibration command with noupdate VB OPER CMFI DIST GRID 8 NOALIGN COVER NOUPDATE The accuracy required will depend on the application Repeating fullcal should improve the accuracy to better than 0 02 Part Number 878275 Vectorbeam Operator Manual Page 223 19 4 3 5 Autostitch overall calibration check Use the measure autostitch routine on marks on a substrate see Section Autostitch in Chapter Diagnostics This tests whether the mainfield subfield BEF field corrections and height corrections are OK 19 4 3 6 Progress until exposure check When the exposure has been started see next Section monitor the progress until the actual pattern exposure begins 19 5 Expose finally When all the preparation has been done type VB OPER Gfilename com The correct sequence of operations to expose the layout will be carried out assuming that the parameters are correct by the file wlvd com 19 5 1 Interrupting exposure 19 5 1 1 Pause The operations can be paused by pressing pause in the
363. scribes the level of training required to perform operation and service tasks The user is responsible for analysing the risks of the work not described in these manuals Allowed Operation Training required Operation of the tool as described in Vistec Course Operator Training this Operator s Manual Limited service operations and routine Vistec Course based on 893116 maintenance as covered in 893116 manual Full Service operations Vistec service engineer training includes hands on based on 893117 manual Table of Training Requirements for various operations This section lists the main topics that are covered by this manual for reference only e Machine start up from cold e Machine shut down e Vacuum system start up and shutdown e Bakeout e Routine maintenance covers water baths vacuum pumps gun cathode replacement Maintenance of safety features Safety features require no regular preventative maintenance to maintain their effectiveness However checks are recommended at intervals to ensure that no faults have arisen These are detailed in the Routine Maintenance manual referred to in 893116 or the Service Manual Part number 878275 Vectorbeam Series Vectorbeam Operator Manual Page 268 26 Diagnostics Various jobfiles have been written to diagnose problems with the machine and these are described in the manual Acceptance and Operator Jobfiles part number 892777 Part
364. see perpe edit a 12 UEM Dmm 12 WG WB Gicateicactonssesnstesech A E E tL SELL Sd 12 Compressed a SUBDIY aci rp piece d Uu ot ode iHe Va Fia bata APP eR dines 13 Meaning of labels found on machine components sss 13 1 10 1 Labels on all Vectorbeam machines seemR m 13 1 10 2 Labels only found on machines with wafer loading automation 15 do ul LT 2 1 2 2 2 3 2 4 Bear production QUII usa ioecres cui vce aei rird naa EUR AN RR Lb ds 18 eedem RESI UOS D S UTERE 19 vH EE INESSE EE EUIS 19 22 JADORDUIBS sicot cobi n a bestes uehrderpa bes eue a 19 2 259 lLensesTor TEE QUIN sic ta os rn is e kx ees eie ei pier ters ietasix est poised 19 2 24 BeanmiblahkiliQgs scheitert em d sie heap ka esce aueh crg e Pese RID PUUR IRAE 19 2 2 5 FNE TOCS COl s esee dein et tree R 20 22 0 SUGMALON CO cass cob ed D A EA 20 BOO OBTBO Mice odas tide CU aed dtd eftt dua n Fa FAX De ua UE CEU dni e duda 20 C AGING node eain rex de aed cen A dal en e Db nad en eet c oap renis pects e treme 21 Yy ACUI Sy SEIN PANIER 5 CR 3 2 3 3 3 4 3 5 3 6 3 7 Conversion of different pressure UMS iss seessccci scree tacente pe etra senecta eer dati deed aes 23 P uncibles er Bpalt atio oo booten dia datura dora ood iris a adc bc o eb Bead 23 TIROL BS CI sessed devia ixi Elo Se DUO Rod repo lot Do Ri dn 24 NES VACHUNT SYS IOs coi d etr tw a d eo d pL pis 24 VBG vac
365. should not be contaminated and should have a small mark slope to ensure accurate spot size measurements 4 Make sure that the correct aperture size has been defined Part Number 878275 Vectorbeam Operator Manual 15 9 15 9 1 15 9 1 1 15 9 1 1 1 15 9 1 1 2 Page 139 VB_OPER gt QSET APERTURE 800 A jobfile similar to the following should be created and run in order to carry out the calibration Spot table calibration job file QSET MODE FAB MVSP FM LOCM FM FM POSM FM Adjust the focus and stigmation on the 10 uim octagon mark The fine focus should be adjusted to be about 0 0 This command also sets the parameters to be used throughout the calibration QCAL STIG FM FILT 16 POINTS 8 DACPOS 5 EDGE 5 ACC 0 01 Measure the beam diameter on the 10 um octagon mark This command sets the parameters to be used throughout the calibration LOCM FM FM POSM FM QDISP DIA FM EDGE 5 SQUARE FILT 16 LINESCANS 1 POINTS 8 Calibrate the spot table QCAL SPOT FM FM FC FM FM TABLE 1 DIAG POINTS 40 Calibrate the Mark Slope QCAL SLOPE DP DP FC FM FM SFAB RESTORE FINE MVSP FM LOCM FM FM POSM FM Stage mapping modes There are three stage modes natural absolute and machine The stage co ordinate mapping modes ABSOLUTE mode and MACHINE mode may be used to correct the stage for positional errors and various distortions or to match the stage characteristics of a machine on
366. shown in Tables 1 2 and 3 and correspond to standard Vectorbeam pattern generator fieldsize capabilities The VB field parameter is used to calculate the maximum VB Resolution and thereby automatically ensure that VB Resolution is set correctly There are many variations of Vectorbeam pattern generator hardware with different fieldsize ranges and the VB field parameter needs to be set to match the particular Vectorbeam The following command can be used to change the VB Field VB Field value Usually a Cats installation only generates pattern files for one Vectorbeam so this setup can be done in the Cats startup file Once this has been defined for each kV as shown below the format can be changed and the correct value will always be present Format VB20 VB Field value Format VB50 VB Field value Format VB100 VB Field value VB Max field The VB Max field parameter is only used for the VB20 VB50 and VB100 formats where it is not used for any calculation but is simply a check on the value of VB Field When the VB Field is set larger than VB Max field a warning is given but the VB field will be accepted Therefore the VB Max field is usually set to the same value as the VB Field The VB Max field parameter is not the same as the Maximum Fieldsize parameter on the Vectorbeam Resolution The resolution defines the grid to which all shape vertices are snapped in Cats For the format VB the Resolution corresponds d
367. square connecting the positions of the centre of the calibration mark when calibrating the mainfield When calibrating the centre of the mark cannot be placed at the corner of the deflection field as the outside edges of the mark would be outside the deflection field and therefore the mark could not be located The mark is typically placed within half the mark s size of a corner plus a few microns to allow for scanning over the edge For example using the 10 um octagon to calibrate with a maximum fieldsize of 1024 um and a fine scan length of 3 um the central 1011 um 21024 13 um could be accurately calibrated The cover qualifier for the calibration would have to be set to 1011 1024 0 987 2 The blocksize for a multi block pattern is usually chosen to be as large as possible to maximise the throughput 3 The blocksize for a multi block pattern may be set to be considerably smaller than the maximum in order to use only the centre of the field where the resolution and accuracy is better 4 The blocksize for a multi block pattern may be chosen so that periodicity of the blocks matches some periodic structure in the pattern This can allow the block boundaries to be placed in non critical areas of the pattern 5 Due to the normal discard of shapes with height or width smaller than the beamstep the blocksize must be an integer number of Part Number 878275 Vectorbeam Operator Manual Page 206 beamsteps VRU pattern generator
368. sseeeee 148 16 515 MBBL ais cts ceuatn consi dantisiaqanenseeancedscsyaniagecadutiansavsaupeeusedss eoo e MA sv dune 148 16 52 MBOL Gies ois eese ue oe a eS escis E ARSi 149 16 5 3 Stopping a substrate transfer iii eiii a 149 16 5 4 Holder initialisation sequence 149 16 5 5 dHolderiparamleters eiii ipii riesce tesi Eso Oei eSee 150 16 5 6 Setting up the holder parameters sssssseesess 151 16 5 7 Datum target layout sciiicet 152 16 5 8 Datum mark contamination 154 16 5 9 Holder position table CaSpoS srice 155 17 Job specific machine set ups eeeeeeeererereennee L57 Wen C Oe E UD OESTE DD 157 17 1 1 Coarse gun alignment HR and UHR final lens 157 24 1 2 dines aligmnnebibuisce iioc etai reii cacet io n 157 UR EM eut RR 160 17 1 4 Setups affected by gun alignment sessssssessss 160 17 20 Video dal and DACKOM set US uiuunt oa ELLE EE I dde cadre Ea a dank das 160 17 2 4 Manual gain and backoff adjustment ssssssssss 161 17 2 2 Automatic gain and backoff adjustment sessssss 161 17 2 3 Video Calibration for UHR machines ssssssesss 162 WO Tij 3Bo B ecce added aieo bd elut ena eh Ted AMA etin pae eee a MS 163 17 91 Tuanalaperture Seleolon ire mto n eR d eR ERES 163 17 92 JIinalaperture allgmmelibss iie dai eoe RR hib t d
369. stigmation corrections and the magnetic map remain valid The procedures is as follows 1 Move to the calibration mark FM VB_OPER gt mvsp fm 2 Select SEM mode 3 Move a corner of the mark to the centre of the screen using the joystick Figure 7 21 4 Set the magnification to maximum keeping the mark corner centred 5 Display the height meter reading VB_OPER gt dhgt tab 7 6 Multiply the height reading e g 7 5 um by the focus for height coefficient and reverse the sign of the result The focus for height coefficient can seen in the Sensitivity Panel under the Display menu of the Emma status window e g 7 5 x 0 0081 0 061 reversing the sign of this result gives 0 061 Set the fine focus to this value VB OPER sfoc 0 061 7 If the mark is a long way out of focus bring lens 3 to the approximately correct value based on the sharpness of the mark in the SEM image This is best done by displaying the current value by typing VB OPER dIn3 example result 20 407 and then varying the value until the best focus is obtained by typing VB OPER sln3 value example value 0 410 8 Switch the beam blanking operation in Cabinet B to Test mode 9 Rotate the potentiometer knob RV8 on the front of the beam blanking unit in Cabinet B fully anticlockwise and then fully clockwise The beam will be fully blanked at one end and fully un blanked at the other Confirm this from the SEM image it might be necessary to first s
370. stment mechanism A program has been written to automate this procedure Move to an alignment mark and type VB OPER QG VBSSEQ ALIGN APERTURE COM 15 FM FM 0 005 The program measures the mark position at two Final Lens currents and displays the difference in the positions The aperture positioning will be done automatically until the difference is smaller than the limit defined on the command line 15 nm Part Number 878275 Vectorbeam Operator Manual 17 3 3 17 3 3 1 17 3 3 2 17 3 4 Page 164 Defining the aperture adjustment mechanism positions Manual aperture adjustment mechanism On HR machines the aperture selected can be read from a scale next to the left knob On UHR machines the aperture selected and the position can be read from the micrometers Automatic aperture adjustment mechanism 1 Using a combination of the move commands i e VB OPERo aperture move abs position x position y and VB OPERo aperture move rel distance x and distance y Move the aperture blade to approximately centre an aperture The distance between apertures is about 1600 and a move in the positive direction corresponds to moving from a high number to a low number aperture 2 Align the aperture as described above 3 Store the current position as a present using the command VB OPER aperture set number between 1 and 16 4 The current position and status can be displayed using VB_OPER gt aperture display 5 The
371. stortions off VB_OPER gt SCOR OFF MAIN 3 Run jobcal VB_OPER gt JOBCAL 4 Switch the main field distortions on VB_OPER gt SCOR ON ALL Part Number 878275 Vectorbeam Operator Manual Page 185 Calibrate the distortions using the currently set cover The maximum distortions on the second iteration should be small VB_OPER gt QCAL MAIN FM DIST NOALIGN GRID 8 COVER F TRNLMN VB_COVER ITER 2 DIAG Run jobcal VB_OPER gt JOBCAL Check the corrections as described in the section Checking the deflection field corrections above When satisfactory results have been obtained save the distortions to the appropriate database see Chapter Databases 18 8 3 Subfield distortion 1 2 Align final aperture Run jobcal VB_OPER gt JOBCAL Calibrate the distortions using the currently set cover The maximum distortions on the second iteration should be small VB_OPER gt QCAL TRAP FM DIST NOALIGN GRID 8 COVER F TRNLMN VB_COVER ITER 1 DIAG Run jobcal VB_OPER gt JOBCAL Check the corrections as described in the section Checking the deflection field corrections above When satisfactory results have been obtained save the subfield distortions to the appropriate database see Chapter Databases 18 8 4 Beam error feedback distortion Do not use until further notice Zero any existing corrections using VB_OPER gt QCAL BEF DIST INIT LOAD 18
372. sts Job file creation Although software and job files are supplied to enable many common operations to be carried out the user is expected to be able to create job files if necessary for their particular applications Editing can be carried out in the Job Control Window but is usually done in a separate DECTerm so that the machine can be operated at the same time To create a spare DECTerm window select with the mouse left button Application in Session Manager Menu at the top left corner of the Workstation screen keeping the mouse button pressed down slide the arrow down to highlight DECTerm then release the mouse button After a few seconds a new DECTerm window will appear File editing can be carried out Part Number 878275 Vectorbeam Operator Manual Page 62 8 5 8 6 8 7 8 7 1 8 7 2 8 7 3 while the machine is executing commands but not the files the machine is using Refer the document Job file standards part number 892815 for details of the rules and tips for writing job files Job file examples Job files described in this manual in the Chapters Exposing a substrate and Direct write alignment can be used as examples These job files define and carry out the exposure of a matrix of dies Running jobfiles At the vb_oper prompt type followed by the name of the jobfile VB_OPER gt my_jobfile com Stopping jobfiles During DCL command execution 1 In the DECTerm running the j
373. t G BlankTimeMax The minimum time is that for a mainfield move of 1 subfield The maximum time is that for a mainfield move of the maximum number of subfields that are currently in use All moves inbetween are calculated from a linear relationship between these two points These values can be changed directly by entering gt G_BlankTimeMin value in us gt G_BlankTimeMax value in us followed by gt setBlankDelayTab The mainfield settling values will be reset to any values specified in the EMMA CTRL DCP_CONFIG VW file for 16 bit pattern generators or EMMA CTRL WFDCP_CONFIG VW for 18 bit and 20 bit pattern gnerators whenever the DCP is rebooted These values are set by the same statements as above in the relevent file G_BlankTimeMin value in us G BlankTimeMax value in us setBlankDelayTab Subfield deflector settling time 16 bit pattern generator The current subfield settling delay is displayed in the Display Database Strategy panel Figure 7 14 as SS Settling This value may be changed using the command VB_OPER gt aset ss value in ns Part Number 878275 Vectorbeam Operator Manual 24 2 2 2 24 2 3 24 2 3 1 24 2 3 2 Page 255 18 and 20 bit pattern generators The subfield settling delay is hard coded as 2 us for shape to shape movements of 1 4 of the subfield size or less and 8 us for shape to shape movements of more than 4 of the subfield size Stage settling time The various param
374. t Number 878275 Vectorbeam Operator Manual Page 281 caution is required when recovering the situation and a Vistec engineer should be consulted NOHOLDINAL No holder in airlock Check the source position specified Check the occupancy using the qdisplay air read command If it is known that a holder IS present then extreme caution is required when recovering the situation and a Vistec engineer should be consulted NOHOLDINCP No holder in crane pouch Check the source position specified Check the occupancy using the qdisplay air read command If it is known that a holder IS present then extreme caution is required when recovering the situation and a Vistec engineer should be consulted NOHOLDINFP No holder in front pouch Check the source position specified Check the occupancy using the qdisplay air read command If it is known that a holder IS present then extreme caution is required when recovering the situation and a Vistec engineer should be consulted NOHOLDSTAGE No holder on stage Check the source position specified Check the occupancy using the qdisplay air read command If it is known that a holder IS present then extreme caution is required when recovering the situation and a Vistec engineer should be consulted NOMRKFND No mark found in the data base Use mlis to show the marks in the database Define mark if needed NONEWHOLDER Could not create a new holder definition NOT PAT
375. t and the holder sequence and then try jobcal If the same error occurs during jobcal then enter the qcal trap init load Part Number 878275 Vectorbeam Operator Manual 27 6 27 6 1 Page 271 command again before moving on to point 3 below 3 In SEM mode observe the position of the datum mark when moving the subfield in X and Y e g VB OPER strp 5 0 VB OPER strp 0 0 VB OPER strp 0 5 VB OPER strp 0 0 There should a movement of 5 um in both X and Y of about half the size of FM If not then there is probably a fault with the subfield power supply or cabling Errors and warnings message list with suggested recoveries Emma will report information warnings and error messages in a text form for most of the normal Emma command errors or machine hardware and software problems The DCL command interpreter will however deal with all other commands The information warning or error will be typed below the command issued in the job control DECterm and a list of all such messages is given in the Sections Emma information messages and Emma error messages The logical estatus will also be set simultaneously to an error number and list of such error numbers is given in the Section Emma error numbers Jobfile automatic recovery The estatus number can be read from a jobfile and the appropriate recovery action taken automatically in some cases Note 1 ON ERROR only works for DCL commands and not for Emma commands Include a
376. t meter offset for a 500 um 10 6 3 10 6 4 10 6 4 1 10 6 4 1 1 10 6 4 1 2 blocksize For UHR systems there will be a rotation change of 7 nm across the 800 um field in addition to the scale change Sign of offset The heightmeter measures the separation of the substrate from the final lens and if this measurement is increased by setting a positive offset the deflection angle will be reduced to compensate If the substrate were truly further away the fieldsize would remain the same However if an offset is set on table 7 which is reserved for the datum jobcal will introduce this offset into the calibrations and all exposures using any other table will be affected in an opposite sense Setting the height meter offset The height meter offset can be set either on the MUP or by using the gset reg command Height meter offsets The height offset can be set by logging into the MUP if using a 16 bit pattern generator or the DCP if using an 18 bit or 20 bit pattern generator and setting numbers in a table of offsets for each heightmeter table provided for this purpose 16 bit pattern generator VB SUPER rlogin MUP Vectorbeam Pattern Generator Master login pgdiag Password gt 18 bit and 20 bit pattern generators VB_SUPER gt rlogin DCP Vectorbeam Pattern Generator DCP Part Number 878275 Vectorbeam Operator Manual Page 78 login pgdiag Password gt 10 6 4 1 3 16 bit 18 bit and 20 bit
377. t open file for ADC plot Check protection setup on terminal AGAFAIL AGA Failed Check video levels in SEM mode on suitable target AGASIGHIGH AGA Signal Too High Check video levels in SEM mode on suitable target AGASIGLOW AGA Signal Too Low Check video levels in SEM mode on suitable target AXIALDRIVE Axial stig drive beyond range Check parameters and column adjustment BADDBSECN Invalid database section Use alternative database or set up appropriate machine parameters BADHOLDERINPUT Incorrect holder ID specified Specify correct holder ID BADLENSGRID qCal Lens Grid size is too large Specify smaller grid BADLENSGRY qCal Lens Grid size in Y is too large Specify smaller grid in Y BADLENSID qCal Lens lens map number is out of range Specify valid map number BADLENSORD Lens map order specified is out of range Specify valid map order BADLENSORGX qCal Lens Origin offset in X is out of range Specify valid X origin BADLENSORGY qCal Lens Origin offset in Y is out of range Specify valid Y origin BADLENSSEND Transmission of lens map coefficients to PG failed Reboot PG and retry Part Number 878275 Vectorbeam Operator Manual Page 273 BADLENSSTEP qCal Lens Y step size is negative or out of range Specify valid Y step size BADLMDATSND Transmission of lens map assignment to PG failed Reboot PG and retry BADSTAGEBLK Stage move block positions outside stage limit
378. t otherwise defined As clock 0 is also normally assigned a relative dose of 1 0 the resist sensitivity parameter set will define the dose which will be used for exposing the entire pattern Clocks A clock refers to one of the frequencies supplied by the dose controller Emma will allow up to 32 clocks with numbers from 0 to 31 to be defined anywhere between the maximum 256000 kHz and minimum 0 3 kHz limits of dose controller hardware The dose controller 2 board can actually provide any one of 65536 clocks which are spaced at regular intervals of frequency between the upper and lower limits If a pattern file has been produced by a converter and no clocks or doses were assigned the clock assignment for each shape will default to clock 0 Setting clocks The qset clock command accepts any frequency in the hardware range The frequencies are only set in the pattern generator when the qexpose pattern or qadjust clock commands are issued Adjusting clocks The gadjust clock command is used when defining area or line doses By typing the command VB OPER QADJUST CLOCKS the frequencies are calculated for clocks to which a dose has been assigned based on the beam current see below the doses and exposure grid VRU resolution The frequencies are set in the pattern generator Displaying clocks The qdisplay clock command displays the dose determining parameters and the frequencies as computed with these parameters The comm
379. t pattern generators to 1 Reduce the deflection settling overheads for the exposure of a grating as consecutive complete lines by improving on the method used by the Part Number 878275 Vectorbeam Operator Manual 24 9 1 Page 260 current 16 bit pattern generator using nosort exposure mode to expose consecutive complete lines 2 Provide a mechanism for fine tuning the tradeoff between settling overheads and subfield stitching accuracy These enhancements are possible by changing the order in which shapes after fracturing into subfields are exposed The term sub shapes is used to refer to those shapes produced by fracturing the original shape into subfields The order in which subfields are addressed affects the amount of subfield stitching error on a shape passing through adjacent subfields due to beam drift The order in which subfields are addressed also affects the settling time There are 3 different settling delays used by the PG when switching from one subfield to another 1 Short Used for jumps between adjacent subfields 2 Medium Used for jumps of 2 to 6 subfields 3 Long Used for jumps of more than 6 subfields Defining the grating The grating must be defined in terms of the linewidth lineheight pitch start position the number of lines with dimensions in micrometres consistent with units of sman and the relative dose for the lines The origin for the start x position and y position is the centre o
380. t processed Part Number 878275 Vectorbeam Operator Manual Page 240 DWA is similar to DW2 and DW5 is similar to DW3 except for the following When the machine only finds 0 1 or 2 alignment marks the mapping of the last die with 4 alignment marks is used together with an X Y shift found from the first mark of the current die to be located By using the mapping of last die with 4 alignment marks it is hoped that the scale rotation and keystone corrections will be approximately correct for the current die If no previous dies with 4 alignment marks have been found then the defined dies further on in the matrix will be aligned in turn until one with 4 marks is found and mapping coefficients can be generated The processing resumes at the first die If no dies with 4 marks are found then an error is reported no dies are processed OffsetX 0 25 The offset in mm in the x direction of the lower left corner of the lower left alignment mark from the centre of the die see Figure 20 4 OffsetY 0 25 The offset in mm in the y direction of the lower left corner of the lower left alignment mark from the centre of the die see Figure 20 4 MARKSIZEX 0 01 The X dimension in mm of the alignment mark This can be given as 0 00 if it is desired to set the offsetx to be the distance to the centre of the mark instead of the corner MARKSIZEY 0 01 The Y dimension in mm of the alignment mark This can be given as 0 00 if it
381. t the fine scans are confined to a region that does not encompass the diagonal edges Part Number 878275 Vectorbeam Operator Manual Page 100 FSLIM 8um MW tum MLEN 1 5um ANA A Fine Vertical Scans J E d Fine Horizontal Scans z o W 4um Figure 11 10 Diagram showing fine search scans for rectangular marks with edge 11 10 Part Number 878275 algorithm Cross locate algorithm The cross locate algorithm is separated into a raster search and a fine search locate fine i i e locate raster fine success I il fail gt Figure 11 11 Cross locate algorithm search flow diagram 11 10 1 Raster search A raster search is used as a coarse search for crosses The raster scans consist of a series of horizontal scans about the expected position displaced in the vertical direction by the coarse search offset CSOFF The raster scans expand out from the initial position by an amount equal to 80 of the cross height alternately above and below until the Coarse Search Limit CSLIM is reached or until a scanned line contains a video profile indicating the presence of one of the cross limbs When the horizontal raster scans succeed giving an x position of the cross centre a vertical confirmation scan is made The vertical confirmation scan is made at a distance CSOFF away from the cross centre along the expected Vectorbeam Operator Manu
382. tercept Long Ethernet message send failure Couldn t queue Ethernet message for output Couldn t pass message to Ethernet output Ethernet ifunit failure Host is not connected Missing ethernet packet during block transfer Already connected to a different host Already connected to the subsystem Couldn t queue Ethernet message for input Command rejected qEtherlP queue would overflow Ordinary Ethernet command received by a slave Command Interpreter process suspended No action On axis amp Standalone EHT errors Couldn t create qErrors queue Couldn t create message queue Couldn t send to message queue Couldn t spawn a task Couldn t spawn Emma command interpreter task Error creating semaphore Error getting semaphore Error allocating memory Error freeing memory Couldn t allocate memory for Emma reply packet Couldn t free memory for Emma reply packet This register is write only This register is read only This register does not exist Wrong data type for this register Data out of range Illegal function attempted Illegal data type Task suspended because of exception Cl cmnd receive from qEtherlP queue failed Error in Emma command interpreter Receive from gErrors queue failed Error in DMA write Requested VME DMA xfer over 256 byte boundary Can t set up mailboxes DMA xfer not on 4 byte boundary Can t understand vxWorks version number CCU Error in digital readback CCU Error in analogue readback CCU Miscellaneous
383. tern generator has a 16 bit output i e the maximum fieldsize is divided into 65536 exels The maximum fieldsize is split into 64 subfields in the X and Y directions 4096 in total and each subfield contains 1024 exels Pattern generator digital crate Part Number 878275 Vectorbeam Operator Manual Page 39 LOGAL ETHERHET 4 3 1 1 2 PGD CRATE B2 s TD L w INTERFACE TO 3UB FELD AHD MAIN FIELO Ta HEAD AHP Figure 4 2 16 bit pattern generator digital crate The pattern generator digital crate is shown in Figure 4 2 The pattern generator digital crate contains the Master Microprocessor MUP Flattening Instancing Sort Processors FISPS Shape to Line Converters SLCs and a Digital Correction Processor DCP Some systems may contain PowerPC microprocessor boards instead of a FISP and SLC pair 4 31111 MUP The MuP Master Microprocessor sequences and controls the operation of the Pattern Generator during pattern writing 4 3 1 1 2 1 FISP The FISPs Flattening Instancing Sort Processors take the structured and unstructured shapes from the CFPS Central Field Pattern Store on a field or part field basis They flatten structured shapes into individual shapes and generate an instance of each shape for each subfield that the shape appears in They sort the shapes into subfield order and output the shapes to the Shape to Line Converter The timing and destinations for the output data are determined by the MUP The subf
384. that if the number of shapes in a block exceeds a limit Writefile partitions the data into separate blocks all at the same stage position The need to do this was eliminated with the PowerPC s and therefore the default limit of 80 000 shapes should be set higher by setting the logical as follows VB_OPER gt define te max vb figures 10000000 Setting pattern file attributes on DEC computer The converter software may be running on a workstation not running VMS but an operating system such as NT The most straightforward way to transfer the binary FRE pattern file to the machine is using FTP in binary mode This will result in a file with 512 byte records FRE files Any FRE file on the VB should have fixed length 2048 byte records If this is not the case the file attributes can be changed using File Definition Language FDL under VMS The command is VB OPER exchange network fdl fre fdl transfer convert lt input file gt lt output file gt or VB OPER convert fdl fre fdl lt input file gt output file gt where FDL FRE is record format fixed size 2048 carriage control none VEP files After binary FTP to the machine the file format should be already correct with Part Number 878275 Vectorbeam Operator Manual 19 1 3 19 1 3 1 19 1 3 2 19 1 3 3 19 1 4 19 1 4 1 19 1 4 1 1 Page 199 fixed length records of 512 bytes Otherwise follow the procedure for FRE files above but substi
385. the beam current manually 1 Move to the focus mark VB_OPER gt MVSP FM 2 Switch the beam on enabling the SEM scanning VB_OPER gt SSEM 3 Using the joystick panel adjust the magnification to display the focus mark 4 Enable the automatic video gain VB_OPER gt vid_p 5 Display lens 1 setting lens 1 is the gun focus voltage VB OPER DLNI 6 Increase the value of lens 1 a little from that displayed to decrease the beam current and vice versa Be careful that the resulting focus voltage does not go Part Number 878275 Vectorbeam Operator Manual 17 8 17 8 1 17 8 2 Page 174 above the extractor voltage as this is likely to induce outgassing in the gun VB_OPER gt SLNI1 new value 7 Adjust the value of lens 2 to bring the focus mark into focus see section Focusing the beam and in particular Manual lens 2 adjustment 8 Measure the beam current and repeat the procedure until the desired current is reached If the beam current drops unexpectedly it may be necessary to adjust the gun alignment to bring the beam back onto the final aperture Save the settings if required see Section Databases Beam diameter adjustment Automatic The machine can adjust all the required lenses the gun alignment and the focus and stigmation to give the desired beam diameter based on the spot table calibration Type VB_OPER gt QSET DIAMETER spot size nm mark name mark pos gt BCM FC Manual The beam
386. the deflection coil Bath 5 controls the temperature of the Brooks Transport Module Compressed air supply The Vectorbeam system requires a compressed air supply e For the pneumatic system to operate the valves e For the vibration damping system This system mounted under the electron optical column reduces the effect of floor vibrations on column functions e For cooling the beam blanker the stage motors and the deflection coils When the column chamber or airlock is vented it is filled with dry clean nitrogen to prevent dust contamination and water vapour from entering the system Meaning of labels found on machine components This table shows the labels that may be found on machine components The meanings are explained next to the label and the precautions that an operator must take to avoid the hazard are also given where necessary Note that some detailed precautions that are only relevant to service engineers are not shown here as they are beyond the scope of this table Labels on all Vectorbeam machines Symbols that do not have a warning triangle are usually just for information not hazard warnings Warning risk of electric shock Consult manual for details Caution risk of danger Possible hazard during service Invisible laser radiation operations DO NOT STARE INTO BEAM refer to service manual OR VIEW DIRECTLY WITH OPTICAL INSTRUMENTS No hazard in normal operation Class
387. the fieldsize dependent main field distortions e g VB OPER dfile save file spec text dist cal This saves the distortion calibrations section of the database in a named VMS file 12 4 Database loading The calibrations may be restored either independently or together with the total database If the file is not a total database only the appropriate part of the database is loaded VB OPER qFILE LOAD file spec lt db section gt TOTAL The user is warned if certain elements of the file being loaded are incompatible with the present database such as the field size calibration sensitivities pivot points the aperture number and the EHT In this case the database will not be loaded unless the OVERRIDE qualifier is specified The qualifier TOTAL should be included when saving loading whole databases 12 5 Database management The following scheme is recommended although users are of course free to use different database management The scheme makes use of the fact that most of the database consists of machine set ups that do not change and a few machine set ups which are specific to the job The machine set ups which do not change from job to job are however likely to change if mechanical work is carried out on the machine The machine set ups that usually change from job to job are the field corrections The field corrections are the main field distortion calibrations sub field distortion calibrations
388. this is shown in Figure 19 12 Substrate Top level layout Centre of top level layout Second level layout Figure 19 12 Nested layouts Nested layouts can be used by defining the top level layout to call a second level layout at each position The important points to note are that e thecentre of the top level layout is defined as a stage coordinate as usual but the centre of the second level layout must be taken as the centre of each top level die This is done by using the option in the layout command to define the centre as relative e The workfile should be defined as none in the top level layout to avoid any offset in the second level due to pattern centring An example jobfile is shown below Top level layout parameter file The dose dosestep and operator in this top level file are not used for the exposures carried out by the second level jobfile ON CONTROL C THEN GOTO FINISH ALL ON CONTROL Y THEN GOTO FINISH ALL TYP FSENVIRONMENT PROCEDURE SJOB F ENVIRONMENT PROCEDURE holder 999 6 inch mask 999 dont init holder Part Number 878275 Vectorbeam Operator Manual 19 6 2 Page 230 calibrate calibrate workfile none No pattern file workfile_vru 12 5 nm beamstep vru res startdose 135 uc cm 2 20 KV dosestep 0 00 1 x overall operator 0 no update or beam 20kV 5NA 600um max 5 2
389. tion Switch to SEM mode Switch to the transmission detector with the appropriate level of attenuation MAX HIGH LOW MIN and with the bias on to give the maximum detector bandwidth see Chapter Transmission detector VB_OPER gt QSET DET TRI ATT BIAS Adjust gain and backoff so that the signal from the mark is a little below the maximum level and the signal from the background is a little above the zero level 10 Focus on the edge It may be necessary to adjust lens 2 if the fine focus range is not enough 11 Zoom in on particle on the knife edge in SEM mode and adjust the focus and stigmation manually as accurately as possible with the slider panel fully stretched 12 Move the stage a few microns parallel to edge to place the particle off axis 13 Type VB_OPER gt QDISPLAY DIA A0 LINELENGTH 2 EDGE 0 FILT 16 POINTSAMPLES 8 LINESCANS 1 and record the result Note that AO will need to be replaced by A90or A270 for a knife edge parallel to the X axis requiring scanning parallel to the Y axis 14 Switch back to photomultiplier when finished VB_OPER gt QSET DET PM Mark slope calibration The edge sharpness discussed above can be automatically measured by using the function qcalibrate slope This function measures the beam diameter at two different currents and by making use of the relationship below between beam current J and the measured beam diameter d sharpness can calculate
390. to from the stage Figure 6 2 below shows the orientation of the holder on the stage Main Chamber Laser height sensor Diode And direction of laser beam Figure 16 2 Diagram showing the orientation of the holder on the stage For ultimate pattern placement accuracy and low drift it may be necessary to wait for an hour or longer to allow the temperature of the holder and substrate to stabilise This time can be reduced by matching various temperatures See Section Temperature control and set up 16 5 1 Find out which positions are occupied by holders and whether the stage is in the Load Position type VB OPER QDISPLAY AIRLOCK The status will then be displayed in the Job Control window In order to automate operation with a jobfile a number of holder tracking logicals are defined by Emma which keep track of which positions are occupied with which holder see the Emma Command Set manual for details part number 878274 VB5 To transfer a substrate from one position to another enter in the job control window VB_OPER gt QSUBS LOAD lt N gt This moves the stage to the load position first and then transfers a substrate from the lt N gt th position in the airlock to the stage Part Number 878275 Vectorbeam Operator Manual 16 5 2 16 5 2 1 1 16 5 2 1 2 16 5 3 16 5 4 Page 149 The substrate is transferred back to the N th position in the airlock using VB_OPER gt QSUBS UNLOAD
391. tor signal FILTER should be adjusted so that the repeatability of the mark locate position is suitable 5 The locate method LOCATE should be set to either PIT EDGE or CROSS see Sections Pit locate algorithm Edge locate algorithm and Cross locate algorithm Part Number 878275 Vectorbeam Operator Manual Page 94 11 4 4 Choosing scan parameters 1 The oversampling OVSAM is typically set to 8 as long as the noise in the signal is low enough to give reproducible measured positions Reducing this value does not significantly reduce the mark locate time The value must be increased if the signal is noisy giving a large scatter in the measured mark position The number of lines LINES is typically 1 as long as the noise in the signal is low enough to allow detection when the coarse search is carried out The number of scans carried out on each edge during the fine search is PARA LINES giving greater noise immunity This means that the mark locate time can be minimised if coarse searching is carried out The number of lines must be increased only if coarse searching fails due to a noisy signal and OVSAM and FILT are already set to maximum The scan resolution SRES sets the number of PG deflection bits between adjacent points during the fine scan A value of 4 means every fourth bit is used A value of 1 means every bit is used and gives maximum resolution The typical value for 16 bit and 18 bit pattern generato
392. total range of about 12 5 of the field size This means that for a field size of 0 32768 mm it has a range of about 41 um The software however does not actually load all 18 bits of the main field scaling DAC The two least significant bits are initialised to zero and then the 18 bit DACs are used as 16 bit DACs The subfield has its own scaling DAC which is 12 bits As the subfield deflection DAC is 10 bits 1 Isb on the subfield scaling has the same weighting as 1 Isb on the main field This means that the smallest change of the maximum fieldsize is about 1 2 nm or 4 ppm Limits The maximum range of subfield scaling around the nominal that the hardware can apply and it is 12 5 However due to the hardware arrangement this is only if there are no other corrections to be applied This is because the hardware uses only 1 corrections DAC for the X direction and 1 for the Y direction Corrections are updated on a point by point basis during exposure of a shape within a subfield Scaling in X is carried out by adjusting the X shift and rotation in X is carried out by adjusting the Y shift All scaling and rotation corrections are fed through digital multipliers and summed before being applied to the correction DAC on a point by point basis This means the subfield corrections DAC must apply both the on axis and off axis scaling and rotation and any additional scaling and rotation required through the direct write mapping If the on a
393. trate mounting in holder then a valid reading close to the start position prior to the exposure is required Possible incorrect height meter table selected Slave Micro Modules Slave Command Interpreter Module 128 80 Reboot PG 0x10800001 More SupCl commands waiting Reboot PG 0x40800002 SupCI mailbox interrupt but no source found Reboot PG 0x40800003 Interprocessor comms semaphore Give error Reboot PG 0x40800004 Interprocessor comms semaphore Take error Reboot PG Slave SLC errors Module 129 81 0x40810001 SLC 1st opcode in field not SubField or EndOfField Possible faulty pattern Reboot PG 0x20810002 SLC Invalid command received Reboot PG 0x20810003 SLC Invalid VRU received Reboot PG 0x20810004 SLC Invalid invert strategy received Reboot PG 0x10810005 SLC Info Empty field received Reboot PG 0x40810006 SLC Output buffer overflowed Reboot PG 0x40810007 SLC Couldn t allocate linewriter output buffer Reboot PG Slave FISP errors Module 130 82 0x20820001 FISP Invalid command received Reboot PG 0x20820002 FISP Invalid VRU received Reboot PG 0x20820003 FISP Field too large data will swamp SLC Old software 0x40820004 FISP Division by zero in CadDig Possible invalid shape in pattern file Use smaller VRU Reboot PG 0x20820005 FISP End of last shape is after the end of buffer Reboot PG 0x20820006 FISP Transfer too long for i p buffer Reboot PG 0x20820007 FISP Array overflow field too large Reboot PG 0x4
394. trix of height readings is typed on screen after measurement by a jobfile The centre and extent of the map depend on the substrate parameters defined in the holder sequence Type VB_OPER gt VB ACCS acc_plateheight com For maximum accuracy all heights should be within 10 um and the tilt should be less than 1 um mm Height map mode During pattern exposure the height sensor may either be run in real time mode or in height mapped mode Real time mode In real time mode a height meter measurement is taken after every stage move This is the usual mode of operation as it is simpler than using a heightmap and usually gives the maximum accuracy To use in real time mode type VB_OPER gt QSET HEIGHT REALTIME Height map mode In height map mode a calculated value of the height is obtained from a previously calibrated height map This mode of operation is usually only used for special applications To use in height map mode type VB_OPER gt QSET HEIGHT HEIGHTMAP Height map calibration To create a height map ensure that there is a holder on the stage and the appropriate marks are defined The following command file will create a height map for the current holder The height map will need to be tailored to the individual requirement for each size of substrate Part Number 878275 Vectorbeam Operator Manual 10 13 10 13 1 Page 83 SJOB HEIGHTMAP COM Set the height sensor to real time mode QSET HEIGHT RE
395. trol panel VACUUM CONTROL WARNINGS TRANSPORTER e Figure 3 4 VB6 FEG with 1 roughing pump vacuum control panel 3 7 6 VB6 mimic panel key e LEDs above the buttons are lit when the button has been activated e LED DOOR is lit when the loading chamber door is properly closed e LED GUN LID is lit when the emission chamber is properly closed e Vito V18 are lit when the associated valve is open as seen by the valve detector e LEDs for rotary pumps PVP are lit when the associated PVP is working e LEDs for turbo pumps are lit when the associated pump is running Part Number 878275 Vectorbeam Operator Manual 3 7 6 1 Warnings Page 31 LEDs for ion pumps IGP1 4 are lit when the associated pump is working Pe1 Pe2 Pe3 Pl1 PI2 and PI3 are lit when the vacuum in the related areas is better than the required level LED LOAD is lit when the stage is at the load position LED STAGE FRONT POUCH CRANE POUCH AIRLOCK GENMARK SUBSTRATE HANDLER are lit when there is a holder in that position LED FORK SAFE is lit when the Genmark Substrate handler is in the Fork Safe position LED TRANSPORTER is lit when an error has occurred during the substrate exchange LED STAGE is lit when the stage drive is inhibited by the Genmark Substrate handler not being at the Fork Safe position LED COMPRESSED AIR red is lit when the compressed air is insufficient LED COLUMN WATER red is lit when the water l
396. troller provides up to 65536 different frequencies equally spaced for use during the exposure of a single pattern file The master clock from which the exposure clocks are generated operates at 1 5 GHz which means a period of 0 67 ns Therefore the accuracy of any exposure clock will be to the nearest 0 67 ns The minimum and maximum frequencies of the band will be automatically set up before the exposure They should be within a factor of 100 of each other or a warning will be generated The maximum useable frequency is 25 MHz There are two methods of defining the minimum and maximum frequencies 1 Define the frequencies directly 2 Define the doses from which the frequencies can be automatically calculated using Equation 13 1 or 13 2 Dose controller band set up The dose controller band is set up automatically by Emma to cover all required frequencies both if doses and frequencies have been assigned to clocks in Emma or if the pattern file contains relative doses The only instance when the operator can override this automatic set up is when a pattern contains relative doses If this type of pattern is selected command qset pattern the maximum and minimum relative doses are noted These values may be changed by the qset band command if it is issued after the pattern has been selected If a range of exposure frequencies fmin to fmax is to be defined directly then type VB OPER QSET BAND fmin fmax If a range of area doses dmin to
397. ttern file which is defined below is exposed at that die The plugin process begins with the first group of dropouts using the first plugin pattern defined and continues sequentially through the defined groups using the corresponding plugin patterns This enables several different patterns to be defined within one layout If fewer plugin groups than dropout groups are defined then no exposure is carried out at any of the dies in the remaining groups of dropouts PLUGINWORKFILE_1 ENGINEERING DIE 1 PLUGINWORKFILE 2 TEST DIE 1 Defines the names of the pattern files to be exposed at each die in a group of plugin dies The patterns for each group must be defined by defining the symbols pluginworkfile 1 to pluginworkfile n where n is the same as the number of plugins RESET DATUM INTERVAL 00 30 00 Defines the interval during pattern exposure at which the datum is reset using mcjobs sdp com Set to 999 if this is not required Part Number 878275 Vectorbeam Operator Manual Page 221 CALIBRATE_INTERVAL 02 00 00 Defines the interval during pattern exposure at which jobcal is run using mcjobs reset_ontime_jobcal com Set to 999 if this is not required DISPLAY none Defines whether the cell index is typed up before processing each cell Can be set to none or cell SORTMODE normal Defines the subfield sorting strategy used by the pattern generator Usually set to normal but can be set to nosort to
398. ttern nopostmove Part Number 878275 Vectorbeam Operator Manual Page 86 qset height realtime Part Number 878275 Vectorbeam Operator Manual Page 87 11 Mark locate 11 1 Review of principles The Vectorbeam Series products have been provided with a versatile capability to locate a range of registration marks for use in system calibration procedures and in pattern overlay procedures when two or more exposure levels are required The function of the mark locate subsystem is to accurately determine the position of a mark on the substrate or holder target plate with respect to the axial position of the beam This operation requires the electron beam to be scanned in some manner over the expected location of a mark and the resulting image contrast to be used in an algorithm which reports the actual observed mark location The difference between the expected and observed location provides an error vector value which can be used in calibration and alignment routines to apply corrections to cancel any such errors Generally the beam scanning over the mark is directly positioned by the control system software through the main or sub field trap deflection systems whilst the resulting backscattered electron video signal is amplified and is digitised The mark will be distinguishable from the background by a different signal level and the contrast at the mark edges can be extracted and related to beam position and ultimately mar
399. tuting size 512 in the FDL FRE Pattern generator resolution The pattern generator resolution grid size is the beam deflection obtained by changing the pattern generator output by one bit Defining the pattern generator resolution also defines the maximum fieldsize and vice versa The pattern generator resolution is set by defining the maximum fieldsize in Emma using the FLD command The factor relating pattern generator resolution and maximum fieldsize is given below 16 bit pattern generator For a 16 bit pattern generator Maximum Fieldsize pattern generator resolution x 65536 Equation 19 1 18 bit pattern generator For an 18 bit pattern generator Maximum Fieldsize pattern generator resolution x 262144 Equation 19 2 20 bit pattern generator For a 20 bit pattern generator Maximum Fieldsize pattern generator resolution x 1048576 Equation 19 3 Choosing the pattern generator resolution maximum fieldsize The pattern generator resolution is chosen based on the following considerations 1 Minimise the grid snapping that occurs when the shapes in the designed pattern are placed on the pattern generator resolution grid To do this it is necessary to know the design grid If you have designed the pattern then this will be known already Otherwise the design grid can be determined by inspecting the shapes making up the pattern As the shapes are defined digitally all the vertices will be an integer number of reso
400. ue in percent is shown in place of the x xx 3 ERROR Frequencies out of range The frequencies are out of hardware range 4 Error Bad parameters One of the dose determining parameters resist sensitivity VRU etc has an illegal value 5 ERROR Couldn t set clockband Emma couldn t communicate with the pattern generator 6 ERROR Couldn t set clocks Emma couldn t communicate with the pattern generator VB_DCLK_STATUS The logical VB_DCLK_STATUS is set by qdisplay clocks and qadjust clock to one of the following strings 1 OK 2 WARNING Clocks need adjusting 3 ERROR Frequencies out of range The frequencies would be out of hardware range if an attempt were made to set Part Number 878275 Vectorbeam Operator Manual 13 2 3 3 13 2 3 4 13 3 13 3 1 13 3 2 Page 121 them in the pattern generator 4 ERROR Bad parameters One of the dose determining parameters resist sensitivity VRU etc has an illegal value VB BCM The logical VB_BCM is set by qdisplay bcm to the string Beam current x xxxxeyyy nA where x xxxxeyyy is the measured value of current VB DBCM The logical VB BOM is set by qdisplay bcm to the measured value of current e g 0 5345 25 2753 Exposing proximity corrected patterns The VB enables the dose to be varied within a pattern in order to compensate for the proximity effect see also Section Dose control
401. ues by measuring the point half way between the last two showing intensities of the mark and background 11 8 3 Fine search Fine search scans are carried out centred on the edge positions found in the bisection search If the fine qualifier is used the fine scans are centred at the height and width specified The fine scans are positioned in the other axis If the fine scans locate the mark then the mark locate routine finishes If the mark cannot be located then a bisection search is carried out If the bisection search can locate an approximate position the fine scans are then carried out again at that position The fine scans consist of a series the number is given by PARASCANS of parallel scans of length MLEN placed on each of the four expected edges The scans are horizontal or vertical as appropriate for vertical and horizontal edges These scans are spaced out to cover a length of edge for averaging purposes length defined by MH or MW If any of the scans fail to find an edge they are discarded unless more than half fail in any particular grid in which case an error is returned Once a satisfactory set of scans on an edge has been gathered a least squares fit is carried out on the edge data to provide an equation describing the edge The two equations describing the vertical edges are averaged to give a vertical line in the centre of the mark and the two describing the horizontal edges are averaged to give a horizontal line i
402. uld be used in conjunction with the Vectorbeam Command Set Manual 878274 EMMA commands the VB6 Process Module Controller PMC Operator Manual for machines with Brooks handling 893157 the Machine Managers Guide 892813 the Acceptance test and Operator Jobfiles User Manual 892777 the Jobfile standards 892815 the Manual procedures for obtaining Demag tables 892878 the Demag Tables 892894 the VB6 Pre Installation guide 878215 Documentation on VMS and DCL is supplied on CD by HP Compag with the computer The manuals are available in PDF generally being supplied on CD This document covers topics very roughly in the order in which they would be encountered if an operator were faced with the Vectorbeam in a powered off state Cross referencing between chapters ensures that topics which may be encountered at several different sequences of operation are covered only once This version relates only to Emma release version given on the Abstract page at the front of this manual This version relates only to PICs version 11 and Applications Jobfiles release V02 04 D WARNING IMPORTANT SAFETY INFORMATION If this equipment is operated in a manner not specified by the manufacturer the protection provided by the equipment may be impaired It is impossible to fully predict or list here all the possible safety hazards which could occur in using the Vectorbeam The equipment is designed to minimise any risks if used as it was int
403. ure to keep it closed On power off and reboot V11 will not be opened deliberately however the drives to the open Part Number 878275 Vectorbeam Operator Manual Page 34 inlet and the drive to the close inlet will both be switched off This enables the valve to leak especially if the chamber is at atmospheric pressure after an isolated vent Consequently when switching PICS off either a full vent must be done or an isolated vent can be done after arranging an external compressed air supply on the close line When PICS has been restarted it sets the loader arm to coupled and assumes that the stage is occupied Then at the first substrate transfer command or qdisp air PICS performs a dummy unload to ensure that the chamber is empty 3 8 2 VB6 The valve V17 requires compressed air to remain closed Component Machine power off Power on or reboot eae o Se ow er EC eT chamber Penning gauges C Bt The stage MUST be in the load position when the substrate handler which is under PICS control is initialised Note that V17 requires compressed air pressure to keep it closed On power off and reboot V17 will not be opened deliberately however the drives to the open inlet and the drive to the close inlet will both be switched off This enables the valve to leak especially if the chamber is at atmospheric pressure after an Part Number 878275 Vectorbeam Operator Manual 3 9 3 10 3 11 3
404. urement should be less than 0 02 um mm This will contribute a stitch error on a 1 mm field of 20 nm The maximum and minimum residual errors are given under the Observed X Y Part Number 878275 Vectorbeam Operator Manual 15 8 15 8 1 15 8 2 Page 138 Rotation Statistics heading The magnetic map should be carried out initially for each beam energy to be used Once the calibration has been carried out the results should be saved to the appropriate database to be recalled if required Demagnification and spot table calibration In order that the QSET SPOT and QSET CURRENT commands may work a demagnification table and a spot table must first be created The demagnification table consists of pairs of C1 and C2 settings which cover the range of useful beam currents their gun alignment settings and the relative demagnification factors The spot table consists of these same parameters with the addition of beam currents and beam diameters The demagnification table calibration is normally performed by Vistec engineers on delivery and whenever a significant change is made to the column The demagnification table does not depend on the age of the source or the final aperture used For FEG columns however the demagnification table does depend on the extractor voltage and both the demagnification and spot tables will need recalibrating if the extractor voltage is changed Normally the extractor voltage is fixed for long period
405. us and stigmation manually in SEM mode Check the autostig autofocus parameters and try again 0x40830027 DCP Afoc stg curve fit Div by zero ScaleVector Reboot PG 0x40830028 DCP Afoc stg curve fit Div by zero in fwd elim Reboot PG 0x40830029 DCP Afoc stg curve fit Div by zero in back subst Reboot PG 0x4083002A DCP Asig plane fit fewer than 3 points Check the autostig autofocus parameters and try again Reboot PG 0x4083002B DCP Astg plane fit Div by zero finding X coeff Reboot PG 0x4083002C DCP Astg plane fit Div by zero finding Y coeff Reboot PG 0x4083002D DCP Astg plane modify points coincident Reboot PG 0x4083002E DCP Autostig timed out Reboot PG 0x4083002F DCP Autostig history list overflow Reboot PG 0x40830030 DCP Used CheckTime uninitialised Reboot PG 0x20830031 DCP Afoc astg parabola fit gave minimum not maximum Reboot PG 0x30830032 DCP Warning Autostig closed off accuracy dubious Reboot PG 0x40830033 DCP Tried to start a mark locate when one was already running Change mark locate parameters to speed it up 0x40830034 DCP Tried to abort non abortable function Reboot PG 0x40830035 DCP Tried to abort a function that wasn t running No action 0x30830036 DCP Abort came too late to stop function No action 0x40830037 DCP Cannot query the state of this function No action 0x20830038 DCP Unable to append to image file Check that the directory protection for vb dat dir allows Write on operator terminal 0
406. username oadiag and password espritoa Part Number 878275 Vectorbeam Operator Manual Page 43 5 Machine start up from cold Please see Vectorbeam Customer service procedures manual document number 893116 Part Number 878275 Vectorbeam Operator Manual Page 45 6 Machine shut down Please see Vectorbeam Customer service procedures manual document number 893116 Part Number 878275 Vectorbeam Operator Manual 7 3 Page 47 Emma control software The primary functions of the Vectorbeam control software Emma are e Provision of sophisticated commands to allow fast and accurate operation of system making full use of its potential e Provision of manual override through such features as the joystick and data display e Transmission of the pattern data from the hard disk to the pattern generator on exposure e Provision of test and diagnostic software Emma runs on an Alpha station workstation OpenVMS provides all the editing facilities for command sequences There is an interactive command language shell which is DEC Command Language DCL compatible The user can use the full line of DCL features to work with the command sequences for the Vectorbeam Emma works with 2 windows on the workstation display they are e The Status Window where the current status of the machine is continuously updated e The Job Control Window where 1 The operator issues commands 2 Measurements requested such as heig
407. ut from the expected mark position alternately above and below the expected centre until the coarse search limit CSLIM is reached or until a scanned line contains a video profile indicating the presence of a possible mark When a horizontal coarse scan succeeds a vertical confirmation scan is made The horizontal coarse scan must contain a falling and rising edge It is the position half way between these two edges upon which the vertical confirmation scan is centred If the two outer edges in the Part Number 878275 Vectorbeam Operator Manual Page 99 vertical scan are consistent with the marker dimensions then another confirmation horizontal scan is performed This procedure is for circles and octagons which only have well defined dimensions near the middle of the sides of the shape If a vertical confirmation scan fails to confirm the presence of a mark then the search with the horizontal scans will continue until a mark is confirmed or the search area is exhausted A coarse search area exhausted error will be displayed If the vertical confirmation scan succeeds then the position of the possible mark found is passed to the fine scan routine The position found by the coarse scans and passed to the fine scans routine is taken to be the mid point of the successful vertical confirmation scan CSLIM 30um W 4um 4um Coarse Horizontal H Oo Scan Coarse Vertical Scan Figure 11 9 Diagram showing coarse search raster scans fo
408. ut of range MRKCSLEN Mark cross coarse search length parameter out of range MRKMSRLEN Mark measurement length inappropriate for mark size MRKCSLIM Mark coarse search limit MRKFSLIM Mark fine search limit MRKCSRCH Mark locate exhausted coarse search area MRKHEIGHT Height of feature found does not correspond to param MRKWIDTH Width of feature found does not correspond to param MRKEDG Mark edge data insufficient or of poor quality MRKMFDAC Mark main field DAC MRKSFDAC Mark trap field DAC MRKLEDGE Failed to find left edge MRKREDGE Failed to find right edge MRKTEDGE Failed to find top edge MRKBEDGE Failed to find bottom edge MRKARITH Arithmetic error during mark locate MRKNONSPEC Error occured in module not specific marklocate MRKABORT Operator aborted mark locate MRKRTT Mark rise time tolerance MRKCTT Mark contrast tolerance MRKBDW Mark bandwidth MRKFXSRCH Mark fine search x MRKFYSRCH Mark fine search y MRKFSRCH Mark fine search MRKFXFIT Mark fine x fit MRKFYFIT Mark fine y fit AGAFAIL AGA Failed AGASIGHIGH AGA Signal Too High AGASIGLOW AGA Signal Too Low EDGEATXTREM Edge too close to extremity of scan EDGEDATANON Edge data does not make sense NOVIDEO No video signal detected LIMITEDACC Beam diameter of limited accuracy AXIALDRIVE Axial stig drive beyond range DIAGDRIVE Diag stig drive beyond range
409. utomatically to provide a suitable signal Height meter tables define the ranges over which the laser drive and detector gain can be varied by the system to obtain a reading so as to minimise the recovery time when a reading is taken off the edge of a substrate The greater the gain adjustment between readings the longer it takes The time required to obtain a reading on a surface after attempting a reading off the edge of the substrate will be relatively large and can take up to 20 seconds in the worst case 10 tables are available arranged in order of ascending maximum and minimum gain This means that table 1 will work for the most reflective substrates and table 10 for the least reflective substrates Table choice A suitable table must be chosen for each substrate type The lowest number table which works reliably should be chosen to reduce the amount of time required in the case that a reading is taken on the surface after taking a reading off the edge of the substrate somewhere during exposure Working reliably means that no warnings or errors such as Laser height sensor reading poor Part Number 878275 Vectorbeam Operator Manual 10 1 2 10 2 Page 74 occur If in doubt a higher table number can always be used at the risk of taking longer for recovery As the height meter must be used to measure both the datum plate and the substrate two appropriate height meter tables are used for processing a holder Height meter table 7 is
410. uum SVBIOITI a ier tiae dl Pon a AR a Les aai 25 Vacuum controller PICS or Brooks control icone rete tatnen to bee tana 25 Vacuum contol PINE m HON 25 37 1 VBS FEG Vacuum control palle toten eot nece tenet ect nata 26 Sica VBE MIME panel Key iioi euni taie oett ince orate erase tob oo Helga 26 3 7 8 VB6 FEG with 2 roughing pumps vacuum control panel 28 Pagei Vectorbeam Series Vectorbeam Operator Manual Page ii 3 7 4 VB6 FEG with 1 roughing pump and no bypass vacuum control panel29 3 7 5 VB6 FEG with 1 roughing pump vacuum control panel 30 9 770 VBM panel Key inerte Dar eno nU 30 SJ WentsystemibDbUllOm s s sen eti e Eoi OD HE EGER EHE ERG 32 940 JunpisystemblbtOh secius etes eite o E RD tei EU ta 32 3 79 Airlock VenbDUllor nectit e tec Pe ve i o T E E RE Ed Ha edd 32 3 710 Airlock full vent DUHOM 2 2 rrt eite iin caution uei Le d deae a 32 Sd dupockpumpibilliohas cni ta oet t M tt ben E ees 32 3 42 Bakeout DOON sopircr eec e ci ir Have dec E ERE 4 32 9 7 19 BakeouttesebDULtOW cie triti paren c pU ERR 32 9 4 5 14 JBeSOoUDULLORU iiie Rire E Io Porto IIR docs cu e e eiie ie oie 32 3 8 PICS and Brooks power up and reboot actions sssssss 33 38l MBBLsene isse nan i a aaa 33 dig WBO LSsOovedsbetdehotei aon eae a 34 3 9 Vacuum system Star UD oiii eic na ti ru iu Ee Sedes a Mera Uim EOob Hid a dE 35 8 10 Vacuum system Shut GOW sod sidiro d
411. valid cassette number Cassette move timeout SCALP link test failed Stage not at load position Cassette contents undefined Command accept timeout Invalid gauge type selected Error initialising ISCOS O P PIO Error initialising ISCOS I P PIO Transfer source is same as destination Timeout for substrate transfer Unknown source for substrate transfer Unknown destination for substrate transfer Timeout reading sensors Airlock not fully pumped down Vacuum timeout waiting for transfer Vacuum alarm set after substrate transfer OCC Tx readback error OCC Rx link error OCC No new data timeout Not safe to move the cassette Cassette not at index position Not safe to perform transfer Not safe to move loader arm Error occurred during transfer Cassette motor didn t start General cassette error Could not open main gate valve V4 Could not close main gate valve V4 Could not perform dummy load because cassette too full Stage contents presently unknown Substrate handler is not at forksafe Source was empty Destination was full There is a holder on the substrate handler already Substrate handler generated an error Airlock gate valve still open Cassette at up overtravel Cassette at down overtravel General PICS error Airlock door open not safe to move cassette No action Close the airlock Check the proximity switch 0x200f0045 0x200f0046 Substrate handler failed to collect holder Substrate handler failed to deliver holder
412. verage Part Number 878275 Vectorbeam Operator Manual 17 6 2 1 17 6 2 2 Page 171 of these measurements uncorrected and the average of these measurements corrected for the edge sharpness corrected The parameters FILT POINT and LINESCANS must be chosen to be large enough so that noise on the signal does not make the measurement inaccurate Measure diameter using metal calibration mark Measuring the diameter on the calibration mark is relatively quick can be done automatically and can be done on any holder It is limited to spot sizes above about 30 nm 1 Move stage to the 10 um octagonal calibration mark 2 With a 16 bit pattern generator in order to be able to scan the beam in small steps to measure the smallest spot sizes more accurately the pattern generator resolution should be set to a small value Set the maximum fieldsize to e g 0 16384 mm VB_OPER gt FLD 0 16384 For the 18 bit and 20 bit pattern generators this is not necessary 3 Switch to SEM mode and graph mode 4 Adjust gain and backoff so that the signal from the mark is a little below the maximum level and the signal from the background is a little above the zero level 5 Adjust the focus and stigmation 6 Type VB_OPER gt QDISPLAY DIA FM ALL LINELENGTH 2 EDGE 5 FILT 16 POINT 8 LINESCANS 1 Measure diameter using knife edge Measuring the diameter using a knife edge is a relatively long procedure must be done manually and usua
413. ween the substrate and the calibration height by default the height corrections are on However due to non linearities of the height meter system and differences between X and Y the compensation will not be perfect Therefore further coefficients are available to take out any inaccuracies in the rotation for height correction as shown in the table below These coefficients are applied to the height meter reading that is used for the rotation for height correction and calibrated on the basis of measurements on the Vectorbeam There are coefficients for a 2 order polynomial for both X and Y allowing independent adjustment These coefficients allow fine tuning of the rotation for height and non linear rotation for height The constant terms effectively produce a rotation offset but it recommended that the calibration offsets are used instead see Section Calibration offsets in Chapter Calibration Part Number 878275 Vectorbeam Operator Manual Page 81 X Y Nominal value NonlinearRotConstX NonlinearRotConstY NonlinearRotScaleX NonlinearRotScaleY 1 000 NonlinearRotSquareX NonlinearRotSquareY The use of the coefficients to adjust the height meter reading can be enabled or disabled using NonlinearRotOn In order to access the coefficients on the DCP directly login to the DCP VB SUPER rlogin dcp Vectorbeam Pattern Generator Master login pgdiag Password gt To display the current value type gt double NonlinearRot
414. wever unfortunately a segment number still has to be supplied even if the qSet Grating command contains only Offset or Scale qualifiers When the machine is in Wafer mode any scale tweak that is being applied via the MOGG is included in the calculation of the interpolation coefficients for the corrections tables In Absolute mode a direct copy is done not interpolation Displaying grating definition The command to display all sets that have been defined is VB OPER qdisplay grating Clearing grating definition A command to clear all the sets that have been defined is required VB OPER qset grating none This will clear all data for all blocks but qSet Grating None Block will clear only the data for the nominated block Selecting the algorithmic grating generation The current command to select the pattern should be expanded to include the algorithmic grating generation alternative e g VB OPER qset pattern grating Defining the order of the shapes The existing nosort and normal sorting modes set by the qset sort command or the qset pg strategy command do not apply and will have no effect The order in which sub shapes of gratings are exposed is configurable to include the orders as follows 1 Expose each line all sub shapes of the shape completely before starting the next as done in the existing nosort mode on the 16 bit pattern generator however start the exposure of the next line
415. which a previous or subsequent layer of a multi layer device has or will be written The mode of the machine when no correction coefficients have been applied is NATURAL mode Absolute stage map mode set up The absolute stage map mode allows the movement of the stage to be corrected for scaling rotation keystone scaling linearity and bow errors By correcting for these errors the true grid placement of the stage can be improved considerably Determining the absolute stage map corrections Theory The positions of an array of marks on a substrate covering the area of stage travel is measured by locating the marks with the beam on axis to avoid any deflection field calibration errors Any measured deviations to the nominal positions of the marks will be as a result of errors in the positions of the marks on the substrate combined with errors due to the stage positioning and errors due to the mark locate accuracy The errors due to the substrate and the stage can be separated mathematically if measurements of the grid are carried out with the substrate at four different rotations 0 90 180 and 270 degrees Procedure 1 The temperature of the system should be set up and stable Part Number 878275 Vectorbeam Operator Manual 8 9 Page 140 The system should be in a fully calibrated state Load a substrate with a suitable array of marks e g an autostitch plate and allow to stabilise Execute a data collection job f
416. witch into SEM mode and then blank the beam using the beam on off button on the Emma status window Part Number 878275 Vectorbeam Operator Manual Page 136 10 Rotate RV8 back and forth to find the range where the image brightness changes from normal to almost completely dark i e almost blanked 11 If the corner of the mark moves during the preceding step gt 0 5 um then adjust lens 2 C2 so that there is no movement of the image between the two pot positions This is best done by displaying the current value by typing VB OPER din2 example result 20 107 and then varying the value until the best focus is obtained by typing VB OPER sln2 value example value 0 110 12 Run the automatic fine focus and stigmation adjustment as described in Section Automatic fine focus and stigmation adjustment in Chapter Job specific machine setups 13 Compare the fine focus value from the preceding step with the value calculated above Adjust lens 3 and repeat from the preceding step until the fine focus values match to within about 0 010 14 Having found the correct lens 3 value the beam blanking operation in Cabinet B should be returned to Normal and the pot returned to its former value 15 Save the new fixed value of lens 3 C3 in the appropriate databases 15 7 Magnetic map calibration On a VB6 a dedicated magnetic map is required for all 8 inch wafer holders which have the datum in the corner and anot
417. x40830039 DCP Image collection command received when one was already running No action 0x4083003A DCP Timed out getting BEF update semaphore Reboot PG 0x2083003B DCP Unable to append to dump file Check that the directory protection for vb dat dir allows Write on the operator terminal Slave PGP errors 0x40840001 PGP Couldn t spawn SLC task Reboot PG 0x40840002 PGP Couldn t spawn FISP task Reboot PG Stage error messages 0x400b0001 Couldn t create qErrors queue Reboot stage 0x400b0002 Couldn t create intertask comms queue Reboot stage Vectorbeam Operator Manual 27 6 7 1 27 6 7 2 27 6 7 3 27 6 7 4 27 6 7 5 27 6 7 6 Part Number 878275 0x400b0004 0x400b0005 0x400b0006 0x400b0007 0x400b0008 Task spawn errors 0x400b0101 0x400b0102 stage 0x400b0103 0x400b0104 0x400b0105 0x400b0106 0x400b0107 C library errors 0x400b0201 Semaphore errors 0x400b0301 0x400b0302 0x400c0001 0x400c0002 0x400c0003 0x400c0004 0x400c0005 0x200c0006 0x200c0007 0x200c0008 0x400c0009 0x400c000a 0x400c000b 0x400c000c 0x400c000d 0x400c000e 0x400c000f 0x400c0010 0x400c001 1 0x400c0012 stage Expose Errors 0x400d0001 0x400d0002 0x200d0003 0x400d0004 0x200d0005 0x400d0006 0x200d0007 0x400d0008 0x400d0009 Stage Errors 0x400e0001 0x200e0002 0x200e0003 0x200e0004 0x400e0005 0x400e0006 0x200e0007 0x400e0008 0x400e0009 Couldn t create vacuum system queue Couldn t create stage queu
418. x40AA0F13 SAEHT Low voltage PSU fault 15V Ox40AA0F14 SAEHT Low voltage PSU fault 15V Ox20AA0F16 SAEHT GS type not specified Ox20AA0F17 SAEHT EHT turned off Ox40AAOF18 SAEHT EHT disabled Ox40AAOF19 SAEHT Filament disabled Ox20AA0F20 SAEHT Focus supply off Ox20AA0F21 SAEHT Extractor supply off Ox40AAOF22 SAEHT Hot Box link problem E HBTIMO Ox40AAOF23 SAEHT Hot Box not enabled Check the EHT power supply is switched on Ox40AA0F24 SAEHT Hot Box tripped out Ox40AAOF25 SAEHT Hot Box link problem E HBCMD Ox40AAOF26 SAEHT Hot Box link problem E HBILPA Ox40AA0F27 SAEHT Hot Box link problem E HBPAOR Ox40AAOF28 SAEHT Hot Box link problem E HBLINK Ox3S0AAOF29 SAEHT Analogue readback error focus Ox20AA0F2A SAEHT Value out of min max range focus Ox30AAOF2B SAEHT Analogue readback error extractor Ox20AAO0F2C SAEHT Value out of min max range extractor Ox40AAOF2D SAEHT Extractor ADC not zero Ox40AAO0F2E SAEHT Focus DAC not zero Ox4O0AAOF2F Error in reading from hotbox Ox20AA0F30 SAEHT Illegal Runup profile parameter Ox40AAOF31 SAEHT Checksum not OK Ox20AA0F40 SAEHT Invalid unit specified Ox30AA0F41 SAEHT Digital readback error Ox30AA0F42 SAEHT Analogue readback error Ox20AA0F43 SAEHT Beam turned on Vectorbeam Operator Manual Part Number 878275 0x40AA0F 44 Ox20AA0F45 0x40AA0F46 0x20AA0F50 Ox20AAOF51 0x10AAOF60 Ox40AAOF61 0x40AA0F62 0x40AA0F63 0x40AA0F64 0x40AA0F65 0x40AA0F70 Ox20AA0F71 Ox
419. xample VB OPER qfile save filename test test col 20 kV 30 nm spot Magnetic map database generation If this is required calibrate the magnetic map as described in the Section Magnetic map calibration This calibration can then be saved using for example VB_OPER gt qfile save filename mag cal mag cal 20 kV 8in wafer Part Number 878275 Vectorbeam Operator Manual 12 6 12 7 12 8 Page 114 Database selection prior to exposure To ensure the machine has all the correct parameters loaded prior to any exposure or job 1 Load the basic database See Section Basic database generation for the beam energy VB OPER dfile load lt filename dbase gt total A warning may appear if certain elements of the file to be loaded such as the field size calibration sensitivities pivot points aperture number and EHT are not the same as those presently in use The database will not be loaded unless the override qualifier is used Alternatively set the elements such as the field size to match the database first 2 Load the field corrections for the maximum fieldsize to be used from file If these are not available see Section Field corrections database generation VB OPER dgfile load filename dist dist cal VB OPER dfile load filename stig stig cal 3 Load the column settings for the beam current to be used from file If this is not available see Section Beam database generation VB OPER dgfile l
420. y istec SAS Lithography Vectorbeam Series Vectorbeam Operator Manual CONFIDENTIAL Part number 878275 Vectorbeam Series Vectorbeam Operator Manual Abstract This document describes the operating procedures for Vistec Vectorbeam series systems This version of the manual is for VMS based workstations Please read the important safety information on page 1 All reasonable steps have been taken to ensure that this publication is correct and complete but should any user be in doubt about any detail clarification may be sought from Vistec Lithography Ltd or their accredited representative The information in this document is subject to change without notice and should not be construed as a commitment by Vistec Lithography Ltd Vistec accepts no responsibility for any errors that may appear in this document Copyright 2006 Vistec Lithography Ltd Cambridge England All rights reserved The contents of this publication may not be reproduced in any form or communicated to a third party without prior written permission of Vistec Vectorbeam is a trademark of Vistec All other referenced trademarks are the property of their respective owners Part Number 878275 Date 18 Nov 2005 Issue 3 1 draft Software Version EMMA v6 38 Author P Hoyle Printed in England This document is originally written in English S Corrections and changes to this manual should be forwarded by emai

Download Pdf Manuals

image

Related Search

Related Contents

（動作確認用） ライナー ＊重量 *505g  Réf. 50360-2 Manuel d`utilisation  Programming manual (CC  取扱説明書 - 三菱電機  Trevi PJ 3028  Manual de usuario. Envío de formato electrónico de Servicio para  Operating Instructions  xbr_x830c service manual  Development Board User Manual    

Copyright © All rights reserved.
Failed to retrieve file