Senior Design Spring 2008 Paper
Contents
1. lt Inside Maximum J Minimum Syringe Diameter Rate Rate uL mra ul hr pl hr Hamilton Microliter 0 001 0 001 0 002 0 004 0 001 0 001 gt w fe b s 20 60 wasa 8 20 s BUT masr 20 os NEED 08 IO IO e jo e le o H G Publication 41200 01 49 Model NE 1000 Multi Phaser 05 26 06 54 E5 Infrared LED Infrared LED _ 1 2791 series Small emission spot LED using current confined chip 12791 is infrared LED with microball lens cemented to the current confined chip surface This combination ensures high directivity and improved emission uniformity In particular L2781 02 uses a lens cap that delivers even narrower directivity As a variant type not using a microball lens 12791 03 is also available with the LED chip potted with resin which gives a small emission spot of 180 Small emission spot Automatic control systems L2791 400 um Optical switches 12791 03 160 Auto focus Uniform emission L2791 02 Narrow directivity L2791 02 gg Absolute maximum ratings Ta 25 C Parameter Symbol Condition L2791 03 is guaranteed to resist temperature cycle test of up to 5 cycles W Electrical and optical characteristics Ta 25 C2. sung Figure ue CTA ET M Figure Cl 3 aca PERS 27 Fiure 2 uqa a desi ayy naya Guinan trate 30 2 2 S 3l 2 M 31 PT OD A tee vias DD 32 Iip rg 2 5 33 PigureE2 eck 36 E22 2 RE 38 Chapter I Introduction As of right now the main procedure for distinguishing cells apart is flow cytometry We are working on a more cost efficient method called optofluidic intracavity spectroscopy OFIS This procedure has the advantage of not using fluorescent tags which are quite expensive and add to the time needed for the sample preparation By using the refraction of light through the cells instead of fluorescents we can observe the cells in a more natural or pure form and therefore do not have to worry about the chance that the tags may interfere with normal cell function The components of our system are a microscope an infrared LED a cell detection circuit a data acquisition unit with analog to digital convertor and digital output computer control software a function generator a RF switch a DEP chip with the micro fluidic channel in it and a spectrometer See figures below To date we have made a channel to flow our sample solution through
3. 1 lt Figure C2 2 a b Figure C2 3 a RS232 input to Oriel Controller Figure C2 3 b DB9 input to computer 31 A Labview VI controls the signals that are put on the different wires of the RS 232 Cable The front panel for the Labview VI looks like this String to Write IDN V Time On ms 2 Delay Before Read 2 500 Delay Before Write ms 2 500 Delay Before Read ms 2 soo Time Off ms 500 7 Read String Pump Stop Manual Cmd ab ab 6 Figure C2 4 Front Panel of Oriel Controller VI The switches in the lower right portion of the front panel control what the switch actually does Only one switch can be on at one given time in order to function properly When the VI is run with only the pump switch on the pump will run for a total of 2 Delay Before Read the time value ms specified in the top left input box When the VI is run with the stop switch turned on it will stop all pumping functions With only the manual command switch turned on a string will be sent to the Oriel controller For example the string V200 n will command the Oriel actuator to move at a velocity of 200 um second There many commands like this that the VI already uses to control the actuator within the VI itself such as run and stop The way that the pump command works is that it will send a run command Then there will be a specified delay time followed by the stop command being sent to the c
4. 5 Figure C 2 1 a Control Panel of Oriel Controller b Microscope stage being controlled The Oriel Instruments Encoder Mike Controller is normally used as a microscope stage controller to move a microscope stage very slowly Our application of this piece of equipment was as a syringe pump The main reason that this was used is because it was able to move a syringe plunger slow enough to get reasonably slow flow within the channel of the chip The minimum rate at which the actuator can move is 0 5 um s Remember that the chip only had a cross sectional area of approximately 5000 um The syringe that we are using is approximately 0 5 cm in diameter This makes it so that the stage must be moving extremely slow in order to get a flow rate within the channel to be as slow 40 um per second which is the flow rate by which it is suitable to trap cells Again we used Labview to control the actuator This was much more difficult to control because there were no VIs available on the internet This made for a much bigger challenge because we were unfamiliar with RS 232 communication that was required for control of the Oriel actuator Using a wiring diagram that was in the User s manual of the Oriel actuator we created our own RS 232 to DB9 cable to interface between the control panel and a computer 30 Block Diagram TREES DB9 gt RS 232 8 2 2 2 Set Yes Yes gt unb Beis 0 unb
5. 0 55 va 5 VO lo Dark Current 5 0 nA VR 5 0V Ap Peak Response Wavelength 860 _ nm tr _ output Rise Time zm 0 6 ns Vg 50 50 Q 10 90 tr Output Rise Time 5 ns 15 500 10 90 Output Rise Time ns Vg 5 0 V 50 1026 9096 Total Capacitance Typical Performance Curves DARK CURRENT vs AMBIENT TEMPERATURE 100000 10000 1000 100 DARK LEAKAGE CURRENT nA 0 25 50 75 100 125 T AMBIENT TEMPERATURE C DARK CURRENT vs REVERSE VOLTAGE 3 0 25 20 3 15 lt 10 8 5 EM 0 0 0 10 20 30 40 50 60 70 80 90 100 REVERSE VOLTAGE VOC CAPACITANCE pF pF Normalized Responsiviity Va 5 0V Relative Spectral Response PT AISNE EET Wavelength nm Typical Capacitance vs Reverse Voltage FIBER OPTIC 2 2 2 a o REVERSE VOLTAGE VOC Optek reserves the right to make changes at any time in order to improve design and to supply the best product possible Optek Technology Inc 1215 W Crosby Road 8 91 Carrollton Texas 75006 972 323 2200 Fax 972 323 2396 41 E2 DAQ Comparison Number of Measurement Channels range Resolution Sampling Rate 4 1440 Variable Depends on internal gain DI 158 setting 12 bit 0 Hz NI USB 6008 20V
6. 1 97 L2791 Parameter symbol Conaitior Min Typ Max Min 55 infrared LED 1 2791 series Emission spectrum Radiant flux vs forward current 25 50 MA imetOD 01 1 z E 3 E 5 2 2 a WAVELENGTH inm Li Forward current vs forward voltage Bi Directivity 4c22 Typ T2225 tw 100 us 1 KT PY 4 JEN b eo z m 150 b UL 5 m H eL rae Q S 5 SACO aa on E Di RELATIVE RADIANT OUTPUT g L27S1 03 Except for refecton ingredient of the base FORWARD VOLTAGE V Lil d Lili W Radiant output vs ambient temperature Allowable forward current vs ambient temperature 422 q g 115 55 14 5 82 A TY m 4 22 a AMBIENT TEMPERATURE AMBIENT TEMPERATURE Lih gm up 56 infrared LED 1 2791 series Allowable forward current vs duty ratio Typ Tae25 ALLOWABLE FORWARD CURRENT mA IB Dimensional outlines unit mm 12791 2L2791 02 L2791 03 2582 02 254202 Sars COMMON TO CASE HAMAMATSU p amp be bowers no assumed kq pean 26 om addons me subject fo change without notion Nu pater rights are to any ct the crc descrited Seer C2008 Haramatiu Photenicw HAMAMATSU PHOTONICS K K Solid State Division 1126 1 Ichino cno Hamamats
7. A 120nm layer of gold is then placed over the chrome layer Gold is deposited because of its low resistance and low chemical reactivity Once this deposition is completed photolithography follows AZ1512 Photoresist is placed and then spun on the chip This resist is thinner and has very good resolution A soft bake at 110 degrees Celsius for 1 minute follows Next the chrome DEP mask is used to pattern the resist A chrome mask is chosen for this exposure because of the small features it encompasses Chrome masks are expensive about 400 for a 3 resolution 4 x4 plate but have a high resolution and have an antireflective coating to enhance the exposure This mask contains patterns for two different trapping circuits to select from A characterization of the AZ1512 resist found and exposure time with the soft bake of 10 seconds The resist is exposed for 10 seconds which maximizes the resolution of the resist The exposed resist is then developed leaving the unexposed area covered A two step wet etch removes the uncovered gold and chrome from the glass slide The DEP trap chip is then ready for two of the three types of bonding The DEP channels were composed of either polydimethylsiloxane PDMS or glass PDMS channels were the majority of the channels prepared since the fabrication success rate of these channels was much higher than glass The success of PDMS is attributed to its exceptional adherence to glass The PDMS used was made of bulk P
8. This channel has metal leads DEP traps for setting up the electric field to trap our cell when it is detected by the detection circuit to be analyzed by the spectrometer We first had to build the channels with the traps on them The next step was to determine the proper flow rate of the solution so as to optimize trapping We also made a circuit to detect the presence of a cell to trigger the trap of the control was done with Labview software The way our system works is as follows First light from an inferred light emitting diode IR LED wavelength peak centered at 850nm is focused through the cavity of the micro fluidic channel on the chip The light passes through up to the microscope and out from the microscope via a fiber optical cable into our detection circuit This light gets converted into a current in the nA range using a photodiode and then goes into a transimpedance amplifier to get converted to a voltage and to be amplified This voltage is sent to a data acquisition device DAQ to compare the voltage with a preset reference voltage If the input voltage is below the preset value the DAQ will send a signal to close the RF switch in order to send a 6 or 10 MHz 10 Vpp signal to the DEP traps on the chip This will stop and hold a cell in the trap for analyses by the spectrometer to distinguish the cell In order for the DEP trap to be able to capture a cell the sample solution must be flowing at a slow enough rate so as to not have
9. New Era Pump Systems Inc www SyringePump com 12 7 Syringe Diameters and Rate Limits Inside Maximum J Minimum Syringe Diameter Rate Rate Manufacturer cc mm mL hr D 1 B n un c 4 li ho SEDI pb w i m tale gt ID d H o 1 pe e pe Jal fos gt eo e iun 6 le e w un c b HSW Norm Ject 5 3 un JN f w go joo Ro to cg m p o ut T jw ho rm joo W m o e n in co ur w oT to lela olin r heo 21156115 nfe e or 0 FO Monoject Terumo Air Tite eo m 4 1 w oo w un fw H j o JW t3 Jw jN F in o amp pd e 1 Uu w a Iho e H us o ju 2 un to un m pb j 15 un tle e LSA N eo 29 13 co EISE IN v 13 55 16 72 M ut Du b gt N 1
10. in diameter The majority of the time 9 77 um spheres were used This water with glass spheres will be referred to as fluid in parts of this section When the project first began it was thought that the rate at which the end of a syringe plunger needed to be pushed could be calculated based on the fact that the volume in is the same as the volume out Ideally chips are fabricated with a 200 um wide by 25 um deep channel This gives an ideal cross sectional area of 200 um x 25 um 5000 um If the syringe plunger is pushed at a rate of x um s then the volume of fluid flowing into the nanotube and the chip is m x 0 5 x diameter of syringe um x x um s z um s This flow rate must logically be the same in the channel since the volume in must equal the volume out This result can then be used to calculate the velocity at which fluid will travel through the channel The equation is as follows v um s z um s x 5000 Substituting the first equation into the second gives direct relationship between the velocity of fluid in the channel v um s and the velocity at which the plunger of the syringe is pushed x m s This resulting equation is v um s x um s x 0 5 x diameter of syringe um x 5000 um The only syringes on hand at the beginning of the project were 3 cc syringes with a diameter of 8 585 mm 8585 um manufactured by B D Using the equation above equation to achieve a 40 um s flow rate in the channel the plu
11. included as when nanoports are purchased Nanotubes are very small a kit adapter tubing that are 150 um in inner diameter that are frequently used in micro fluidics The figure shows the basic structure of a nanoport 8 nanoport Figure 3 1 Nanotube Nanoport Interface Before nanoports can be placed on a chip holes must be drilled into the chip in the appropriate places To do this a pattern of the chip fabrication is placed over the chip and the two places for holes are marked Then a drill press is used with a small drill bit Great care must be taken when drilling the holes because glass cracks and chips very easily Water is applied at the drill area and only very small depths are drilled at a time This process is very dependent on minimal human error Many of the fabricated chips have non ideal holes when drilling is complete It is also very important to have a way to securely place the nanoports on the DEP chip so that there is no leakage as cells and liquids are pushed into the channel and pulled back out of the channel sometimes under high pressure The nanoport attachment process is quite simple and never fails no matter how dirty the sample is even for the non adhesive type described below Adhesive o rings are used to create the bond between the nanoport and the chip Adhesive o rings are a sticky glue like material in an O shape that act as double sided tape These rings are very carefully placed around the cen
12. is done using a wet etch of HF The chip is placed in HF for approximately 3 4 minutes The chip is removed from the HF and washed with de ionized water The chip is then r placed in an acetone bath to remove the hardened resist Acceptable glass etching was never achieved in the semester The channel surface must be as uniform as possible to provide better transmission of light Using an Alpha Step surface profilometer to measure the surface height deep indents were found in the glass One explanation to this result is that there are lattice impurities in the glass that etch faster than the uniform lattice If this were the case then higher quality glass would be needed to perform suitable glass etching Once the layer of resist is removed the channel chip can be bonded to the DEP trapping circuit chip Two bonding methods have been implemented The more successful method is Indium bonding A piece of Indium metal and the channel with gold on the outer area are placed in a solution A positive voltage is applied to the Indium metal and a negative voltage is place on the of the channel chip This creates a current of Indium to be deposited on the gold The channel chip is then ready to bond to the DEP circuit chip The two pieces are roughly aligned by eye and pressed together by hand The chips are then observed under a microscope to view the alignment Fine adjustments are made to optimize the alignment Once the
13. junction temperature T max Note 7 Unless otherwise specified the specifications apply over the full temperature range and for V2 420 V for the LF412A and for Vg 15V for the LF412 Vos and 55 are measured at Note 8 The LF412A iz 100 tested to this specification The LF412 is sample tested on a per amplifier basis to insure at least 85 ol the amplifiers meet this specification Note 9 The input bias currents are junction leakage currents which approximately double for every 10 C increase in the junction temperature T Dus to lrnited production test time the input bias currents measured are correlated to junction temperatura In normal operation the junction temperature rises above the ambient temperature as a recut of internal power dissipation Te Ta amp y Po where Gja the thermal resistance from junction to ambient Use cf a heat sink is recommended if input bias current is to be kept to minimum Note 10 Supply vokage rejection ratio measured for beth supply magnitudes increasing or decreasing simulkanecusly in accordance with common practice Vs 6 15V Note 11 Refer FETS412X fcr LF412MH and LF412MJ military specifications Note 12 Max Power Dissipation is defined by the packages characteristics Operating the part near the Power Dissipation may cause the part to operate outside guaranteed limte Note 13 Human body model 15 kf in series with 100 pF Typical Performance Char
14. per second 40 um s This requirement comes from the DEP traps on the chip If the flow rate exceeds 40 um s the force of the DEP trap to stop a cell will not be sufficient to stop a cell and hold it in place Another limitation of flow control is that there must not be too much 10 pressure on the nanoport chip seal or in the channel The bonding on glass to glass chips is not very strong and high micro fluidic pressure will cause the chip to leak If leaking is severe it is possible for cells to flow outside of the channel Obviously this is very undesirable because all of the traps and optical detection are near the center of the channel It is also worth noting that all of the glass chips have gaps between the two pieces of glass due to an imperfect bonding process This creates flow outside of the channel and different fluid dynamics than the PDMS chips Because the glass to glass chips glass chips essentially have larger channels the flow rate is significantly slower than in the PDMS chips when the same amount of pressure is applied to the fluid Most cell samples such as blood are obtained using a syringe Therefore a syringe is used to push samples into a nanotube and into the channel Although the end goal of the project is to pump cells through the channel and analyze them it is impractical to actually use cell samples to develop and test flow control Therefore de ionized water was used with glass spheres ranging from 5 um to 26 um
15. the cells have too much inertia This is done by using a micro pump and software Labview to create a pulse width modulation to have the proper duty cycle to get the flow rate we need Chapter 2 will discuss the fabrication of our chips and how the channels are made and how the traps and leads are added to the chip Chapter 3 will be on DEP chip micro fluidic ports Chapter 4 will discuss the flow control of the microfludics we used plus the pump system Chapter 5 will discuss the detection circuit we designed and some of the components used Chapter 6 will discuss DEP trapping Chapter 7 will cover the future work and recommendations for next year Chapter II Channel Fabrication Photolithography was used to fabricate these cielectrophoreris traps in the CSU Cleanroom This process allows building contacts wires and effectively traps on a micron scale The DEP chips were built with two distinct components a DEP trapping circuit and a microfluidic channel These two components are bonded together to form a DEP trap The DEP trapping circuit is composed of three contacts and three discrete lines running through the area under the channel The conductive material used was gold and chrome These metals were first deposited onto a glass slide using an evaporator in the cleanroom A layer 30nm thick of chrome was first deposited onto the slide This thin layer of chrome is deposited because chrome can adhere to glass much better than gold can
16. the general set up of the entire system The size of the optic fiber coupled into the beam splitter should be around 50 125 to 62 5 125 and needs to be a multimode fiber The percentage of light sent through the optic fiber is 90 while the remaining 10 of light goes to a camera In the system an infrared LED is used as a light source Using a lens this light is focused to a small spot on the microscope stage where chips are placed to provide high light intensity The light that goes into the microscope is then split as described above Optical fibers then transport the light to the spectrometer and cell detection circuit not pictured camera 50 125 multimode fiber beam splitter Infrared or 6 infrared red a source alignment n LEDs laser Figure D2 1 Diagram of microscope being coupled with light source 36 Calibrating the optical system is a fairly involved process with many steps and can take several hours To calibrate the optical system follow the steps below 1 This first step is not truly necessary but serves as a good reference and is a good way to start To get a starting point for calibrating the system the use of a LED that emits light in the visible spectrum is useful Setup the visible light LED below the microscope stage and drive the LED to emit light Place a white piece of paper on the focusing plane the microscope stage where chips will be placed and use the focusing lens for the LED t
17. 12 bit 10 kS s LabJack U3 16 2 4 12 bit 2 5 50 kS s NI USB 12 Yes 6008 LabJack U3 20 Programmable Ports Yes Labvew Gost DI 148U Labvew5 50 http www dataq com products startkit di148 htm 50 DI 158 http www dataq com products startkit di158 htm NI USB 6008 159 http sine ni com nips cds view p lang en nid 14604 LabJack U3 Yes 99 http www labjack com labjack_u3 php prodid 25 42 E3 LF412 Operational Amplifier National Semiconductor LF412 August 2000 Low Offset Low Drift Dual JFET Input Operational Amplifier General Description These devices are low cost high speed JFET input opera tional amplifiers with very low input offset voltage and guar anteed input offset voltage drift They require low supply current yet maintain a large gain bandwidth product and fast slew rate In addition well matched high voltage JFET input devices provide very low input bias and offset currents The LF412 dual is pin compatible with the LM1558 allowing designers to immediately upgrade the overall performance of existing designs These amplifiers may be used in applications such as high speed integrators fast D A converters sample and hold circuits and many cther circuits requiring low input offset voltage and drift low input bias current high input imped ance high slew rate and wide bandwidth Typical Connection Ordering Information LF412XYZ indicates electrical grade
18. 48 1 10 100 10k 100k 10M 10 100 10 100k 1M 10M FREQUENCY Hz FREQUENCY Hz Power Supply Rejection Equivalent Input Noise Voltage 22 150 29 C ES E NET TS EQUIVALENT INPUT NOISE VOLTAGE Hz tk FREQUENCY Hz Open Loop Voltage Gain if HE HATTE EE ULL LU 1k 10k 100k 1M SUPPLY VOLTAGE VJ FREQUENCY Hz OUTPUT IMPEDANCE 0 rao LF412 Typical Performance Characteristics continued Inverter Settling Time OUTPUT VOLTAGE SWING FROM OV V M didis ws Pulse Response 2 c 10 pF Small Signal Inverting L TIME 8 2 DIV 5 85 i Large Signal inverting OUTPUT VOLTAGE SWING TIME 2 ay DIV www national com 8 OUTPUT VOLTAGE SWING 90 e 0 av OUTPUT VOLTAGE SWING Small Signal Non inverting TIME 0 2 5 00 Large Signal Non inverting TIME 2 4 7 MV 50 Pulse Response 2 10 pF Continued Current Limit 1000 OUTPUT VOLTAGE SWING now TIME 5 ws DIV Application Hints The LF412 series of JFET input dual op amps are internally trimmed BI FET 1174 providing very low input offset voltages and guaranteed input offset voltage drift These JFETs have large reverse breakdown voltagee from gate to source and drain eliminating the need for clam
19. 82 connected to a 62 5 um core diameter multimode fiber optic cable The light from an infrared LED goes into the microscope and then through the fiber optical cable into the photodetector A photodetector is a photodiode which is a component with a p n junction When a photon light of sufficient energy strikes the diode it excites an electron in the valence band thereby creating a mobile electron and a positively charged electron hole If the absorption R occurs in the junction s depletion region or on Vin Vous average one diffusion length away from it these carriers are swept from the junction by the built in field of the depletion region producing a 6 photocurrent Photodiodes can be used in either A low pass electronic fiter realized by an RC circut the zero bias mode known as the photovoltaic mode or in the reverse bias mode known as the photoconductive mode the mode we are interested in In the zero bias mode light striking the diode causes a current across the device which leads to a forward bias of the diode which in turn induces dark current in the opposite direction to the photocurrent Dark current is the relatively small electric current that flows through a onion iN photodiode even when no photons are entering the device This is called the photovoltaic effect and is the basis for how solar cells work which are just a large number of big photodiodes Onto Vout reverse bias which only induces
20. DMS and the PDMS curing agent with a mass ratio of 10 1 respectively This PDMS is then placed in a vacuum and degassed Degassing will remove the impurities in the PDMS which will improve the optical properties of the molded part A mold of the channel was needed to form the channel This mold is composed of a silicon substrate and SU 8 resist This epoxy based resist is much thicker and viscous than the AZ1512 and can be used as a mold when it is baked and fully developed Once the mold is constructed PDMS solution is poured into the mold again degassed Once degassed the mold is placed in an oven to bake for at least 4 hours The curing agent transforms this viscous liquid into an elastic solid The PDMS can then be cut and peeled off the silicon substrate A glass cover slip is then placed on the opposite side of the channel which keeps the PDMS chip clean and allows the channel to be transported easier In order for the PDMS channel to be bonded to the DEP trapping circuit one more step in the DEP circuit must be performed A gold and chrome sheet still remains outside of the trapping circuit which is used for other bonding processes This area is removed with the same lithography steps as creating the traps except that a mask covering the trapping circuit is used Now the two chips are ready to be bonded together The DEP chip is taped to a glass slide with the circuit facing out Both chips are then oxygen plasma treated with a micro RIE After treatme
21. Optical Biosensors Second Semester Report Spring Semester 2008 by Allan Fierro David Sehrt Doug Trujillo Evan Vicek Michael Bretz Prepared to partially fulfill the requirements for ECE402 Department of Electrical and Computer Engineering Colorado State University Fort Collins Colorado 80523 Report Approved Senior Design Coordinator ABSTRACT Currently cell sorting is both an expensive and arduous task Optofluidic Intracavity Spectroscopy OFIS is a low cost attractive alternative to these other cell sorting methods such as flow cytometry Unlike flow cytometry OFIS doesn t contaminate the samples with fluorescent tagging OFIS relies instead on index of refraction of the cells which allows for observation of these cells in their natural state The current procedure for OFIS demands an observer to monitor cells flowing through a microfluidic channel trap a cell and set up a spectrometer reading This demands valuable time of the observer and makes cell differentiation unnecessarily slow In order to improve the speed of OFIS an automation system must be implemented to detect cells trap them and take a measurement To achieve acceptable speed through OFIS new hardware was engineered which includes new fabrication techniques for making the polydimethylsiloxane PDMS glass chips with micro fluidic channels with dielectrophoretic DEP traps flow control a detection circuit with a high gain transimpedance amplifier a RF swi
22. Supply voltages less than these may result in lower gain bandwidth slew rate The amplifiers will drive a 2 2 load resistance to 10V over the full temperature range If the amplifier is forced to drive heavier load currents however an increase in input offset voltage may occur on the negative voltage swing and finally reach an active current limit on both positive and negative swings Precautions should be taken to ensure that the power supply for the integrated circuit never becomes reversed in polarity or that the unit is not inadvertently installed backwards in a socket an unlimited current surge through the resulting forward diode within the IC could cause fusing of the internal conductors and result in a destroyed unit As with most amplifiers care should be taken with lead dress component placement and supply decoupling in order to ensure stability For example resistors from the output to an input shoukl be placed with the body close to the input to minimize pick up and maximize the frequency of the feed back pole by minimizing the capacitance from the input to ground A feedback pole is created when the feedback around any amplifier is resistive The parallel resistance and capacitance from the input of the device usually the inverting input to AC ground set the frequency of the pole In many instances the frequency of this pole is much greater than the expected 3 dB frequency of the closed loop gain and con
23. a little current Vin known as saturation or back current along its Voltage follower 52 direction But a more important effect of reverse bias is widening of the depletion layer therefore expanding the collection volume and strengthening the photocurrent Circuits based on this effect are more sensitive to light than ones based on the photovoltaic effect and also tend to have lower capacitance due to the greater separation of the charges which improves the speed of their time response because so the smaller the capacitance the smaller the time constant On the other hand the photovoltaic mode tends to exhibit less electronic noise Another type of photodiodes is the avalanche photodiodes which have a similar structure but are operated with a much higher reverse bias which allows each photo generated carrier to be multiplied by avalanche breakdown resulting in internal gain within the photodiode which increases the effective responsively of the device Our photodetector is used in reverse bias with a large resistor 1 8MQ to have the majority of the 14 current flow into the circuit The next step of the circuit is a low pass filter with a cutoff frequency of 1 000 Hz to reduce the noise from the photodetector The output from the filter goes into the operational amplifier or op amp The op amp a LF412CN is setup in a non inverting configuration with a gain 1 of 2 500 After the op amp stage we have voltage buf
24. act the Natlonal Semiconductor Sales Office T max 150 C 115 C Distributors for availability and specifications ep Typical OM denn mem Operating Temp Range Note 6 Note 6 Storage Temp 65 C lt T 150 C 65 C lt T lt 150 C Supply Voltage 22V 18V Range Differential Input Voltage 38V 30V Lead Temp Input voltage Range Soldering 10 sec 260 C Note 3 19V 15V ESD Tolerance Output Short Circuit Note 13 1700V Duration Note 4 Continuous Continuous H Package N Package LF412A Power Dissipation DC Electrical Characteristics Note 7 Conditions Lenza Units LE m Cos frof ome TC of Input 10 Note 8 Offset Voltage Input Offset Current 25 100 25 100 mewe 2 2 51 J a Input Bias Current 29 so 200 mec 4 4 Signal Voltage Vgzt15V 10 Gain Ry 2k 25 Over Temperature Over Temperature Swing Vs 15V Input Common Mode mE Voltage Range Jms Common Mode Rast 0k 440 Supply T Note 10 Ll Supply Current IVos0V R IVos0V R Note 2 Absolute Maximum Ratings indicate limits beyond which damage to the device may occur Operating Ratings indicate conditions for which the device is functional but do not guarantee specific performance limits AC Electric
25. acteristics Input Blas Current INPUT BIAS CURRENT pA INPUT BIAS CURRENT pA 5 50 25 0 25 50 75 100 125 COMMON MODE VOLTAGE V TEMPERATURE o esseD www natlonal com 4 46 Typical Performance Characteristics Continued SUPPLY CURRENT mA E is 53 EE 3 z NEGATIVE OUTPUT VOLTAGE Positive Common Mode input Voltage Limit POSITIVE COMMON MODE INPUT VOLTAGE LIMIT V 0 5 20 25 SUPPLY VOLTAGE V POSITIVE SUPPLY VOLTAGE V Negative Common Mode Input Voltage Limit Positive Current Limit il OUTPUT SOURCE CURRENT mA Output Voltage Swing SWING Vp p OUTPUT VOLTAGE OUTPUT SINK CURRENT mA SUPPLY VOLTAGE V 6 www nallonal com 47 LF412 Typical Performance Characteristics Continued Output Voltage Swing M EDS DUTPUT VOLTAGE SWING Vp p UNITY GAIN BANDWIDTH MHz 25 50 25 0 25 50 75 100 125 AL OUTPUT LOAD kit TEMPERATURE 50 50 25 0 25 50 75 100 125 inicie TEMPERATURE C SLEW RATE V s 00549521 Undistorted Output Voltage Swing 10 meee I 111 FREQUENCY DISTORTION OUTPUT VOLTAGE SWING Vp p 10k FREQUENCY Hz www natlonal com Typical Performance Characteristics Continued COMMON MODE REJECTION RATIO
26. al Characteristics Note 7 Parameter aza Dus we usc s Amplifier to Amplifier T4725 C f 1 Hz 20 kHz PT Stew Rats ven 6 GsmBendwidh Product eso 4 27 3 www national com LF412 AC Electrical Characteristics Continued Note 7 Total Harmonic Dist Ay 10 10 20 Vp p 20 Hz 20 kHz Equivalent Input Noise Ta 25 C Re 1002 ies Equivalent Input Noise 25 f 1 kHz ae we P Note 3 Unless cthenvise specified the absclute maximum negative input voltage equal to the negative power supply voltage Note 4 of the amplifier outputs can be shorted to ground indelintsly however more than one should not be simultanscusly shorted as the maximum junction temperature wil be exceeded Note 5 Fer operating at elevated temperature these devices must be derated based cn a thermal resistance of Note 6 These devices are available in both tha commercial temperature range C CZT4z70 C and the military temperature range 55 lt 7 2125 The temperature rangs is designated by the postion just before the package type in the device number A indicates the commercial temperature range and an indicates the military temperature range The military temperature range available in package only all cases the maximum operating temperature limited by internal
27. alignment is made the two chips are sandwiched between two copper sheets These sheets are then put in a vise The vise is screwed together with four screws This puts equal pressure on the bonding area The bonding vise is then placed into a heater A vacuum is pulled in the heater to prevent oxidation from the high temperatures The temperature in the heater is controlled by a programmable unit attached to the heater The temperature is then programmed to ramp up to and hold at a certain temperature and then ramp down Once the desire temperature program completes the bonding vice is pulled from the heater The channel and the DEP trapping circuit chips are now bonded together A similar bonding procedure is gold to gold bonding The advantage gold has over Indium is that the height of the material required for bonding is minimal compared to Indium Minimizing the height for bonding is critical because the height dictates how leaky the channel will become Indium bonding adds another layer in the bonding area outside of the chip This layer adds an additional height on the scale of microns This will produce a leakier channel than chips bonded with a gold to gold method This gold to gold method has not successfully been completed on test chips These test chips are half the size of the DEP trapping circuit chips and have gold with no patterning The two gold chips are pressed together and put between the copper plates and into the vice The temperatur
28. all percentage of the time For details on how the Oriel was implemented into the system see Appendix C2 To achieve 40 um s velocity inside the channel the Labview VI was set so that the actuator moved for 0 825 seconds at 0 5 um s 0 4125 um and off for 7 seconds When the VI is used the optimal flow rate occurs at this duty cycle of approximately 10 5 This duty cycle description applies to the movement of the actuator However duty cycle is not an accurate comparison because the flow rate in the channel takes much longer to decay than the 7 second period during which the pump is turned off This is discussed more later Another consideration is that the stage mounted to the actuator is spring loaded and adds complication to the analysis This spring action actually allows the plunger to be pushed at a rate slower than the 0 5 um s that the actuator is moving actuator Labview program spring loaded stage syringe Oriel Figure 4 Through experimentation it was quickly discovered that the derived equation does not hold true for glass chips but is reasonable for PDMS chips This is because the actual cross section in a glass chip s channel is larger and cannot be accurately estimated Experimentation also revealed that due to complex pressure dynamics the flow rate in the channel does not follow the syringe plunger To be more precise when the plunger is moved and then stopped the flow in the channel does not stop im
29. e current flow control method can achieve desired flow rates the velocity of particles in the channel and the volume of throughput are currently human estimates Different chips sometimes require different setting in Labview to achieve the same result If an accurate way to measure flow rate were achieved a feedback control system could allow for the same flow rate even if there are variances in the effective cross sectional area of chips channels The optical system will need to be fine tuned to allow for adequate light modulation on chips with dielectric coating This coating significantly reduces the transmitted light intensity transmitted power If we are able to obtain any donations a higher output infrared LED could improve the system Also if a high performance photo diode was obtained the performance may improve Last the optical detection circuit could be improved in the area of time response Using a high performance transimpedance amplifier designed for photocurrent amplification may enhance performance This amplifier should have a larger gain bandwidth product to allow for higher frequency response while maintaining the necessary gain and signal to noise ratio In the future of the project the spectrometer will become an essential part of the cell analysis Data acquisition will need to be automated using Labview software and resulting data of the cells will allow for cell differentiation Because it is the shifting of the wavelengt
30. e fixed value If a false is generated in this boolean variable then the algorithm goes back to stage 1 to read in a new value If a true is generated then the Hytek s capabilities are exploited again The Hytek has drivers to 27 send out digital signals of either 0 or 5V When the boolean variable is true the Hytek sends a digital signal to apply an RF signal to the traps theoretically trapping the cell Stage 4 future work Once the cell is trapped the spectrometer will take it s reading store it in a file to be read by the user This has not yet been implemented but it is planned to be done in the future Currently we are modeling this as a delay of 500 ms Stage 5 Once the spectral reading is taken the digital signal that is applied to send the RF signal to the traps is set to OV and the sphere will be released Stage 6 In order for there to be no re trapping of the same cell there is a small delay of 500 ms after the trap is released to allow time for the recently trapped cell to clear the trapping area before another voltage reading is taken From this stage process starts over again until terminated by the user 28 Here is a flow chart of the algorithm Read Input from Circuit Compare Input Value to Fixed Value Wait for Cell to Is Input Value No Leave Trapping Less than Fixed area Value Release Cell Trap Cell C2 Oriel Instruments Encoder Mike Controller 18011 E j
31. e is programmed to be much higher than indium bonding This is due to the fact that Indium bonds much easier to surfaces than gold does to gold or any other surface The melting point of gold is also very high In order to fuse the two pieces together a very high temperature must be obtained An appropriate temperature and pressure have no been obtained which prevent acceptable bonding Small bonding areas have been obtained but not large enough to be a reliable method for chip bonding Given the difficulty of metal bonding alternative approaches were explored but have not yet been incorporated in actual micro fluidic channels Chapter III DEP Chip Micro Fluidic Ports In order to get cells into the DEP chip with a micro channel there must be a way for cells to physically enter the chip It is also important that there is a way for the cells to exit the chip This is accomplished through the use of nanoports There are two nanotube nanoports on every chip one for the cells to enter into the channel and one for cells to exit the channel Nanoports are cylindrical shaped parts that are on the order of 1 2 cm in diameter and 2 3 cm tall They are composed of a plastic like material On the bottom of the nanoports there is a small circular opening with a rubber seal surrounding it From the top 1 nanoports are hollow with threads so that a nanotube be secured into the nanoport using a hollow screw like adapter These adapters are
32. fer or voltage follower used as a buffer amplifier which is used to eliminate loading effects or to interface impedances Vout Vin with Zin oo in theory but in reality it is the input impedance of the op amp which is usually to ITO Some of the difficulties we ran into were a noisy signal which is why we added a low pass filter having the op amp oscillate so we added some capacitors to eliminate this We also had a weak input signal which is why we increase the resistor to 1 8 Below is the final circuit 5 VDC fe C5 100n D1 PHOTODIODE A R2 U1B d AA 3 2k L A R1 z 250k 9n LF412 0 0 Figure 5 3 15 Second Semester Updates Due to many problems with the previous circuit design such as noise causing errors in our results and simply not enough gain we redesigned the entire light modulation detection circuit The redesigned circuit Figure 5 4 is still based around the premise of the previous circuit s transimpedance stage but uses a different configuration It was seemingly problematic to use a resistor based transimpedance such as R1 in Figure 5 3 to generate a voltage based off of the photocurrents as the photocurrents ranged from 900 nA to 5 and the amount of external electromagnetic interference seemed to be enough to cause erroneous values The redesigned circuit takes advantage of a transimpedance amplifier TIA configurati
33. ggest issue with the circuitry The limited bandwidth due to the LPF has caused the overall system performance to decrease The amplification stage produces the largest time constant of the system and is the cause of the issues with a slow time response In order to achieve a faster time response we could either tweak the LPF for optimal values or use an op amp made for transimpedance configurations such as Texas Instrument OPA380 What this amplifier provides over our current op amp the LF412 is a much larger gain bandwidth allowing for a greater input signal bandwidth and also offers great precision and low noise which is important when trying to amplify currents as low as 1 nano Ampere A future design idea which would call for a complete redesign of the circuit would include an op amp like the OPA380 and a voltage buffer To achieve the best distortion reduction and performance in a transimpedance circuit all required gain should be taken care of with the TIA The output of the gain stage should be followed by a voltage buffer The same noise cancellation ideology and capacitive decoupling should be followed from previous designs however the low pass filter should be tweaked to acquire a stable phase margin Ideas for improving performance with a transimpedance circuit can be found in the datasheet for the OPA380 PI 18 Chapter VI RF Switching To fully automate the system the cell traps would have to be turned on when a cell i
34. h that 21 differentiates cell types the spectrometer is the heart of the system Triggering DEP traps based on light modulation i e intensity is very difficult and may not be feasible with current project limitations An alternative to trapping based on light intensity would be to use the spectrometer and attempt to detect changes in wavelength and trigger the DEP traps accordingly It may even be possible in the future to skip the trapping step entirely and simply take data of the cells spectrums as they pass Such alternatives should be investigated along with other possibilities to improve the rate at which cells can be analyzed Another key element to automation is software Once the spectrometer is implemented and spectral data is taken the data needs to be stored in an efficient filing system and analyzed for the end user Without software each spectrum would have to be analyzed by a person and compared to other data The software should have statistical data about the analyzed cells and should have tolerance parameters that can be set by the user to specify what characteristics determine different cell types 22 REFERENCES What is Optofluidics Optofluidics 2007 10 Dec 2007 http www optofluidics caltech edu optofluidics index html Yang Changhuei and Demetri Psaltis Optofluidics Optofluidics Can Create Small Cheap Biophotonic Devices Laser World 1 July 2006 10 Dec 2007 http www laserfocusworld com article
35. indicates temperature range M for military C for commercial indicates package type H or N BI FET 117 is a trademark of Nations Semiconductor Corporation 2004 National Semiconductor Corporation DS005656 Features Internally trimmed offset voltage 1 mV max Input offset voltage drift 10 y V C max Low input bias current 50 pA Low input noise current 0 01 pA Hz Wide gain bandwidth 3 MHz min High slew rate 10V ys min Low supply current 1 8 mA Amplifier High input impedance 101202 Low total harmonic distortion Low 14 noise comer 50 Hz Fast settling time to 0 01 2 ps 30 02 Connection Diagrams Metal Can Package MON ARVERTING NON NVERTING INPUT r Note Pin 4 connected to case TOP vew 2266642 Order Number LF412MH LF412CH or LF412MH 883 Note 1 See NS Package Number Dual In Line Package OUTPUT A INVERTING INPUT A burruta 3 NOS INVERTINGZ INVENTING INPUT NON INVERTING TOP view Order Number LF412ACN LF412CN or LF412MJ 883 Note 1 See NS Package Number JO8A or NOSE www national com 134f 39sJJO MOT 21 1 LF412 Simplified Schematic Note 1 Available per JM38510 11906 Detailed Schematic www natlonal com 44 Absolute Maximum Ratings nots 2 H Package N Package If Milltary Aerospace specified devices are required Note 12 Note 5 670 mW please cont
36. lace turn on the infrared LED source Using the marked spot on the computer monitor verify that in the open channel there are a few volts output from the detection circuit Also by positioning the marked spot over DEP trap leads and or other objects in the channel verify that the minimum voltage outputs are obtained when the spot is directly over these objects Optical calibration is now complete The difficulties encountered in the calibration of the microscope were numerous and were mainly due to the poor maintenance of the entire instrument The second difficulty is the basic concept of the optics 38 The microscope used for our experiments has been poorly maintained Dirt and scratches on the lens are present which may alter the light collection but to a lesser degree The beam splitter stage was unstable and poorly fitted onto the microscope and due to budget constraints has to be fastened to the microscope with zip ties If our budget constraints allow we can machine better housing for the beam splitter The focusing plane where the micro fluidic channels sit is also difficult to use with the channels Scotch tape is required to steady the chip in place and leveled with the focusing plane The use of tape leads to oils from the fingers being deposited onto the chip which alters the light characteristics of the chip Again if our budget would allow we could machine a better focusing plane that would allow us to secure the chip into
37. mediately but rather decays over the course of several minutes This behavior is analogous to the behavior of an RC circuit Both have a time constant associated with how long it takes the flow rate to die down The larger syringe 8 585 mm 12 controller diameter builds up a higher pressure and the time constant is several times longer than the time constant for the smaller syringe 4 699 mm diameter This suggests that the cross sectional area of the syringe is proportional to the time constant for the flow rate inside the channelAll of these more complex dynamics make experimentation the best indication of actual flow inside the channel There was no precise way to obtain the velocity of spheres moving in the channel but the duty cycle described earlier is estimated to produce a flow rate of 40 um s in a PDMS channel Due to limited fabrication of glass chips near the end of the semester the duty cycle for the desired flow rate in a glass chip has not been found However it is known that the duty cycle must be higher than that for the PDMS channel Second Semester Updates Spheres often get trapped on the edges of the drilled hole when pumping a solution through the channel To deal with this problem a new set of solutions were made and documented These solutions are more dilute and should allow for operation with less clogging The new 9 77 um sphere solution is 0 05 cc of sphere solution to 7 cc of de ionized water The new 7 um sphere s
38. mpler terms the force to push an object into a trap is proportional to the positional gradient of the electric field The implications of such a thing is that the particle will experience a force until it reaches an extrema in the intensity of the electric field the positional derivative of the electric field is zero and therefore the force will equal zero This translates into basically a restoring force that keeps the particle trapped at a certain point until the electric field is removed or until a much larger force physically removes the particle from the trapping area In the scope of this project the non uniform electric field comes from a 5 Vpp sinusoidal signal across the two electrodes of the traps 20 AC yoltage Ground AC Voltage Figure 6 2 a Electromagnetic Modeling of DEP traps Reproduced from 2 Based upon the field lines shown in Figure 6 2 the point of highest electric field intensity is approximately in the center of the trap This provides an optimal trapping point for which the spectrometer can take a reading Chapter VIII Future Work Chip bonding has shown to be a critical step in chip fabrication Failures in PDMS and thermocompression bonding have not been reliable methods for bonding the trap and channel pieces together and more reliable methods need to be found With so much variance in chip fabrication it would be beneficial to create a feedback control system for flow control While th
39. nger must be pushed at a rate of 0 00346 um s The original plan for flow control was an industrial syringe pump Borrowed from the chemistry department the NE 1000 syringe pump made by New Era Pump Systems Inc can push this B D syringe at a rate of 2 434 uL hr This is the equivalent of 6 761 x 10 um s This is 1 954 x 10 times too fast Obviously this syringe pump is not a valid solution to the flow rate problem In addition when the syringe pump was set to its slowest setting the torque broke the encoder coupler because the pressure in the chip created a force pushing back against the syringe plunger that was beyond what the pump was designed for As a solution the Oriel Instruments Encoder Mike Controller 18011 was chosen to function as a custom pump The Oriel Instruments Encoder Mike Controller 18011 Oriel is designed to be used as a precise way to move a microscope stage It can move the actuator at a minimum velocity of 0 5 um which is still about 150 times faster than the desired rate for the larger syringe 8 585 mm diameter To reduce the rate at which the plunger must move a 11 smaller syringe 4 669 mm diameter was chosen to be used in the system This syringe plunger only needs to move at a rate of 0 0115 m s to achieve the desired flow rate of 40 um s in the channel Because this is still slower than what the Oriel is capable of Labview software was used to create a duty cycle where the actuator is only moving for a sm
40. nt the two chips are bonded together using the mask aligner The PDMS channel is placed on the stage facing up The DEP chip is positioned facing down being held in the mask holder The DEP traps are then positioned over the channel The channel is then brought into contact with the DEP circuit chip The two are now bonded together The mask holder is then unscrewed and pulled out vertically from the aligner Glass channels were experimented with as well These glass channels are more desirable than PDMS because a dielectric coating can be deposited on the surface of glass producing a optical cavity necessary for cell differentiation Pyrex glass is used for channel etching and chrome and gold are deposited on the surface A thick layer of P4400 positive resist is deposited onto the surface This resist was chosen because it is much thicker than the AZ1512 resist which will be necessary with HF etching The tradeoff with using P4400 is that it doesn t have as good a resolution as AZ1512 In this application giving up resolution is certainly acceptable The soft bake for this resist is 110 degrees Celsius for 2 minutes The resist is then exposed with the channel mask for 35 seconds The exposed area is washed away in development The chip is then baked at 110 degrees Celsius for 10 minutes This turns the resist into a hard film which can withstand the wet etching process The gold and chrome are etched away The next step is to etch the glass This
41. o focus the visible light on the same plane the piece of paper The goal of the lens above the LED is to make the light emitted from the LED confined to a very small area in a circular shape Note that the lens should be fairly level if the LED lens and focal point are aligned vertically Once complete remove the piece of paper 2 Turn the microscope light on Center the infrared LED below the lens where the microscope light falls after passing through the hole in the stage and the lens and connect it to a current source for operation Now turn off the microscope light 3 Carefully rotate the lens out of the path between the LED and microscope without changing its height or tilt 4 Turn on the spectrometer and set it up to take intensity counts Note that you must ensure that the optical fiber is connected between the microscope and spectrometer Set the reference level for the spectrometer with no light source 5 Turn on the infrared LED with a50 mA current Adjust x and y tilt of the LED to maximize the intensity counts that the spectrometer reads 6 Now reposition the lens over the LED and under the stage Fine tune adjust the position tilt and even height of the lens to maximize the intensity counts on the spectrometer We achieved counts of around 3000 Note that although the height should not have to be repositioned after step 1 fine adjustments often improve the intensity 7 Turn off the spectrometer and place a chi
42. olution is 4 cc of sphere solution to 10 cc of de ionized water Perhaps the most prominent problem that hindered the project is leaking chips Chips often leak at the nanoport glass barrier Another common place that chips leak is the edges of the chips where the glass to glass bonding or PDMS to glass bonding occurs Glass to glass bonding problems cause leaking and this problem is more extensively discussed in Chapter II Channel Fabrication Many leaks cause fluid to flow outside the channel or not in the channel at all The most effective method of fixing leaking chips is UV glue However UV glue can only be used to patch the outside of a chip not where glass to glass or PDMS to glass bonding has failed Therefore patching the perimeter of a chip or the nanoports can prevent leaking but not flow outside the channel within the chip The UV glue is mostly transparent and does not harden until exposed to ultraviolet light After applying UV glue to the necessary part s of the chip the glue must cure in the ultraviolet light for at least 5 minutes Experimentation has found that 15 minutes of curing seems to be more appropriate 13 Chapter V Detection Circuit We needed to be able to tell when a cell was present in our system For this we decided to design a detection circuit This circuit would detect a cell by a modification in the light intensity In order to design the detection circuit we started with a ST connectorized photodetector OPF4
43. on shown in figure 5 4 as well as a voltage amplifier vie E ik C2 luF T C3 luF a HK i M Cl 5nF R3 700k A D Input Figure 5 4 16 The first stage of the circuit Figure 5 5 is the TIA with a transimpedance gain of 4 4MQ The second stage of the circuit Figure 5 6 is the voltage amplifier taking the voltage output from the TIA and applying a gain of 700 V V R3 700k Figure 5 6 Figure 5 5 Noise was still seen as a problem with this new circuit Noise of about 60 Hz was seen on the output of our amplification stages which caused unreliable results while automating as it created a 30 50 mV difference on the output of the circuit The noise was dealt with by the use of the low pass filter LPF Figure 5 7 as well as decoupling the power supplies from the circuit Figure 5 8 The LPF can be seen as the capacitor C1 in parallel with the feedback resistor R1 This LPF limits the overall bandwidth of the circuit causing a slower time response but is required to limit the amount of noise from the input signal The capacitors on the power supplies act to decouple the supply lines from the circuit which greatly limits the destabilizing effects of the inductive response the supplies cause when applying a load We also created an aluminum enclosure to house the circuit in order to stop any electromagnetic interference Cl onF Ri 44M Figure 5 7 Figure 5 8 17 The time response is now the bi
44. ontroller then a second specified delay The loop starts over starting with arun command This allows the user to specify a duty cycle for the pump The values that we used for this VI were time on 725 ms delay before write 50 ms delay before read 50 ms and time off 7000 ms This roughly gave a flow rate of 40 um per second The VI block diagram shows all of the internal logic that occurs with the switches and with the delays as well as the Run and Stop commands sent to the Oriel actuator 32 Enable Termination Char T timeout 10sec 10000 icone r Before Read ms BEEE R Instr Bytes at Port Time On ms 2 Before Read Time OFF ns IsHeHsHsHeHeHaHeH Read String Figure C25 33 C3 Ocean Optics HR2000 High Resolution Fiber Optic Spectrometer The overall goal of the project is to differentiate cells as they pass through the channel In order to do this the method of intracavity spectroscopy will be used In order to detect the changes in wavelength as a cell passes through a spectrometer is used to analyze the diffraction of the LED light source At this point in the project the spectrometer has only been used to assist with the setup of the optical system The spec
45. p on the microscope stage Use the camera to focus the microscope on the center of the chip s channel using the infrared LED as the light source 8 Once focused turn the infrared LED off Connect an alignment laser preferably in the infrared region to transmit into the viewing plane through an optic fiber This is 37 basically shinning light in the opposite direction back onto the chip like from the spectrometer back to the beam splitter and then the microscope 9 Light from the laser is seen in the viewing plane via the camera Adjust the beam splitter until the laser light is confined to a small circular area Our best calibration had a focal point that was on the order of 10 um to 15 um in diameter This ensures that the output from the beam splitter is collecting light from an area where the LED light is focused on and that a cell in this focal point will produce significant light modulation It is important while doing this that the laser is at a low power output so that the camera is not saturated With our infrared laser we used 0 2 to 0 3 mA of current to drive output Figure E2 2 Example of alignment using a laser 10 Mark the spot on the computer monitor with a marker Do not move the camera window on the computer after this mark is made Turn off the laser and connect the optical fibers for normal operation with the detection circuit 11 Turn on the power supplies for the detection circuit With the chip still in p
46. pe across the inputs Therefore large differential input voltages can easily be accommodated without a large increase in input current The maximum differential input voltage is independent of the supply voltages However neither of the input voltages should be allowed to exceed the negative supply as this will cause large currents to flow which can result in a deetroyed unit Exceeding the negative common mode limit on either input will cause a reversal of the phase to the output and force the amplifier output to the corresponding high or low state Exceeding the negative common mode limit on both inputs will force the amplifier output to a high state In neither case does a latch occur since raising the input back within the common mode range again pute the input stage and thus the amplifier in a normal operating mode Exceeding the pasitive common mode limit on a single input will not change the phase of the output however if both inpute exceed the limit the output of the amplifier may be forced to a high state The amplifiers will operate with a common mode input volt age equal to the positive supply however the gain band width and slew rate may be decreased in this condition When the negative common mode voltage swings to within of the negative supply an increase in input offset voltage may occur Each amplifier i amp individually biased by a zener reference which allows normal circuit operation on 6 0 power sup plise
47. pensive units were about the same or even worse We designed a Labview VI that would take an analog input compare it to some specified reference and output a digital signal based upon whether or not it was higher or lower than the given reference Here is the front panel of the Labview VI iuo Figure 1 2 Front Panel of DAQ VI The interface is quite simple The triggering voltage is input in the field labeled Comparison Voltage The channel of the analog input is easily configured Any errors are easily read on the error out field Since we configured the DAQ to take inputs on analog channel 0 the default value getting a digital output based upon comparing to a preset voltage was as easy as inputting the comparison voltage and hitting the run button in Labview The VI for controlling the DAQ was not difficult to program Another reason we went with this DAQ was due to the fact that Labview VI s were available on the Hytek Automation website This made it so that it was a plug and chug type of program 26 Here is the VI block diagram DeviceType Figure C1 3 Block Diagram of DAQ VI The algorithm used for trapping is described as follows Stage 1 The first thing that needed to be done was to be able to read the output of our light modulation circuit This was accomplished by using the Hytek Data Acquisition Unit The Hytek has analog to digital capabilities and drive
48. pherals in this project is abundantly apparent We used many different measurement and control systems to be able to achieve our goals These peripherals were Data Acquisition Unit USBDAQ U120816 from Hytek Automation Ocean Optics Spectrometer HR2000 Oriel Instruments Encoder Mike Controller 18011 Microscope heavily customized Olympus 230997 Logitech USB camera Beam splitter Newport parts manufactured in house C1 Data Acquisition Unit Figure C1 1 Hytek Data Acquisition Unit The data acquisition unit was brought on board due to a need for triggering a circuit based on amount of light collected by the photodiode This made it so that instead of creating a new logic circuit every time we changed the gain stage of the amplifier the logic could be controlled with a Labview VI This also took out the guess work for a logic chip In the specification sheet for a 7400 series logic component there was a large range of voltages that were specified to be undefined logic The DAQ took a lot of the uncertainty out of the logic The DAQ is accurate to within 3 mV DC on its 8 available channels of analog inputs It also can 25 output digital logic at 5 V DC on one of its 18 Digital I O channels This may seem like it s overkill to have so many but the DAQ was 2 3 of the price of some of the other data acquisition units that are on the market It has a sampling rate of 13 000 samples per second on a single channel More ex
49. place and avoid taping it The fiber optic output from the beam splitter is also difficult to use at time as there is no relief at the ST junction for the fiber optic cable A relief at the ST junction would allow for the focal length from the optic fiber to the beam splitter to be changed without causing stress to the fiber optic cable or having to dissemble the output of the beam splitter The knowledge of optics was one of the easier difficulties to overcome with the help of Dr Lear Understanding why we needed a 62 5 125 optic fiber as opposed to a much larger optic fiber is an example of the problems he helped us overcome The reason why a 62 5 125 cable is used to minimize the area in which the light from the beams splitter is collected 39 Product Bulletin OPF482 August 1996 Appendix E Data Sheets and User Manuals E1 OPF 482 Optic Fiber Photodiode D Fiber Optic High Speed PIN Photodiode Type OPF482 Features Component pre mounted and ready to use High speed low capacitance Pre tested with fiber to assure performance Popular ST style receptacle Electronically isolated from case Description The OPF482 consists of a low cost plastic cap PIN photodiode pre mounted and aligned in an ST receptacle This configuration is designed for PC board or panel mounting Includes lock washer and jam nut two 2 56 screws and dust cap The PIN Photodiodes are designed to interface with mul
50. rs available on the Hytek website to allow for a conversion of an analog signal to a 16 bit floating point number This allows for approximately three decimal place precision on the readings from the light modulation circuit Stage 2 Once the reading is taken from the light modulation circuit and converted to a floating point number the second thing that is done in this algorithm is to compare that number to a fixed value The fixed value that we have been using is the average voltage reading from the light modulation circuit when no sphere is present within the trapping area and the reading when there is a sphere present This creates a buffer by which particulates and various fluctuations in the output voltage would not set off the traps For example generally the output voltage without a sphere in the trapping area is approximately 3 V When a sphere is in the trapping area the voltage drops to 2 V Therefore the comparison voltage would be 2 5 V That way if a small particulate were to flow into the trap area we would not have a false trapping Also if the voltage of the output were to fluctuate by 10 without a sphere being in the trap the worst case scenario would be a 2 7V output vs 3V This would still allow operation to work without creating a false trapping situation Stage 3 Once the comparison occurs it sets the value of a boolean variable This boolean is true if the voltage was less than the fixed value and false if greater than th
51. rt While holding the nanoport in place use a pair of tweezers to grab the sides of the nanoport and press it against the chip while applying pressure on the center of the opposite side of the chip with a finger This should allow the placement of a clamp over the nanoport and chip holding the 2 together without fingers in the way Place a clamp over the nanoport and chip holding the 2 together Make sure that the clamp is centered and applying even pressure across the nanoport surface before releasing the tweezers hold on the nanoport Put the chip with the nanoport clamped on in the oven and bake it at 170 C for 1 hour Remove the chip from the oven and allow to cool 35 D2 Optical Calibration At the heart of the OFIS process is the microscope This is the instrument that we use to gather light to use for spectroscopy cell detection and for the camera No other instrument needs to be calibrated as frequently as the microscope and no other instrument in the process can dramatically change the resulting data by a slight adjustment Therefore the correct calibration of both the microscope and the light source is vital for any reasonable data collection from a micro fluidic sample In this section the calibration process of the microscope will be covered The difficulties in learning how to calibrate the microscope will also be explored To begin the basic knowledge of how our system works is essential Below is a figure on
52. s article_display html id 259933 1 J Voldman R Braff M Toner M Gray and M Schmidt Holding Forces of Single Particle Dielectrophoretic Traps Biophysical Journal Vol 80 pp 531 541 January 2001 2 W Wang H Shao Lear Lab on a Chip Single Particle Dielectrophoretic DEP Traps powerpoint presentation given March 2006 Lear Kevin L Hua Shao Weina Wang and Susan E Lana Optofluidic Intracavity Spectroscopy of Canine Lymphoma and Lymphocytes IEEE Explore4 Dec 2007 4 McGoldrick Paul ed TI OPA380 Wireless ZONE 10 May 2004 20 Mar 2008 http www en genius net site zones wirelessZONE product reviews hfp 051004 5 OPA 380 Texas Instruments Sept 2007 20 Feb 2008 lt http focus ti com lit ds symlink opa380 pdf gt 23 APPENDIX OR APPENDICES Appendix A Abbreviations AC Alternating Current DAQ Data Acquisition Unit DC Direct Current DEP Dielectrophoretic LED Light Emitting Diode OFIS Optofluidic Intracavity Spectroscopy OpAmp Operational Amplifier PDMS Polydimethylsiloxane VI Labview Virtual Instrument Vpp Volts Peak to Peak Appendix B Budget e 432 Digital Switch 15 Various circuit elements including Op amps and digital chips 10 Hytek iUSBDAQ 0120816 105 TOTAL EXPENSES 130 Starting Budget 500 over 2 semesters Money left 500 130 370 24 Appendix C Peripherals The need for peri
53. s in proximity of a trap To do so the Data Acquisition Unit DAQ would monitor the voltages from the cell detection circuit until the given voltage exceeds the threshold showing the presence of a cell The DAQ would then create a digital high output on an analog switch or relay and cause a signal of radio frequency RF to be applied to the traps thus causing a cell to be trapped in an electromagnetic field The components we tried to use to switch the RF signal such as an analog switch no longer performed as an ideal switch due to the high frequencies At RF the capacitance associate with the analog switch caused a percentage usually 10 of the signal to leak when the switch was considered open To avoid problems with RF signals we chose to use a Teledyne 172 5 RF DPDT relay which is designed to handle RF signals Chapter VII Dielectrophoretic DEP Trapping DEP trapping is necessary in this project for one main reason If the sphere cell particle is not held in a particular position without moving a spectral reading of light shined through that object will be extremely difficult to gather data with the equipment and limited budget that we have DEP trapping is a method of using electromagnetic forces to hold an object in place One might ask the question how do you exert a force upon a charge neutral object using an electric field The answer to this question is not trivial DEP trapping occurs when a non uniform electric field is passed
54. sequently there is negligible effect on stability margin However if the feedback pole is less than approximately 6 times the ex pected 3 dB frequency a lead capacitor should be placed from the output to the input of the op amp The value of the added capacitor should be such that the RC time constant of this capacitor and the resistance it parallels i amp greater thian or equal to the original feedback pole time constant www national com 51 LF412 Typical Application www national com Single Supply Sample and Hold 52 LF412 Physical Dimensions inches millimeters unless otherwise noted Continued 85 45 0 310 0 410 0300 2 326 0 62 8 126 1916 D 381 www natlonal com SN 0 008 80 610 Eu DE b 0 220 0 310 VAX 0 291 GLASS Mb ae R0 023 0 045 0 055 GLASS SEALANT 0 055 MAX BOTH ENOS 0 100 0 010 nu DuaHn Line Package J Order Number LF412MJ 883 NS Package Number JOSA 23 0 480 19 474 18 16 0232 0 006 10 RAD PIN 1 IDENT 0 250 0 005 16 35 3 0 127 OPTION 2 0 445 0 200 13 653 5 088 0 018 0 203 10 457 20 676 8 108 t 0 010 2 548 10 254 0 45 20 015 1 142 2 0 3811 0 050 1 270 Dual4n Line Package Order Number LF412ACN or LF412CN NS Package Number NOBE RA PRY 12 53 E4 Relevant Syringe Pump User Manual Page
55. tch and Labview software to control the system Each of these components required innovative thinking to improve cell differentiation with OFIS TABLE OF CONTENTS Introduction E Ne Chaniiel PabrICdUOD u HI DEP Chip Micro Fluidic Ports EE Y EE S HP ERE IV Flow Control E Eo S ER EP TOS Edd V Detection duri MA EE VIP IDE u sore resi wot oce bea Su ales ea rave VIII Work du D eee sos I Appendix Appendix B B dgel 2 n Re DA EE Appendix Peripherals a AN segs tuu es Appendix D Project Tips and Tricks ere Appendix E Data Sheets and User Manuals EU Pera P CLA DLE cs D duc gru cs EUN 2 Ec ERNEUT Figure gm aa PAgure D2 T NS Nu UT ME S2 EET M MER NM E MNT aaa tte batt cae lui mc M LT Figure SG maya P Pr D upya Dans a ee Com has Oy Saad eee FISUrE SS NER Figure UE leur M
56. ter of the drilled holes and pressed firmly into place After this the nanoports are cautiously placed on the adhesive o rings making sure to center the nanoport over the hole Again this is pressed firmly into place Once complete the chip is clamped to the nanoport for one hour to allow the adhesive o ring to dry and seal The figure below shows the placement of a nanoport TOP VIEW SIDE VIEW drilled hole nanoport J B channel pem adhesive from pump waste o ring e o nanotube Figure 3 2 Nanoport Attachment Second Semester Updates The company that made the adhesive o rings no longer produces them Their replacement o rings are non sticky and bond via heat The process for attaching a nanoport is nearly identical however the final step is to bake the chip in 170 C for 1 hour Placing the o ring and nanoport is much more difficult because the new o ring is not sticky The bonding strength of the new o ring is very dependent on the cleanliness of the surfaces that are being bonded A detailed description of the process of attaching the new nanoports is in Appendix E Chapter IV Flow Control An essential part of the project is to be able to get cells into the chip There are several considerations for acceptable means to accomplish the desired flow rate of cells in the channel of the chip The first and primary requirement is that the flow rate must be less than or equal to 40 um
57. through a neutral body Internal to the neutral object the molecules within the neutral object polarize much like the depletion region that occurs when a voltage is put across a pn junction The imbalance of localized charges results in a virtual electric dipole being formed within this neutral object Therefore the electric field will apply a force to the surface dipoles in the overall neutrally charged object Since the objects we are concerned with are on the order of 10 um in diameter it is easier to exert a significant enough force on these small particles with large surface to volume ratios to push them into a suitable area to take a spectral reading 19 Figure 6 1 Picture of DEP trap electrodes In order to achieve a non uniform electric field a time varying voltage must be applied across the electrodes of the DEP trap The approximate force applied to an object being trapped can be defined by the formula F 2r R Re CM w x VE r 1 where F refers to the dipole approximation to the DEP force refers to the permittivity of the medium surrounding the object being trapped R is the radius of the particle r is the spatial coordinate is the angular frequency of the applied voltage E is the complex applied electric field CM is the Claussius Mossotti factor The Clossius Mossotti factor is a frequency dependent function of the permittivity of the medium outside the particle and the inside of the particle In si
58. timode optical fibers from 50 125 to 100 140 microns ST is a registered trademark of AT amp T Optek Technology Inc 1215 W Crosby Road in ST Receptacle 086 56 UNC 2B ANODE CATHODE 235 5 97 MAX 100 2 54 DIA 375 9 53 E 375 24 UNF 2A THREAD 155 3 94 810 20 57 730 20 07 L sud DIMENSIONS ARE IN INCHES MILLIMETERS Absolute Maximum Ratings Ta 25 C unless otherwise noted Reverse Voltage isi sae 100 VDC Continuous Power 200 mw Storage Temperature Range 55 C to 100 C Operating Temperature Range 40 C to 85 Lead Soldering Temperature 1 16 inch 1 6 mm from case for 5 sec with soldering iron 2409 c 2 RMA flux is recommended Duration can be extended to 10 sec max when flow soldering 3 Test Vn 5 V with 50 125 micron 0 20 fiber 10 uW optical power 850 Responsivity levels apply to 50 um 62 5 and 100 um core optical fibers Carrollton Texas 75006 8 90 972 323 2200 Fax 972 323 2396 40 Type OPF482 Electrical Characteristics Ta 25 C unless otherwise noted SYMBOL bee x PARAMETER MIN TYP MAX UNITS TEST CONDITIONS R Flux Responsivity 0 45
59. trometer was used to maximize the light intensity from the LED light source as well as the focused LED light after it has passed through the lens The resulting spectrum also provided verification that the microscope optics were properly aligned Narrow peaks in the spectrum indicated proper alignment from the LED light source to the microscope 34 Appendix D Project Tips and Tricks D1 Attaching Nanoports The recommended procedure is below 1 Clean the chip around the drilled hole and the bottom of the nanoport using acetone methanol and de ionized water It must be cleaned in this order using 3 different Q tips to scrub the surface The next cleaner must be applied before the previous one dries Putting the specimen to be cleaned in a glass dish is usually easier in that there can be an excess of the cleaner making it take longer to dry It must be glass because plastic reacts with acetone Use an air gun to blow the surface dry before the de ionized water dries Remember to wear gloves when working with acetone and methanol Flip both the chip and the nanoport upside down so that the surface of the chip that will be bonded is facing downward and the surface of the nanoport that will be bonded is facing upward Center the o ring on the nanoport surface and bring the nanoport into contact with the chip centering the nanoport over the drilled hole on the chip Notice that the o ring should be between the chip and nanopo
60. u City 435 8556 Japan Telephone 81 053 434 3311 Fax 61 053 434 5184 www hamamatsu com USA Humametsu 380 Foothi Bex 60 0 4 2 08807 0010 USA Telephone 5 006 257 0060 1 908 251 1218 Hamamatsu Deutschland Omah tQ D 52211 Herring am Aimee Germany Telephone 49 05152 3750 Fas 40 08152 2668 Fresco Phonics Pante SARL 19 Roe du Saule Tapu Parc du Moulin de Massy 01822 Massy Codax France Telephone 35 11 69 53 71 CQ 35 7 69 53 71 10 United Hamamatsu Phonics UK 2 Howard Cour 30 Tewin Road Vielen Carden City Hertfcidahos AL 18 Uniled Kingdom Telephone 44 1707 204858 44 1707 325777 Euscpe Hamamatsu Notes AB Seidevdgen 12 SE 7 41 Soha Sweden Telephone 45 5 509 031 00 Fax 48 amp 500 031 01 Photos SRI Stivie deba Mots UE 20020 Mesa Miera ay Teleposs 39 02 995 81 733 Fax 27 00 036 81 741 Cal No KLED1C21E02 Jun 2006 DN 3 ACKNOWLEDGMENTS We would like to thank Dr Kevin Lear PhD for his help and guidance as well as four of his graduate students Bob Pownall Hua Linda Shao Sean Pieper and Weina Wang for all of their help 58
Download Pdf Manuals
Related Search