Home

Primer on Flame Ionization Detectors

1. SSS gt m
2. ji A AIT C Data Rate 100 HzData Rate 50 Hz Data Rate 25Hz Data Rate 12 5 HzData Rate 6 3 Hz Data Rate 3 1 Hz Data Rate 1 6 Hz Data Rate 0 8 Hz Noise 56 Noise 48 Noise 44 Noise 38 Noise 35 Noise 28 Noise 24 Noise 17 Nois ryus 3 Nois rys 6 1 Nois rys 5 5 Nois rys 4 8 Nois rys 4 4 Nois rays 4 1 Nois ryus 3 1 Nois ryvs 2 4 0 5 10 15 20 25 30 35 Minutes 40 Figure 31 Detector noise at various data rates is shown over 5 minute intervals to illustrate the suppression of random noise through selection of the Detector Bunch Rate in the Workstation Detector temperature 250 C IX Summary of Optimum Parameters
3. or for recording devices that mandate a voltage signal The available range settings are 9 10 11 and 12 to represent the exponent in 10 Amperes milliVolt Range 9 is the least sensitive and range 12 is the most sensitive and the latter always should be set when using the direct conversion to digits for full range signal coverage equivalency from 100 microVolts full scale to 1000 Volts full scale sic with a Bruker Workstation Flame ionization detector is a mass flow detector where the mass flux through it is what is detected and its response is independent of analyte flow rate through the flame jet the same electrons are generated whether passing through the detector slowly or quickly As a result its signal is not severely impacted by changes in flow and valve upsets By contrast a thermal conductivity detector is a concentration detector where its response is very dependent on flow rates through it a slower flow allows the analyte to reside in the detector longer and hence yielding a bigger response The thermal conductivity detector is very sensitive to changes in flow and requires special considerations with valved systems especially involving column switching and backflushing actions A good performing flame ionization detector is a near perfect carbon counter for hydrocarbons including alkanes alkenes and aromatics Its response per molar concentration for propane with three carbons is half that of hexane with six
4. 2005 p 25 Sternberg et al Gas Chromatography p 265 266 Reactions involved in the burning process are listed by Sternberg et al including primary reactions matrix reactions giving energetic intermediates and ionizing reactions using excited components from the matrix and are repeated in Table These authors also discussed results of experiments performed with differing carrier gases variations in flow rates and inclusion of premixed oxygen with hydrogen fuel Some of these experiments are reproduced here with modern detector designs and reported below with some differing outcomes undoubtedly due to improvements in detector design over the years Table Chemical Reactions in Flame lonization Typical Primary Reactions 1 Cracking CoHes gt 2CH3 CH3COCH3 CH3 CH3CO CH3 2COH gt CH3 2 CH2 H20 2 Stripping H C2Hs gt C2H7 Ho OH CH3 gt CH2 H2O H CH gt C H2 Matrix Reactions in Hydrogen Flame H OH gt H20 H O M OH M O O M gt O gt M O HO2 gt OH O2 Possible lonization Reactions CH2 OH gt CH2OH e CH OH CH20 amp gt CHO H e CH O2 gt CHO e CH H20 gt CH2OH amp gt CHO H2 e CH CO gt CHO CO e C OH gt CHO e C O gt CO e C H20 gt CH20 e CCI OH gt CHO Cr CCI Oo gt CO CI CO Cl e CCI H20 CHO HCI amp M reaction chaperone invo
5. ebramstoncook msn com
6. ppbCarbon ml for gases and milliVolt sec nanogram for liquid injections Here a bigger number generates a better response This parameter can be used to estimate the size of a peak for a given hydrocarbon Peak Reponse Sensitivity x Analyte Concentration x Injected Volume For example the Bruker 450 GC Gas Chromatograph with a 0 02 flame tip generates a sensitivity of 6 27 microVolts sec ppbCarbon ml fOr siema sansamansys Propane volatile hydrocarbons with nitrogen as the make up gas With an injection volume of 300 ml and a hydrocarbon concentration of 40 ppbCarbon the expected peak area becomes 75 120 microvolt sec Typical sensitivity for the 0 01 tip is 9 4 microVolts sec ppbCarbon ml for volatile hydrocarbons For liquid injections into a capillary injector sensitivity is usually gt 33 000 milliVolt sec nanogram with helium as the make up gas and gt 66 000 milliVolt sec nanogram with nitrogen make up Linear Range A clear advantage of the flame ionization detector over other chromatographic detectors is its ability to properly measure very low concentrations typically around 50 ppbCarbon as well as high percent levels with near Figure 12 Cartesian plot of Area vs perfect linearity Concentration showing compression of points at low concentrations Cartesian plots as shown in Figure 12 display results nicely for limited concentration ranges typically within a factor of ten but become congested at the
7. to position the collector tube in place and to hold the probes in place Overall performance of the detector is critically dependent on the cleanliness of these ceramic pieces a They must remain in their pristine condition and the white parts f faem E eae ake should not be touched by the operator with anything but cotton 5 gloves McMaster Carr P N 9729111 large T12 medium tel 562 692 5911 Oil from fingers on the ceramics becomes extremely difficult to remove even with heat Often contaminated parts are easier to replace with new ones instead of attempting to clean them Figure 38 Location of Ceramics in Tower Flame Tip The flame tip assembly in the Bruker 450 GC is comprised of a ceramic tip stainless steel detector base titanium nut and usually a graphite Vespel ferrule They all have differing thermal expansion coefficients Repeated heating and cooling of the detector cause the various parts to act together in loosening up the flame tip nut To avoid catastrophic failure in the performance of the detector the flame tip should be checked monthly for snugness A 3 8 hollow shaft nut driver McMaster Carr P N 8358A27 tel 562 692 5911 is the best tool to ensure that the nut is properly tightened Cracked Burned Ferrule in Flame Tip Difficulty in lighting the flame and degradation in detector sensitivity are often caused by a deteriorating ferrule holding the flame tip in place A crack here will allow hydrogen and
8. Controller DEFC These values then adjust flows to the detector If actual flows do not match their setpoints the status light nearby will indicate a fault condition Upper range and tolerance for each gas are listed below Flow Range Flow Rate Ready Tolerance Actual vs Detector Gas ml min ml min Setpoint Make up Gas Label The choice of make up gas affects the calibration of the Detector Electronic Flow Controller by the thermal conductivity of the gas and must be specified in Setup Detector This entry is then displayed in all associated detector screens If hydrogen is used for reduction catalyst feed as EFC11 Make up then helium is selected and desired flow must be measured with a separate flow meter and the setpoint adjusted to the desired hydrogen flow rate Figure 18 Gas employed as make up is specified in Setup Detector FID TSD Detector Full Scale The Bruker 450 GC flame ionization detector has the ability to handle signals from 100 microVolts full scale up to 1000 Volts full scale without need to change detector ranges Unfortunately the preset parameter to activate this mode is set to 10 V and should be changed to 1000 V for the appropriate detector position in the Workstation method Figure 19 Workstation entry for Detector Full Scale of 1000 Volts See discussions on detector time constant by R Bramston Cook Peak Detection with Varian Star Workstation for Varian 3800 and 450 GC
9. For the 0 02 flame tip maximum response is generated with these flame parameters 40 ml min 35 ml min 300 ml min For the 0 01 flame tip maximum response is generated with these flame parameters 20 ml min 25 ml min 300 ml min X Troubleshooting The flame ionization detector is undoubtedly a simple detector to diagnose problems as it is not very complicated compared to other chromatographic detectors especially chem illuminesence Hall electrolytic conductivity and pulsed flame photometric detectors Flame Ignition Flame ignition is attempted whenever a method is activated when flame out is sensed by the instrument or manually by the operator using the Ignite button on the screen The ignition process makes three attempts to light the flame If ignition is confirmed by the system then the process is terminated and normal operations are allowed to proceed If not a fault is triggered and operations are halted After correcting the issue and reinitiating the ignition process with the Ignite button in the Detector section of the Bruker 450 GC the error is cleared After ignition is confirmed an automatic 60 second delay is initiated to help recover from the ignition process before the system displays ready FID Auto Ignite Figure 32 Flowchart for automatic process of flame ignition Confirmation of Flame Ignition Available means to sense for a functioning flame are e Instrument sensing backgr
10. Gas Chromatographs Lotus Consulting 2010 p 12 available on request from Lotus Consulting at ebramstoncook msn com These parameters do not apply to manual controls of detector gases R Bramston Cook Processing Very Tiny and Huge Monster Peaks in the Same Chromatogram Lotus Consulting 2009 available on request from Lotus Consulting at ebramstoncook msn com 25 27 Time Programmable Parameters Several detector settings are time programmable to permit changes in operating range and to execute an automatic zeroing of the signal at user selectable points in the chromatogram Initial programming must be set prior to start of data collection but additions can be made for times after the current progress time Range Amplitudes of the chromatographic peaks are directly impacted by this range setting The Bruker 450 GC has the ability to set the full scale range to allow voltage processing through a conventional analog to digital converter ADC or for recording devices that mandate a voltage signal The range settings are 9 10 11 and 12 to represent the exponent in 10 Amperes milliVolt 2 Range 9 is the least sensitive and range 12 is the most sensitive Range can be time programmed to keep various sized peaks on scale over the chromatogram especially with analog recorders With a Bruker Workstation connected directly to the Bruker 450 GC the range setting always should be set to 12 to enable detection of
11. carbons To allow direct correlation of differing hydrocarbons molar concentration units for example ppmVolume are converted into ppmCarbon by multiplying the molar value by the number of carbon atoms in that analyte An exception to this rule is the response of methane relative to the other hydrocarbons methane gives an enhanced signal for its carbon content by typically 13 over the others R Bramston Cook Processing Very Tiny and Huge Monster Peaks in the Same Chromatogram Lotus Consulting 2009 available on request from Lotus Consulting at ebramstoncook msn com 10 MilliVolt is included in this unit from the historical use of analog recorders that had a full scale span of 1 mV full scale Many gas chromatographs including the Bruker 450 GC report peak signals in units of Volts to correspond to the output of the usual electrometer while others use the original signal units of Amperes The interconversion between the two depends on the range setting of the electrometer At a range of 10 Amps mV microVolts is equivalent to femtoAmperes Corresponding area units are microVolt seconds and femtoAmpere seconds both are equivalent with a range of 10 Amps mv Il Construction of a Flame lonization Detector The essence of the flame ionization detector design has remained virtually unchanged for four decades Inclusion of solid state electrometers with Junction Field Effect Transistors JFET through the introduction of
12. exhibit non linearity at the lower end of the scale due to component loss from active sites in the measurement system Measurements can be made into these stages but must be labeled as outside linearity VI FID Linear Range Outperforms Chromatography Chromatographic operating conditions are always a compromise between resolution of peaks speed of analysis and sample capacity Parameters can be selected to achieve optimum performance for any one of these or a compromise with selection of any two but performance for the others are severely compromised For example best resolution of peaks is achieved with a long narrow bore thin film capillary column with low carrier flow rates of nitrogen but speed and sample Capacity are severely compromised under these conditions Fast chromatography requires short thin film columns and high flow of hydrogen whereas sample capacity or how much sample can be loaded onto the sample needs wide bore thick film columns with moderate flow rates Optimum conditions for resolution or speed will cause peaks to be readily overloaded and thus skew the calibration curve s linearity Chromatographic conditions often cannot use the full linear range of the flame ionization detector but having this capability provides assurances that the detector does not become the limiting factor Other variables can dramatically impact linearity including accurate standard preparations over the complete range reactive sites
13. helium or even better with hydrogen Addition of Internal Oxygen Most modern flame ionization detectors now employ a true hydrogen diffusion flame with only hydrogen and inert gases flowing through the flame tip When small amounts of oxygen are added at this point along with carrier and make up gases a smaller inner cone develops in the flame just above the tip Heat generated during this process leads to more thermal cracking but the loss in fuel in the inner cone and pre oxidation of analyte molecules reduces the response to combustion in the outer layers of the flame Even with this limited combustion the inner cone B from Figure 1 retains its hydrogen diffusion character but this area is altered by the decrease in available fuel and a buildup of combustion products from the premixed oxygen hydrogen flame Possible enhancements with the addition of internal oxygen are very compound dependent with the higher inner flame temperatures and the presence of more reactive hydrogen atoms yielding bigger signals for selected analytes notably for methane This boost lessens for higher alkanes where oxidation occurs more readily to cancel the enhancements from cracking and stripping A notable example of this degradation is carbon tetrachloride where it generates a response at about half of methane in a normal hydrogen diffusion flame but virtually 130 disappears with addition of internal oxyge
14. houses promote a stainless steel version as a preferred substitute for this brass one but the potential for galling of stainless stainless connections or cross threading of the fitting with its mate can result in damage to the detector crossover and could mandate an expensive replacement of the detector base The brass version is strongly suggested for normal operations and a special stainless steel version P N CP 743117 is available for very high temperature detector operations above 350 C Bruker recently introduced this special stainless steel nut P N CP743117 for very high temperature applications where the detector temperature is maintained above 350 C but is not recommended for normal applications below these temperatures due to a chance for galling or cross threading Purity of Gas Supplies Hydrocarbons impurities in any supply gas can be major contributors to degradation in performance of the flame ionization detector Even a purity of 99 995 commonly labeled as chromatographic grade can possess 50 ppmV hydrocarbons and still pass this specification Ultimate performance is achieved with 99 9999 purity often labeled as research grade In line hydrocarbons filters Bruker P N 392594492 each with connecting unit for carrier make up and hydrogen are recommended to aid in reducing contaminates further Packed Column Installation The crossover connection for attachment of columns to the detector inside t
15. the Varian Model 1740 Moduline C in 1968 made this detectors performance maxed out at its theoretical limit with the electrometer no longer a noise contributor Few such dramatic improvements have occurred since With electrical currents as low as 10 Amperes Sa popu capable of being measured all efforts to insulate this signal from external influences must be employed Ceramic insulators are used in four a A locations to ensure that generated electrons are fully a pm Dy directed to the collector and on to the electrometer Flame Tip Nut T nitor p a without interference The flame tip is electrically insulated with ceramics to eliminate any noise A p Clunen Wasnel enerated by voltage spikes and electrical i 9 y 9 p Tame Mp akaaka j Flame Tip Ferrule Collector ps w i e ee ene d i disturbances on instrument chassis ground The collector probe has gold plated connectors to ensure accurate measure of extremely low currents Column Oven Detector Cross over generated in the detector And this probe and Figure 2 Cross section of Flame associated collector cylinder are electrically isolated lonization Detector in the Bruker 450 from the instrument with ceramic insulators as well GC Gas Chromatograph Bruker flame tip designs have patented small exposed ceramic insulation at the jet exit to reduce interaction with analytes at this critical junction Figure 3 illustrates the enhancements generated with use of this cer
16. this process A Detector Bunch Point is an arithmetic average of the user specified number of raw data points and effectively changes the data collection rate This point is then transmitted from the gas chromatograph to the Workstation Peak to peak noise is readily determined by visually picking out the highest and lowest noise excursions and then taking the difference in microVolts or picoAmperes Another approach to determining noise values is to compute the root mean square RMS value for noise over a time interval by RMS Notse or one standard deviation from the mean and where y is each data point over the interval The relationship between RMS noise and peak to peak noise is usually considered to be a factor of 6 to include 99 7 of the excursions Direct comparisons of noise characteristics among different detectors of the same style especially differing brands are impacted by the signal processing conditions especially by analog dampening with a resistor capacitor network and digital filtering applied to the detector signal Any use of noise must always include conditions used to generate this value especially ones with extreme averaging that can distort chromatographic peak shapes Elevated background signals directly impact the level of noise with the flame ionization detector High background signals can be generated from impurities in any of the supply gases column bleed or heavy compounds from previous sampl
17. 150 C a a Wiha iy AV i a Wyn NY AWA OANT Wiat 100 C 0 0 2 0 4 0 Minutes Figure 6 Noise level on the detector is directly impacted by the temperature at its base Data collection rate for all segments displayed is 12 5 Hz Flame Background Peak to Peak Noise Root Mean Square Noise Detector Temperature Table Il Summary of impacts from different detector temperatures on noise levels 19 20 If all column eluents have low boiling points such as the hydrocarbon range from C4 through Cs then a lower detector temperature results in lower noise and improved detection limits and detectivity With higher boilers a simple recovery test with different temperatures can ensure that a lower temperature does not affect their responses Realtive Peak to Peak Noise 150 200 250 300 FID Base Temperature C Figure 7 A plot of relative peak to peak noise versus detector temperature illustrates the noise improvement with lower temperature settings All noise levels are normalized to the level at 100 C 350 Bruker Model 450 GC User Manual Bruker Chemical and Applied Markets Fremont Ca p 262 Dependence of diffusion coefficients on temperature for gases can be expressed using the Chapman Enskog theory by T where T is temperature in K Reference en wikipedia org wiki Mass_ diffusivity accessed 5 23 11 Drift Stability of the detector signal is critically important during measuremen
18. 87600 Large Stopper P N 394958600 Air Measuring Plug P N 1600050500 Flowmeter P N 190011500 Detector Flow Measuring Kit includes flow tube large stopper air Leak Detector P N CP87610 GC Column Scale P N 392575001 GC Column Scale Kit includes scale 2 nuts 2 _ Column Nut 0 9 mm bore brass P N 394955100 Column Nut 1 6 mm bore brass P N CP742351 29 Column Nut 0 9 mm bore stainless steel for high temperature applications 350 450 C 2 pkg P N CP743117 Temperature Probe P N 392537401 Heater 120V 125W P N 392539601 230V 125W P N 392539602 101V 125W P N 392539603 Air Measuring Plug measuring plug and flowmeter P N 391725091 Flow Tube and a ee ferrules each for 0 4 0 5 and 0 8 mm ID P N 392575190 GC Column Scale FID Test Sample Capillary column 30 ng ul C14 C15 C46 in isooctane P N 8200504807 Packed Columns 300 ng ul C14 C15 C146 in isooctane P N 8200504800 Column Particle Trap 2 5 meters 0 53 mm ID P N CP4017 Hydrocarbon Filter P N CP17972 Single Filter Connecting Unit 1 8 fittings P N CP7988 Dual Filter Connecting Unit 1 8 fittings P N CP738407 45 Quad Filter Connecting Unit 1 8 fittings P N CP736520 The latest version of this column nut has a reduced opening to prevent ferrules from protruding through when tightened and will not accommodate 1 16 tubing nor two columns inserted into a two hole ferrule If a nut with
19. Primer on Flame lonization Detectors by Randall Bramston Cook Lotus Consulting 5 81 Campo Walk Long Beach Ca 90803 310 569 0128 ebramstoncook msn com June 4 2011 Copyright 2011 Lotus Flower Inc Flame ionization detection FID is a primary mechanism for measuring organic compounds eluting from gas chromatographic columns It is not as selective as other detectors such as a thermionic specific detector TSD often called a nitrogen phosphorus detector NPD but most organics will generate electrons from their combustion in a flame and this process creates a signal related to the analyte concentration Notable exceptions are carbon disulfide carbon dioxide carbon monoxide and formaldehyde Inorganic gases such oxygen and nitrogen do not produce significant responses with this detector but huge amounts passed through can cause a disruption in the flame process and yield a very small response especially when the bulk gas into the detector differs from the eluting bulk sample gas This detector has extensive use in monitoring hydrocarbons in a variety of matrices including gasoline diesel biodiesel jet fuel refinery gas kerosene crude oil fuel oil natural gas liquefied petroleum gas engine exhaust and ambient air Other applications include measurements of essential oils alcohols oxygenates volatile free fatty acids primary amines antioxidants phenols polycyclic aromatic hydrocarbons PAH and residual solvents e
20. Type 11 for this detector allows the operator to set flows either from the Bruker 450 GC front keyboard screen or from the Workstation The chromatograph then establishes the actual flows to agree with the setpoints and then documents flows in the run log of the Workstation Attachment of gas lines are shown in Figure 40 Usually H2 OUT and M U OUT are combined by a manifold into a single outlet connection With connection to a reduction catalyst two hydrogen flows are needed one at 10 ml min through the catalyst and the second at 20 ml min to provide a total flow of 30 ml min to the T j detector This manifold can be inverted to provide a single input Ri me oe it of hydrogen into both hydrogen and make up controllers and f E m u out Hy OUT Air OUT then two outlets for the catalyst The 450 GC Setup does not list AE oam cH H2 as a possible choice for M U gas instead He is selected and k Me i actual flow measured with an external flowmeter and the Figure 40 Gas connections parameter adjusted to get the desired flow into Detector Electronic Flow Controller Type 11 o For packed columns this Valco adaptor is P N ZRU41C Any megabore column or narrower is appropriate here as its effect on chromatography is overwhelmed by the packed column 4 Varian Model 3400 3600 3700 6000 Gas Chromatographs are configured for a distance of 11 5 cm from the bottom of the nut to the tip of the column that goes int
21. a wider hole is not available then the narrow one can be drilled out with a 1 16 bit to provide adequate clearance or the 1 6 mm bore version P N CP742351 can be substituted Postscript In my literature search for studies on the chemistry of the flame ionization process encountered a complete treatise of the topic in an early publication by J C Sternberg W S Gallaway and D T L Jones The Mechanism of Response of Flame lonization Detectors N Brenner et al ed Gas Chromatography Academic Press New York 1962 that provides a basis for much of the theoretical discussions in this monograph Since the original text from 1962 is not readily accessible today portions of their manuscript are paraphrased or quoted directly here to pass along their contributions to the science that still apply today This author thanks Ed Nygren James Pachlhofer Rand Mahoney Wilco Hoogerwerf Jos Curvers Paul Van den Engel Hans van den Heuvel and Nijsse Kees of Bruker CAM Jack Marrin of Scott Marrin Inc and Herb Neumann for assistance in preparing this monograph Copvright 2011 Lotus Flower Inc Screens are copyrighted by Bruker Corporation and are reprinted reproduced with the permission of Bruker All rights reserved Bruker and the Bruker logo are trademarks or registered trademarks of Bruker in the U S and other countries 5 81 Campo Walk Lotus Consulting Cam 310 569 0128 Long Beach California 90803 email
22. age If a fault occurs it can be cleared by correcting the cause of the failure and reigniting the flame A capillary column is inserted into the detector cross over at the roof of the column oven and reaches almost to the top of the flame tip to maintain an inert surface through the detector bowels This distance for the Bruker 450 GC and Varian 3800 is 9 5 cm from the bottom of the brass column nut to the tip of the column to be inserted This cross over specifically has low thermal mass to minimize transfer of heat from either zone to the other Performance of this detector is very dependent on the geometry of the flame The standard 0 02 flame tip orifice Bruker P N 200193800 is useful for most applications All published specifications for the Bruker 450 GC Gas Chromatograph are based on this flame tip When added detectivity is required a narrower tip of 0 01 Bruker P N 200187500 can enhance peak sizes due to the tighter hotter flame generated However this tip gives a bit more unstable flame and is more susceptible to flame outs especially with packed columns and valved systems Special precautions with added flame out restrictors are frequently needed to eliminate this issue lil Even Response Factors for Hydrocarbons Accurate standards are often not available for every possible hydrocarbon found in samples Reliance must be made on the uniform response of flame ionization detectors with hydrocarbons Then the response fact
23. amic tip design over metal tips run under Metal identical conditions and scaling Peaks for high molecular Flame Tip Meat weight hydrocarbons are about 5 10 times larger and extend ATVI out to C 96 Also metal clad columns for high temperature applications can be positioned just below the ceramic orifice to i avoid danger of shorting out as expected with a metal tip and JECT to mitigate sample fragmentation from pyrolysis on the interior surfaces Ceramic Flame Tip III II III JECT A graphite Vespel ferrule Bruker P N CR213200 for detector temperatures to 400 C or graphite only ferrule Bruker P N CR211200 to 450 C for very high temperature operations provides a seal between this flame tip and its ae base And a small washer is inserted between this ferrule and its nut to provide an unobstructed pathway for the air flow to be directed around the flame jet Figure 3 Chromatograms of Polywax 655 with different tips 11 See discussions at en wikipedia org wiki JFET ie Berg J Wells G High Temperature Flame Jet for Gas Chromatography U S Patent 4 999 162 Mar 12 1991 S Graphite ferrules are very soft and pliable and more likely to get stuck inside the detector base The graphite Vespel ones should be used whenever detector temperatures do not exceed the ferrules temperature limit of 400 C To allow measurement of carrier and flame gases through the detector an aluminum cr
24. backgrounds to preclude the error but still maintain monitoring of a lit flame oe E Check for Flame out in Detector Setup some detectors utilize the FID electrometer to process signals through the Bruker 450 GC including photoionization detectors PID and pulsed discharge detectors PDD These detectors do not have a flame per se and need not have the built in flame out sensor active especially when a fault iS emmma triggered by very low backgrounds often generated by these detectors In Setup Detector the check for flame out can be disabled for that detector zone Figure 17 Enabling check for Flame out is chosen in Setup Detector Time Constant This parameter is displayed but has no role in data processing in 450 GC Firmware Version 1 07 and later Now all dampening is performed digitally by adjustment of data collection bunch rates from the Workstation Detector EFC Enable For active operations of flows to the detectors with a Detector Electronic Flow Controller DEFC the Enable box must be checked If unchecked the system will not wait for flows to match setpoints and the associated electronics are automatically disabled This parameter cannot be set through Workstation and method downloads from Workstation will not impact this setting Detector EFC Set The operator can set detector flows through the entries for Make up Hydrogen and Air with the inboard Detector Electronic Flow
25. bleed and impurities from supply gases are minimal and detector noise generated is from the flame itself this noise is related to the temperature of the detector base and tower 60 ilh Ih Lait W Nid Labia M y l Mi i I illl rn hai cat jili j Ii SNM AN HE a Ay AYN yh 250 O 1 Holl BRU 307 ii Ndi a hial AII vA diy yu uu l lha S AYN Aa N 300 2c Lo il yy li yarn y igbai iy earn n MMi Ai PN AWN AMMAN 250 C APPIN A FM ai 0 gt HMPA aN Aaya AMOR 200 C ny Nth iAy vs hi WW At hihi lik HN Wa Hint w
26. cially proper allocations of areas or heights For ultimate detector performance the detector temperature and lit flame should be allowed to stabilize for at least an hour before making runs Ultimate constancy is frequently achieved only after extended standby conditions above operating temperatures often overnight The baseline should be monitored with an expanded ordinate scale to match Figure 8 Column Bleed Many capillary columns now are better constructed to significantly minimize column bleed into the detector especially ones labeled for mass spectrometric applications However some phases can generate significant bleed especially polar ones especially as the column oven temperature approaches the column maximum To avoid detector contamination from the column phase the column ought to be removed from the detector during the initial column conditioning while the column is heated toward its upper limit and then later reinserted into the detector This effect can be negated by executing baseline subtraction operations with the workstation No Peaks or Smaller Peaks than Expected Absence of flame noise and signal background is an indication that the flame is out the electrometer is not turned on refer to diagnoses above or a detector cable is detached Numerous causes for no detected peaks or suppressed sizes also can be attributed to sampling and chromatography issues beyond the scope of this discussion Missing peaks or smaller peak
27. e Hydrogen atoms oxygen atoms hydroxyl and hydroperoxy radicals are continually produced in the reaction zone e Sample molecules introduced into the flame undergo cracking and stripping reactions to form reactive hydrocarbon fragments which diffuse back into the inner cone of the flame e Chemi ionization reactions are exothermic oxidation reactions occurring in the reaction zone between reactive hydrocarbon radicals and the excited molecules and radicals from the matrix flame 1 References to detector design and performance relate specifically to the Bruker Varian 450 GC Gas Chromatograph with Firmware Version 3 03 and most also apply to the Bruker Varian 430 GC and Varian 3800 3900 Gas Chromatographs Older versions of 450 GC firmware may present subtle differences to what is discussed here Generic comments are also valid for older Varian and other manufacturers detectors a Early flame experiments employed carbon monoxide as the fuel but were soon abandoned undoubtedly by the obvious toxic exposure of the fuel to the chromatographer Slee Sternberg W S Gallaway D T L Jones The Mechanism of Response of Flame lonization Detectors N Brenner et al ed Gas Chromatography Academic Press New York 1962 p 249 t RJ Reed North America Combustion Handbook Volume I North America Mfg Co Cleveland Oh 1986 p 10 gt J V Hinshaw The Flame lonization Detector LCGC North America Vol 23 Issue 12
28. eased These impurities can mask desired peaks eluting from the column A dirty detector can be triggered by a number of factors including impure supply gases dirty gas lines into the instrument fingerprints on internal parts column bleed and condensation of high boiling compounds or combustion products from previously injected analytes Even unexpected late elution of compounds from earlier runs can be generating elevated backgrounds Continually monitoring of the background signal over time can provide an indication if the condition is improving with a declining signal Often during bake out the signal will rise and then eventually start falling and then finally stabilize hopefully at a low value If the signal does not drop to an optimum level discussed above then the background must be coming from a constant source such as impurities in detector gases or column bleed Insertion of hydrocarbon filters on carrier hydrogen and make up gases can dramatically reduce this source of background All tubing from the supply tanks to the instrument must be chromatographic grade stainless steel Supelco P N 20526 U tel 800 247 6628 or copper and NEITHER refrigerator grade from a hardware store as this tubing is full of oil to keep the copper shiny look NOR plastic tubing of any sort they readily release monomers and plasticizers and often are permeable to oxygen and water from room air should be installed If either dirty tubing or plas
29. eme caution must be exercised to avoid any possibility that the lubricant is applied to any area in the detector that could possibly reach the active flame zone Removal and Reinsertion of Probes The ignitor probe has two prongs that are in contact with the flame tip metal sheath to polarize the tip Before the tower is to be removed this lower probe must be disengaged from the tip by pulling the ignitor a bit out of its position to avoid distorting the prong shape during servicing The upper signal probe can remain in place unless the collector needs to be extracted for inspection or replacement As the probes ae are pulled out their red silicone o rings will come into view Figure 39 Probes are lined up with indicating that the probes have been pulled out far enough tabs and notches on their sides If these o rings exhibit scrapes or gouges they should be during insertion replaced After detector servicing these probes need to be reinserted into the tower To maintain proper orientation with the flame tip or collector each probe has a tab on one side to line up with the matching notch on the side of the tower side arm To ease the o rings into position the probe can be rotated slightly but must come into final alignment when approaching the ultimate spot Brass Capillary Nuts Bruker intentionally uses brass ferrules for sealing columns to the detector to avoid seizing of the nut to the cross over Several chromatography supply
30. er Bruker P N 190011500 to validate detector flow settings Since four separate flows all converge in the detector make up hydrogen and air are initially all turned off so only column flow is measured at the top of the detector using flow tube P N 200187600 inserted in the large silicone stopper P N 394958600 with a flow meter Then make up is turned on and the combined carrier and make up are measured Then hydrogen is added in and verified and finally air is joined in and confirmed a The air measuring plug P N 1600050500 can be installed in Figure 33 Flows at the detector place of the flame tip assembly to be able to measure only air can be measured by inserting a while all others are blocked With a capillary column installed flow tube and stopper at the top the column must be lowered a bit as this plug will not fit with the of the detector tower column penetrating into where the plug will be placed This obstruction can also be used to pressurize the column system to test for leaks A caution here is that if hydrogen and make up are flowing with tank pressures below the carrier pressure blocked flow at the detector can back up into these streams and the leak test never realizes the steady maximum pressure expected Instead the hydrogen and make up settings must be set to zero to prohibit this backflow Flows into the detector can be assessed by measuring Figure 34 Air Measuring Plug total flows with the fla
31. es continuing to elute In a clean system noise is curiously related to the temperature of the detector perhaps lowering diffusion rates of hydrogen and oxygen atoms within the flame cone from lower temperatures as illustrated in Figures 6 and 7 and Table Il See discussions on finite impulse response FIR by R Bramston Cook Peak Detection with Varian Star Workstation for Varian 3800 and 450 GC Gas Chromatographs Lotus Consulting 2010 available on request from Lotus Consulting at ebramstoncook msn com Peak to peak and RMS noise values can be calculated by converting raw data to ASCII with Bruker MSWS Software and then copying pasting the data into Microsoft Excel Peak to peak noise for a selected region is obtained with MAX and MIN functions and RMS noise is determined with STDEV function 18 M Duff Analog Devices Inc www youtube com watch v KcODSYXiZA accessed June 4 2011 Typically the detector temperature is suggested to be set at 50 C over the maximum temperature of the column oven to ensure that all eluents are kept in the vapor state and do not condense in the detector 1 If column bleed is low and purities of supply gases are appropriate low levels of analytes injected into the column are most likely to remain volatile well below their elution temperatures due to their vapor pressures at lower detector temperatures However if the elevated background signal from column
32. es from both detectors on a single sample injection Thus for example inorganic gases such as oxygen and nitrogen can be measured together with hydrocarbons especially hydrocarbons below detection by the thermal conductivity detector Or aromatic and olefinic hydrocarbons can be picked out from other hydrocarbons with the photoionization and flame ionization detectors plumbed in series Carbon Dioxide Oxygen Nitrogen Volts Carbon Monoxide Methane Hydrogen Reduction Catalyst Conversion of Carbon ToD Monoxide and Carbon Dioxide FID Carbon monoxide and carbon dioxide are not j I l C e responsive to flame ionization but can be measured after their conversion to methane with a reduction catalyst of nickel nitrate The catalyst is heated to 380 thermal conductivity for simultaneous C and has continual flow of hydrogen 10 ml min to measurement of inorganic gases with provide a reducing environment to effectively convert TCD and hydrocarbons with FID An these two analytes into detectable methane included reduction catalyst allows Detection is down to less than ppmV and is linear up carbon dioxide and carbon monoxide into high percentage levels ia be Converted to methane for sensitive detection by FID Figure 5 A flame ionization detector can be plumbed in series with a Total Hydrocarbons Rapid screening tests for hydrocarbons can be performed by injecting an aliquot of sample directly to the flame
33. he column oven is designed specifically for capillary columns and requires an adaption for use with packed columns To minimize detrimental effects of unswept deadvolumes at this critical point a Valco reducer union for 1 8 packed columns Valco P N ZRU21C tel 800 367 8424 a fused silica adapter Valco P N FS1 8 for megabore columns and a short section typically 1 meter of megabore 0 53mm ID capillary column long enough to stretch from the end of the packed column into the detector can be combined to provide the connection Column Insertion into Detector Crossover Optimum performance is aided by keeping the pathway inside the detector inert up to the exit into the flame Obviously an inherently inert surface is the capillary column iz employed By inserting the column up into the detector and just below the orifice of the flame tip the analytes can avoid the reactive surfaces inside the detector This magical distance is 9 5 cm from the end of the column to the bottom of the column nut A convenient marker for this position is a small strip of tape that remains attached during installation but can be removed readily without leaving a trace or use of GC Column Scale P N 392575001 Note that the ferrule in this nut is reversed from normal to present a flat seal at the detector crossover for easier subsequent removal m 9 5cm Gas Connections into Detector Electronic Flow Controller Module Electronic flow controller
34. his tip is more susceptible to flame out from high flow rates from large solvent injections and from pressure surges generated by valve operations such as column switching with backflush to detector Use of this tip is recommended only to achieve ultimate performance in detection limits and sensitivities 1 00 9 1 00 y S z S 3 0 67 gt 0 67 UO Vo San D lt x 5 0 33 9 0 33 a 2 o 2 0 00 l l I I cd 0 00 l l i j l oo oc rd 5 10 15 20 25 30 35 40 45 50 5 10 15 20 25 30 35 40 45 50 Argon Make up ml min Make up ml min Figure 26 Relative responses for Figure 27 Changes in response with 0 01 and 0 02 tips over a range of different make up gases and make up flows various flow rates with 0 01 tip 2 1 00 2 S S 0 67 x q x g 0 33 g 0 00 E x x 100 200 300 400 5 10 15 20 25 30 35 40 45 Air Flow ml min Hydrogen Flow ml min Figure 28 Changes in peak size versus Figure 29 Relative peak sizes versus air flow in the detector with 0 01 tip hydrogen flow in the detector with 0 01 tip E a r anti Ca Effect of Bunch Rate on Noise assy male Kian Digital averaging of the signal occurs in two separate f sesamo functions within the Bruker 450 GC To avoid excessive noise especially at high data rates a 31 point Finite Impulse Response FIR filter is deployed during the initial signal processing and not adjustable by the user Further
35. in the sampling process and chromatographic system especially in the injector module and column and maintenance of appropriate loading into the column without overcapacity With the exceedingly wide linear range for the flame ionization detector its performance does not usually contribute to issues involving non linearity of the experiment from fast narrow bore capillary to high capacity packed columns other factors are likely to be the culprit Response factor is area counts divided by concentration of that calibration point Vil Bruker 450 GC Instrument Method Parameters for Flame lonization Detector Most parameters for the flame ionization detector can be set and conditions monitored through both the instrument screen and Workstation Both the touch screen on the Bruker 450 GC and Workstation are active until the instrument goes into run mode to collect data Once a run starts most parameters cannot be altered in real time until the run ends ws FID cp TSD oP tcp Analog Output Flame ionization detector conditions and programming on the instrument are accessed 7 i Enable Iv Time Range Autoz through the detector icon on the right side of the Be z gt Initial 2a w screen Temperature 500 Flameout Threshold 300 pY Time Constant Enable ug Eee Set Actual O Make up He 25 0 25 0 mLimin Combustion H2 30 0 30 0 mLimin Combustion ir 300 0 300 0 mL min Figure 14 Detector Ignite Clear Autoz Autozero Pa
36. ionization detector without chromatography The result is called Total Hydrocarbons THC Since this detector s response is even across the range of non methane hydrocarbons the signal can be accurately quantified against a propane standard V Performance Specifications Noise Proper signal processing of the flame ionization detector with modern electrometers should generate detector responses that are limited only by noise generated from the combustion process at the flame and not from the signal processing notably from the electrometer This noise then becomes the major limiting factor in how low concentrations can be measured Classically this noise value is assessed as the peak to peak variations over a reasonable time interval typically a few minutes Now with data systems processing digital values for the signal all sorts of signal averaging can be applied to yield any value desired for this noise value For example raw data collected at 100 Hz with Varian 450 GC could display some very noisy chromatograms and make sorting out legitimate peaks from baseline noise very difficult To reduce high frequency noise without significant distortion of peak shapes a digital 31 point Finite Impulse Response filter is applied to raw data points in the gas chromatograph during data collection In addition signal conversions can be averaged into bunched points to still properly describe the peak and allow more suppression of noise through
37. irmed when the signal shows some variations Background of the detector used for diagnostics can be viewed with Clear Autoz button at the bottom of the screen EFC Type 11 Actual Gas flows to the flame ionization detector are displayed as Actual If this value does not match the setpoint the status light in the area will display J to convey a problem and the instrument will remain in equilibrating This parameter does not apply to the manual control option for detector gases User Settable Parameters Enable For active use of the flame ionization detector this box must be checked or all parameters are ignored and the system will proceed even when actual detector conditions do not match setpoints The detector oven is also disabled With Workstation operations this parameter is labeled as Detector Oven Electronics Functional use of the detector mandates that the electronics must be enabled If this box is unchecked then the output signal is maintained at zero and flame out sensing is disabled If the Detector EFC is installed and not enabled this box is grayed out and cannot be turned on Temperature Setpoint for the detector temperature at its base is assigned here This temperature setting can range from 50 0 C to 450 0 C unless a lower maximum temperature is entered in Setup The choice of temperature does have impact on noise and needs to be warm enough to keep analytes water and co
38. is not enough to accomplish sufficient flow extra flow is provided through a make up gas added to the column effluent right at the detector base Hydrogen fuel is added in at this same location Combustion of organic molecules by the flame r Air generates electrons that can be collected and Cai ake up processed through an electrometer to give a signal that yields the amount of that analyte passing through A metal collector cylinder is placed just above the flame and a bias voltage of typically 180 Volts DC is applied between the flame tip cathode and this collector anode to direct electrons for processing into this signal Collector Probe Ignitor Probe Frequently the signal from the collector probe is directed to a conventional electrometer to convert the current generated into a voltage for processing by a data system This approach requires setting of the electrometer range to keep the voltage signal within span of the analog to digital converter for data computations typically 0 1 Volt or 0 10 Volts The Bruker 450 GC Gas Chromatograph instead processes the generated current from the detector directly into a number bypassing the voltage translation to permit coverage of the complete range of the detector without changing the full scale voltage range setting The Bruker 450 GC does possess an ability to set the full scale range to allow voltage processing through a conventional analog to digital converter ADC
39. justments Time Constant Slow Fast EFC Type Typell Make up Flow ml min H2 Flows ml min hir Flows ml min Figure 20 Parameter entries for flame ionization detector from MS Workstation method lgnite Clear Autoz Autozero J Figure 21 Active buttons in detector section of Bruker 450 GC MilliVolt is included in this unit from the historical use of analog recorders that had a span of 1 mV full scale Clear Autoz The natural background signal of the flame ionization detector is a useful assessment of the condition of the detector see discussions below This button clears the imposed zeroing of the signal and allows the raw signal to be displayed All actions with use of this button impact the data collected during a run If Autoz is enabled in the detector section of the method Clear Autoz becomes momentary and reverts back after a few seconds to the zeroed signal Autozero This button allows the operator to manually trigger an Autozero action during a chromatographic run to counter excessive drift in the signal especially after valve actuations All actions involving use of this button impact the data collected during a run with sharp excursions to zero at the point of activation Argon Nitrogen Helium 1 00 Vill Operating Conditions Effects of Detector Gas Flows on Peak Size for 0 02 Flame Tip Inert gases added into the flame tip with h
40. k Area 800 gt 3 c Q O A uVolts 0 Analyte Carbon Amount x Noise p_p x Signal Noise Target 4 0 4 2 4 4 4 6 Minutes Figure 10 Portion of 30 ppmV Propane Chromatogram with 1 ul injection Detector temperature 250 C Conditions used for this measurement are flame tip size 0 02 propane concentration 30 ppmV sample injection volume 1 0 microliter data bunch rate 16 6 3 Hz propane retention time 4 3 minutes detector temperature 250 C carrier gas helium at 5 ml min and make up gas nitrogen at 25 ml min An equivalent measurement with the 0 01 tip results in a detectivity value of 0 58 picogramsCarbon second Detection Limit The concept of detection limit is related to detectivity but it is computed differently Here the result is a measure of how low an analyte can be measured with statistical significance The whole measurement process is thus assessed including injection volume repeatability stability of the response of the detector and random signal variations The normal protocol involves measurement of at least seven identical consecutive runs at or near the expected detection level a calculation of three times the standard deviation of the areas from that series and then dividing by the analyte response factor peak area concentration to convert raw areas into concentration 21 22 200 uVolts wi i M Propa
41. low end when used for concentrations over multiple orders of magnitude common with hydrocarbon measurements This is a noticeable problem when the full range of the flame ionization detector typically 10 is displayed To illustrate the full linear range a more meaningful display is to plot Response Factor versus log Concentration as depicted in Figure 13 Thus maintenance of linearity can be visualized clearly at both low and high ends of the range and a defined tolerance frequently specified as 10 indicated across the range Propane Benzene n 10 aw 2 000 10 m x _ X X L50 1 10 1 50 Response Factor Area Concentration Response Factor Area Concentration Lodi 4 hw 1 10 100 He a e gt tr O Lo oS log Concentration ppbCarbon log Concentration ppbCarbon Figure 13 Plots of Response factor vs log Concentration for Propane and Benzene permit better visualization of conformity to linearity over a wider range including error bars for deviations at 10 Dynamic Range Dynamic range can be defined as the complete range of the detector where changes in concentration still effect a variation in signal but outside the linear range Typically non linearity occurs when the detector is severely overloaded and cannot readily distinguish increasing concentrations a calibration curve then reaches an upper plateau Some analytes can
42. lumn bleed from contaminating the detector innards The detector ready status is triggered when the actual temperature is within 4 0 C and the actual temperature settles at typically less than 0 2 C especially after a 2 minute stabilization interval ooo ra er l SamhHeDeliver Lal The actual temperature of the detector must exceed 100 C for the ignitor to be enabled to prevent accumulation of water Temperate Li um in the detector CAUTION the ferrule sealing in the flame tip has an upper temperature limit depending on the material of the ferrule Brown Vespel ferrules have an upper limit of 350 C dull black Graphite Vespel ones standard have a 400 C limit and soft shiny black ferrules allow the maximum limit of 450 C but are haraa often difficult to extract out of the detector base intact This Figure 16 FID temperature limit can be user fixed appropriately in Setup Detector limit is set in Setup Detector Flameout Threshold The Bruker 450 GC constantly monitors the flame signal to ensure that the flame remains lit through the chromatographic process If the raw signal without zeroing is detected below this threshold a flame condition is considered as out and operations are halted until corrected Very clean systems can generate very low baseline signals where a meem LL false flameout is sensed This parameter can then be comawain T samcedaivery lowered to allow these low flame
43. lved in adsorbing excess reaction energies typically the inert carrier and make up gas molecules Sternberg et al Gas Chromatography p 262 i Sternberg et al Gas Chromatography p 251 253 Figure 1 Zones involved with combustion processes in the flame ionization detector The sample is preheated in the flame tip A The unburned gas emerges to form the inner cone B Higher temperatures at its outer boundary lead to partial pyrolysis with free radical reactions from back diffusion of hydrogen atoms These cracked fragments then enter the main reaction zone C where oxidizing fragments O O2 OH radicals and HO radicals penetrate by inward diffusion of oxygen to accelerate cracking and stripping This leads to highly exothermic oxidation reactions Energy liberated in the oxidation reactions is retained primarily by the product molecules formed Then this excess energy is released by ionization reactions in the form of electronic energy D Surrounding the flame is a sheath of oxygen rich gas E with some diffusing into the flame cone to support combustion and most propelling combustion products and inert gases out of the detector I Principles of Operation The flame is generated by burning hydrogen supported with oxygen from air bathing the surroundings Exit hole at the top of the flame jet requires inert support gas usually the carrier gas to keep the flame stable for best performance If this column stream
44. make up gases to flow around the flame tip instead of properly through its orifice Then this leak can ignite and start burning up the ferrule to further degrade performance This ferrule should be inspected regularly to catch this fault and replaced if defective Bruker P N CR213200 Sealed Detector The Bruker flame ionization detector is sealed at all connections with an aluminum crunch washer at the base of the tower and silicone o rings on both probes except for the open vent at its top This permits accurate measure of gas flows at the top of the detector with the appropriate adapters included with every instrument flow tube P N 200187600 and large stopper P N 394958600 The crunch washer Bruker P N 1500334701 must be replaced whenever the tower is removed for inspection of servicing The o rings Bruker P N 9775204 should be exchanged when cuts and abrasions are observed Applying Anti seize on Tower Screws The detector base is constructed of stainless steel to keep its shape intact under stresses encountered Screws holding the tower to its base are also stainless steel for added strength during tightening Invariably stainless steel screws in stainless steel receptacles tend to seize by galling and can make subsequent removal of the tower extremely difficult To avoid this antiseize lubricant McMaster Carr P N 1367K91 tel 562 692 5911 can be applied only once to the very tip of the long screws used to clamp down the tower Extr
45. me lit Since air and hydrogen are burned to produce water volumetric flow is reduced when the flame is active Established total flow conditions can be monitored for overall flow with the flame lit to become the recognized benchmark Differing readings indicate inappropriate flow settings Also a loud pop on flame ignition is indicative of either excessive hydrogen or inadequate air flow Background Signal Condition of the complete detector can be gauged through monitoring of the detector background signal To view this level Autoz needs to be turned off and the Clear Zero button manually triggered in the Detector section on the Bruker 450 GC screen A low level indicates a clean detector and good purity of the gas sources Typically levels range between 1 and 5 mV and up to 10 mV are tolerable A level approaching 100 mV indicates severe contamination This signal should be continuously monitored over an extended time interval to note if the level is decreasing implying a reduction in contamination or is increasing showing that the detector needs a bit more cleansing with heat or replacement of suspected dirty parts or gases If extreme temperatures are needed to be used here especially exceeding the maximum allowable setting for the column the column ought to be removed and its connection point suitably plugged Contamination High background signals from contamination can directly impact the ability to measure low levels as noise is incr
46. me tip 9 4 uV sec ppbC ml 0 01 flame tip Gas Injection 40 ppbCarbon 300 ml load N make up 10 5 Patented Ceramic Orifice 0 02 and 0 01 isolated from chassis ground Automatic Sensor and Reignition Attempts with fault on failure Platinum Probe Resistive Temperature Device RTD 50 450 C 0 1 C 0 2 C After 2 minute stabilization True Mass Flow Controllers for all flame gases 10 Typically lt 0 2 ml min XIII Part Numbers for Bruker 450 GC Flame Ionization Detector 1 23 Tower Seal aluminum crunch washer pkg 25 P N 1500334701 25 42 DEFC Module FID Type 11 SI P N CP74101180 FID Field Kit with Manual Pneumatics Valve Combo Black Air 0 500 ml min Valve Combo Red Hydrogen Make up 0 50 ml min Valve Combo Blue 0 150 ml min FID Electrometer P N 392502101 Analog out Cable to non Bruker Workstation 9 pin D Shell Analog out Cable to non Bruker Workstation spade connectors Detector Tower Body P N 100099300 Ignitor Probe P N 200187200 Signal cable P N 391783000 Ignitor cable P N 392548701 Signal Probe P N 200187300 O rings silicone for probes P N 390596004 each or P N 9775204 pkg 10 Collector Tube P N 394958700 Collector Tube Insulator P N 2100003200 Screws 8 32 x 74 P N 1214200804 Screw FID Field Kit with Type 11 Detect
47. n due to the ease of preoxidation of the C Cl bond and absence of amp 120 4 detector response to the newly formed C O bond 110 4 3 gt If methane and carbon dioxide carbon monoxide 5 ne converted to methane with a reduction catalyst are 90 4 the only components to be measured then this 0 20 40 addition of internal oxygen 40 ml min can improve Internal Air Flow ml min performance of the detector by about 20 Other Figure 25 Relative enhancements to analytes will yield a lower response with some Methane response with addition of air generating no response into the make up flow balance N for total 50 ml min to the flame ionization detector Source http www wolframalpha com input i heat capacity ofthelium Computed from Graham s Law See http en wikipedia org wiki Graham 2 s_law accessed May 22 2011 Source http www wolframalpha com input i thermal conductivity helium accessed May 22 2011 Sternberg et al Gas Chromatography p 257 Sternberg et al Gas Chromatography p 253 Sternberg et al Gas Chromatography p 260 30 31 32 33 34 35 Effects of Flame Jet Size on Sensitivity A smaller flame orifice 0 01 ID creates a tighter flame profile that produces a hotter flame This generates somewhat different response to hydrocarbons than its 0 02 brother as illustrated in Figure 26 Optimum flows for this tip are also different than for the wider one Due its flame geometry t
48. n through the column bore The error message Not Ready in Run is displayed until the fault is corrected typically through the automatic reignition process The trouble is usually correlated with loss of supply gases or some valve event especially when valving is configured after the column To minimize this interruption a flame out restrictor fashioned from a short piece of narrow bore tubing 0 18 mm ID or for packed columns 0 05 narrow bore stainless steel tubing can be installed just prior to the detector Although more common when packed columns were in wide use massive injections of solvent especially water can extinguish the flame In this case a smaller injection loading is needed to reduce the solvent pulse through the detector either by reducing the syringe volume or increasing the injector split ratio This fault is spotted through an error displayed in the Bruker 450 GC Log and also by three high voltage spikes 10 Volts in the chromatogram indicating attempts to relight it If the flame fails to light a fatal error is triggered the active run is immediately halted along with subsequent automated operations and the current run data is lost Validation of Flows Detector gas flows have a major impact on the performance of the detector Whether flows for the supply gases are set with a detector electronic flow controller DEFC or with manual needle valves flow rates ought to be confirmed with a digital flow met
49. ne rT hy nha all rho ath M hallak ih M l hi N i ay dk j TT i Wl ns ii Nh ii i N Ni Mat h 32 0 33 0 34 0 Minutes 35 0 Figure 11 Typical Propane peak generated from a 0 27 ppbCarbon standard and 300 ml loaded A cryogenic concentrator reduced the sample volume to lt 90 ul for on column injection Detector temperature 250 C McNair H M Bonnelli E J Basic Gas Chromatography Varian Associates Palo Alto Ca 1969 p 87 US EPA Compendium of Methods for the Determination of Toxic Compounds in Ambient Air Second Edition Compendium Method TO 15 January 1999 Section 11 2 2 p 15 30 www epa gov ttn amtic files ambient airtox to 1 5r pdf Raw Area Counts a uVolt sec Detection Limit 872 3 x Std Dev Resp Factor 823 225 3 019 926 oe 0 07 ppbCarbon in 300 ml 386 or 10 3 pgCarbon 898 1074 Table Ill Computation of Propane detection limit with a 0 27 ppbCarbon standard and 300 mi sample loading with nitrogen make up Sensitivity The term sensitivity or detector response is often incorrectly equated with detectivity Sensitivity is a measure of how large a signal is generated by the flame ionization detector for a given concentration of analyte or the slope of its calibration curve normalized by the injection volume r ik Peak Reponse ensitivity SAA y Analyte Concentration x Injected Volume Units for sensitivity become microVolt sec
50. noise suppression is performed through user selection of the data rate by choice of the Detector Bunch Rate in the Workstation Here the Bruker 450 GC executes a simple averaging of the unprocessed data after treatment by FIR Figure 30 Signal averaging is set filtering at 100 Hz to produce the specified data rate Noise through Detector Bunch Rate is reduced somewhat by this action but does not correlate from the Workstation with expected decreases For example a change from 100 Hz to 0 8 Hz should yield a noise reduction of 100 0 8 or a factor of 11 but only a drop of 3 2 is noted Certainly FIR filtering suppresses noise very effectively with the higher data rates without impacting peak shapes Detector Bunch Rate points 6 3 Hz Noise Monitor Length E4 H bunched points eee 10 2 sec P i 7 UPAEEEEEEETETEIN Act ET erer I tae eo oe Spacek oeE Ti STREET begg G2 70 g eaceprawzveigreds TE EES FFIR te i Fii rere a 4 Discussions of Finite Impulse Response filtering can be found in R Bramston Cook Peak Detection with Varian Star Workstation for Varian 3800 and 450 GC Gas Chromatographs 2010 available on request from Lotus Consulting at ebramstoncook msn com ij ih Ii ull li ii AA MN my nn La wee
51. o the detector XII Summary of Performance Data for Bruker 450 GC Gas Chromatogram RMS Noise Peak to Peak Noise Drift Signal Range Amplitude Resolution Detectivity Detection Limit Sensitivity Linear Range Flame Tip Flame out Temperature Sensor Temperature Range Temperature Resolution for Setpoint and Display Temperature Stability Detector EFC Controllers Detector EFC Accuracy Detector EFC Stability Published Typical Specification for Bruker Performance for A are 450 GC Bruker 450 GC lt 7 5 microvolts lt 4 6 microVolts ae 10 Hz data rate 12 5 Hz data rate detector 250 C lt 33 microVolts 12 5 Hz data rate Peak to Peak 5 minutes detector 250 C lt 10 microVolts minute 0 3 microVolts minute Flame on for gt 1 hour 100 microVolts to 1 000 Volts full scale Single Range of 10 Amperes mv 1 microVolt 1 femtoAmpere Range 12 0 87 pgC sec 0 02 flame tip lt 2 pgC sec 0 56 pgC sec 0 01 flame tip 30 ppmV Propane 1 ul injection N2 make up p p S N 2 6 3 Hz detector 250 C 10 3 pgCarbon 0 07 ppbC w 300ml 0 27 ppbC Propane 300 ml load w Cryotrapping detector 250 C gt 3 300 uV sec ng Helium make up gt 6 600 uV sec ng Nitrogen make up Liquid injection 1 ul Cis 30 ng ul splitless 0 02 flame tip 6 27 uV sec ppbC ml 0 02 fla
52. olved must be kept exceptionally clean and all gas supplies need to be the best available and especially free of residual hydrocarbons Installation of hydrocarbon filters Bruker P N CP17972 on all gas supplies can help insure maintenance of the required purity Titanium Flame Tip Nut Older Varian detectors used chrome plated brass nuts to hold the flame tip in place They were very prone to coming loose even when held at constant temperature Several years ago these were replaced with ones made from titanium P N 1600078500 With more compatible thermal expansion with other materials in the detector these nuts are less likely to loosen Titanium nuts are distinguishable from brass by a more machined appearance to the surfaces including sharper corners and sometimes subtle visible machining grooves on the edge planes Chrome plated brass ones have more rounded corners and a very smooth looking surface When the flame tip is reinstalled this nut must be snug enough to prevent gas flows from escaping through the associated ferrule but not overly tightened to cause the interior ferrule to crack or deform Insulators Electrical insulators are deployed in four locations in the eae Bruker 450 GC flame ionization detector to insulate the electric eae E __ Nia mal field between the cathode flame tip and anode collector tube holding Collector Tube 5 from current leakage White ceramics are used to isolate the tip of the flame jet
53. or Electronic Flow Controller DEFC P N 452100101F 120 VAC P N 452100102F 230 VAC P N 452100103F 101 VAC P N 452100901F 120 VAC P N 452100902F 230 VAC P N 452100903F 101 VAC for manual control P N 391706500 for manual controls P N 391706501 for manual controls P N 391706503 P N 392567201 P N 392567501 Microseal 8 32 x 3 P N 391866302 Uncoated 8 32 x 3 P N 391866301 Tower Top Plate P N 3400073100 Flame Tip Assembly with washer and ferrule 0 02 ID P N 200193800 0 01 ID P N 200187500 Flame Tip Ferrule 10 pkg Vespel 100 max 350 C P N CR212200 Graphite Vespel 40 60 max 400 C P N CR213200 Graphite 100 max 450 C P N CR211200 Flame Tip Nut titanium P N 1600078500 Detector Base P N 392547501 Distinguishing the two flame tip sizes is not easy but comparing the two by looking down the center axis of the tip can provide an assessment of the differences with the 0 01 tip being obviously smaller 6 Column Ferrules pkg 10 2 30 31 43 40 Graphite Column ID Teflon Vespel 60 Vespel Graphite Max Teme 450 C No Hole CR2121XX CR2131XX CR2112XX Single Hole 0 18 mm CR212103 CR213103 0 25 mm CR214104 CR212104 CR213104 CR211104 0 32 mm CR214105 CR212105 CR213105 CR211105 0 53 mm CR214108 CR212108 CR213108 CR211108 Two Hole 0 25 mm CR213124 0 32 mm CP212222 CR213125 CR211125 Flow Tube P N 2001
54. or for a limited hydrocarbon set can be applied to all others measured with the same detector Figure 4 illustrates the ability of high performance detectors to measure a wide range of hydrocarbons and still achieve even responses with the same detector Then for example the response for Propane can be applied to all analytes measured near its elution and Benzene for those around it and beyond with confidence that correct concentrations for the full range of hydrocarbons are reported For older Varian gas chromatographs including Varian 3700 3400 3600 this distance is 11 5 cm 1 For example over 300 separate hydrocarbon species can be detected in engine exhaust and ambient air 3000 4 2 1960 4 1920 2940 2880 1880 4 Response Factor Area Concentration 2820 1840 4 Response Fator Area Concentration Figure 4 High performance flame ionization detectors yield uniform response factors over a range of analytes as demonstrated with a 100 ppbCarbon NIST custom hydrocarbon blend certified to 2 The discontinuity between two response factors is due to a switch in detectors with differing flame tips sizes IV Special Applications Flame lonization Detector in Series with Thermal Conductivity or Photoionization Detectors Since the thermal conductivity or photoionization detectors do not destroy the character of analytes either can be plumbed in series with the flame ionization detector to yield respons
55. ound signal flame out error message not displayed Soft audible pop sound on ignition Water vapor on shiny cool surface above detector Peaks detected on chromatogram Signal background greater than 300 uV activating Clear Autoz displays background Visible inspection for a flame is not possible with transparent hydrogen oxygen flame Do NOT use alcohol soaked Q tip inside detector to verify Failure to Ignite Flame Reasons for the flame not to light include e Wrong flow rates Ignitor broken carefully peek at glow inside the detector tower during ignition to confirm Loose flame tip Column not attached Column installed too far up into flame tip Facilitating Ignition Some settings for flame gas flows can make ignition difficult to accomplish A temporary reduction of air flow from the keyboard of the chromatograph during the ignition process can enrich the flame gases enough to light the flame This problem can be a particular troublesome issue with the narrow flame tip 0 01 and relatively high flows needed to attain maximum performance Once the flame is lit it usually remains until the fuel and oxidant flows are removed or the flame lifts off from the tip from high flows or flow surges especially from valve actuations Flame Out During a Run Occasionally the flame may be extinguished in the middle of a run This problem is rare as flow perturbations prior to the head of the column often are dampened out by the restrictio
56. peaks from low microVolts to kilovolts with this single range setting Autoz Zeroing of the detector can be preprogrammed at specific times during the developing chromatogram to bring the signal back to zero Appropriate times for this action are at the start of every run and at severe disruptions in the baseline that can occur with valve actions and column switching among other events Active Bruker 450 GC Display Screen Buttons Three buttons on the detector display of the Bruker 450 GC allow the user to perform specific actions related to the signal These are always active when the detector is enabled even during a chromatographic run Ignite Flame ignition is attempted whenever a method is activated when flame out is sensed or manually by the The process If ignition is If the flame is not validated an error message is triggered The error can be cleared by either manually reattempting ignition or by operator using this button on the screen makes three attempts to light the flame confirmed the ignition steps cease un enabling the detector 28 E Method Builder Randy mth File Edit View Window Help asas Alale exa S 2 Ele S Sm iS Randy mth Front Detector Middle Detector Rear Detector Front Detector Type FD Detector Oven On Temperature C Electronics On Front Detector Type FID Detector Oven f On Temperature C Electronics On Autozero Range Ad
57. pikes can be generated by mechanical vibrations electric motors proximate transmissions from cell phones an unbalanced column oven fan unstable flame from improper flows flame tip partially plugged and particles from columns especially packed columns and porous layer open tubular PLOT capillary columns A special particle trap Bruker P N CP4018 can be inserted after a problem column and gt before the detector to reduce the effects of these column crumbs N Ss mVolts Although the flame ionization detector is not very flow sensitive 1 6 16 17 Minutes occasionally pressure pulsations from single stage regulators on supply Figure 37 Noise gases can generate regular patterns on the chromatogram appearing produced from an active as spikes Dual stage regulators provide extra dampening of the outlet cell phone near flow and should be used on all gas supply tanks A dual stage regulator detector cabling is distinguished from single stage one by a protrusion on its backside to hold the second stage diaphragm single stage have a flat backside Both have two gauges one for the tank pressure and the other for the outlet and they are not distinguishable in this manner Also dual stage regulators are less prone to increase outlet pressures as the tank depletes and thus give no change in flows at the detector as the tank pressure depletes Drift A drifting baseline can dramatically impact the quality of peak processing espe
58. r setting will not greatly alter these optimum points Signal enhancements with nitrogen and Molar Heat Relati Thermal argon over helium or hydrogen as make up Capacity Sa S Conductivity gases are attributable to differences in J mol K Mtusion W m K 2 physical properties of the gases Lower Hydrogen 0 1805 diffusion and lower conductivities for argon Helium 208 O5 0 1513 and nitrogen lead to a tighter hotter flame Nitrogen 291 O27 0 02583 with a higher average molecular energy in Argon 208 0 22 0 01772 the reaction zone The lowest thermal conductivity and diffusion of argon generates Table IV Physical Properties of Make up Gases a higher temperature at the flame tip and slower equilibration of the liberated energy from the flame reactions over the other gases These actions result in its higher sensitivity Heat capacities for all listed carrier gases are nearly the same and have marginal impact on performance compared to the other properties For the same reason that argon and nitrogen provide enhanced signals for the flame ionization detector with their lower diffusion they are not commonly employed as carrier gases for capillary columns The analyte exchange between the stationary and mobile phases of the column system is critically dependent on the relative diffusion of the carrier and peak shape is dramatically altered with these gases due to their resistance to mass transfer compared with
59. rameters are accessed Log through this icon on the Figure 15 Bruker 450 GC Screen Display keyboard of the Bruker for Flame lonization Detector 450 GC Several parameters displayed are live updates of current operating conditions The colored dots to the right side of the screen indicate the status of these components e A green dot designates the component is Ready A I dot indicates the component is Not Ready particularly the detector temperature or flow rates e An emoty dot stipulates the component is switched off by an unchecked Enable e A red dot signifies the component is in ERROR including flame out or failure to reach the set temperature or detector flow setpoints Precise errors can be viewed though clicking on Log at the bottom right of the screen Actual 50 0 EE Signal Actual Temperature Temperature at the detector base is displayed and updated in real time If this temperature is outside 4 0 C of the set point the not ready MAIS dot is illuminated until the tolerance is reached If the temperature does not make any progress toward the setpoint the system faults out red with all temperature zones in the instrument shutting down Only rebooting the system after the deficiency is corrected clears the fault Signal The live flame ionization detector signal is always displayed here and is impacted by the activation of any auto zeroing action Flame ignition can be conf
60. s due to the detector can be caused by a loose flame tip a wider flame tip than expected incorrect detector gas flows or wrong detector electrometer range A remote possibility is an electronic failure of the electrometer An easy test to validate a functioning electrometer is to simply wiggle the signal cable and watch for a response on the signal display as this mechanical perturbation always generates a response on the electrometer 3 Trouble with an unbalanced fan motor can be confirmed by turning off the column oven heater and fan from the Bruker 450 GC keyboard and observing any change in audible and detector noise Stuck in Equilibrating Whenever an instrument parameter is not at its setpoint the system indicates that it is equilibrating and will wait until all setpoints are reached With the flame ionization detector the areas to look for problems reaching the compulsory settings can involve its base temperature and detector flows when using the Detector Electronic Flow Controllers The precise cause for the system not coming Ready can be viewed from the Status Page on the Bruker 450 GC keyboard Remember that the detector gt has an automatic one minute hold after ignition and will show Equilibrating during this wait XI Maintenance Since raw signals from the flame ionization detector are measured into the femtoAmpere range 10 Amperes all components inv
61. specially in pharmaceutical products Distinct advantages of this detector include e Huge number of detectable analytes e Insensitive to many bulk matrices such as air water carbon dioxide carbon monoxide all inorganic gases H2 O2 N2 Ar He and carbon disulfide Extremely wide dynamic range often exceeding capabilities of chromatography Low detection typically less than 50 ppbV Methane The common flame employed with this detector in gas chromatography is a hydrogen diffusion flame with no premixing of fuel and oxidant Combustion takes place at the interface between hydrogen coming through the flame jet and oxygen available from a sheath of air around the flame Typical fuel mixture travels at linear velocity approaching 1 300 cm sec whereas the oxidant flows at about 10 cm sec This outcome yields a very clean stable and transparent flame with temperatures over 2 000 C Even though this combustion is very inefficient in generating electrons estimated conversion around 0 0015 electrometers used to process the currents generated are sensitive enough to detect these very low signals The chemical processes occurring within the flame envelop are combinations of completing reactions Sternberg et af postulated conditions for production of ions as e Dependent on the energy density in the ionization zone which is determined by the composition flow and diffusion properties of the gases introduced into the flame
62. tic tubing is utilized even for a short duration contamination from them can penetrate into the system pneumatics and necessitate wholesale replacement of all tubing and pneumatic components including injectors valves and even columns to return the detector to its perfect state Noise Excessive detector noise severely impacts the ability to measure low level analytes Typical expected noise levels for good performing detectors are illustrated in Section V above Frequent causes for high noise are due to excessive background levels from contamination Conditioning the system at elevated temperatures often reduces this effect Consistent spikes from external sources can appear in an irregular pattern to look like noise All instrument covers should be installed during normal operations to assist in shielding the detector and signal processing from possible radiofrequency interferences including cell phones electric motors and solenoids 800 mVolts mVolts 4 0 5 0 6 0 Minutes Figure 35 External electrical ea 0 6 0 8 1 0 Minutes interferences especially radio Figure 36 Electrometer noise transmitters can cause severe generated from an active Dremel disruptions in the FID signal and must rotary tool close to be isolated away from the instrument the signal cable Spikes Dramatic excursions in signal can greatly impact quantitation by artificially splitting peaks up and adding in extraneous areas These s
63. tivity is a measure of how low the detector can properly perceive hydrocarbons Often this is simply a computation of the signal to noise ratio of a low concentration hydrocarbon analyte and then extrapolated to a value of 2 1 However the sharpness of the peak can greatly impact the amplitude of the signal and the detectivity value Since the measurement involves a flux through the flame tip the better measure of detectivity is the minimum detectable quantity per second This computation then normalizes out the influence of peak widths Basing the calculation on peak to peak noise peak response as peak area and the limit as twice the noise level detectivity can be computed by 7 Detectivity Conventional units for detectivity are picogramsCarbon second pg sec The analyte amount is converted to weight of carbon by means of multiplying the injected amount by the percentage of weight of carbon in the molecule A lower value for detectivity implies a better performing detector Since noise is impacted by the data treatment applied to raw data this specification must be listed with all measurement conditions employed and the criteria used in the calculation The published specification for the Bruker 450 GC flame ionization detector is a conservative 2 pgCarbon sec An actual measurement yields a value of Detectivity 44 picogramsCarbon of Propane x 16 microVolts x 2 ee 1 626 microVolt sec 0 87 picrogramsCarbon second Pea
64. ts at or near detection limits An unstable baseline can make peak detection at these levels difficult and erroneous Drift can be generated from contamination anywhere in the chromatographic process and needs to be minimized for best performance Drift is differentiated from noise by a measure of the long term change in signal levels rather than the higher frequency random noise Figure 8 illustrates a typical drift over 30 minutes for the Bruker 450 GC flame ionization detector A properly conditioned system should exhibit lt 10 microVolts minute drift 40 2 NN A LAMM Satta i hd lid i Figure 8 Typical drift for a flame ionization detector in the Bruker 450 GC Gas Chromatograph Raw data points ASCII are transferred into Excel to compute the trend line showing a drift of 0 3 microVolts minute from the slope of the line Signal Amplitude Resolution The signal conversion process in the Bruker 450 GC can distinguish signal amplitudes at 1 microVolt resolution equivalent to 1 femtoAmpere as shown in Figure 9 This rate is fully adequate to define peaks and noise for the flame ionization detector under all measurement conditions id T 5 h 4 Ka pn F MON 4 8 700 8 750 8 800 Minutes Figure 9 A magnified portion of the signal shown in Figure 7 illustrates the signal resolution of 1 microvolt for the Bruker 450 GC Flame lonization Detector Detectivity Detec
65. unch washer Bruker P N 1500334701 at the base of the detector tower and silicone o rings Bruker P N 390596004 each or P N 99775204 pkg 10 on the probes provide seals around the detector so that all flows are directed to the top of the tower to be monitored with a flowmeter This crunch washer should be replaced every time the tower is removed for inspection or maintenance A heater is installed in the detector base to prevent any condensation of material inside that would contaminate the detector and impact performance and is separated from and usually higher than the column temperature And an associated platinum resistive temperature device RTD temperature probe is inserted nearby to monitor temperature at the base The massive tower is in close contact with the detector base to keep its temperature near the setting and stable Since the detector is capable of being maintained up to 450 C side arms off the tower are fashioned to keep electrical connections at the probe ends away from damaging heat An ignitor is included with the bottom probe and is close to the tip to provide convenient lighting of the flame The ignitor is automatically triggered whenever a method is activated when a flame out is sensed or when the user selects the keyboard ignite button The ignitor makes three attempts at lighting and upon failure to confirm ignition generates a fault and stops all instrument workstation operations with an indicated error mess
66. ydrogen aids in enhancing the combustion process and increasing peak sizes A study of varying flows over the range of 20 to 50 ml min for helium nitrogen and argon shows that argon provides the best sensitivity and a flow of 50 ml min boosts the 15 20 25 30 35 40 45 50 55 response to the maximum level Make up Flow ml min 0 67 0 33 Relative Area Counts Figure 22 Changes in response with 1 00 different make up gases and V various flow rates with 0 02 tip 0 67 a Maximum peak size is achieved with hydrogen flows between 30 and 35 ml min with 50 ml min 0 33 argon make up Similar profiles for hydrogen are realized with the other make up gases Realative Area Counts 15 20 25 30 35 40 45 Hydrogen Flow ml min 1 00 D Figure 23 Relative peak sizes versus hydrogen flow in the detector with 0 02 tip Flow rates for air flowing around the flame have minimal effects on performance of the flame ionization detector over the wide range of 200 to 400 ml min A flow rate of 300 ml min is suggested to ensure that combustion products are adequately swept out to greatly minimize possible contamination Figure 24 Changes in peak size versus inside the detector air flow in the detector with 0 02 tip Relative Area Counts I 150 200 250 300 350 400 Air Flow ml min Carrier gas flow for all experiments is 3 ml min Helium Minor changes from this carrie

Primer on Flame Ionization Detectors

Contents

Download Pdf Manuals

Related Search

Related Contents