Thermal Desorption Technical Support€¦ · tube sorbents for calibration to those used for field...

Markes International Ltd. is pleased toacknowledge the collaboration of Mr MikeWright and colleagues at UK HSL in preparationof this technical note.

IntroductionIt is not uncommon for calibration standards,used during thermal desorption (TD) analysis ofvapours trapped on sorbent tubes, to bedifferent from field samples. The ideal TDstandard is prepared by introducing a fixedvolume of an accurate gas standard of thecompounds of interest onto an identical sorbenttube to those used for field sampling ofvapours. However, accurate gas standards arenotoriously difficult and expensive to generate;particularly in the case of labile, polar or lessvolatile analytes and at low concentrations. Forthis reason, and as specified in most TDstandard methods, routine thermal desorptioncalibration is normally carried out byintroducing liquid standard solutions to thesampling end of sorbent tubes in a stream ofcarrier gas. The bulk of solvent is thengenerally purged from the sorbent tube duringthe standard loading process (See TDTS 7 -Calibration: preparing and introducingstandards using sorbent tubes). However, it is not always possible to completelypurge the solvent from the tube – for examplewhen dealing with volatile compounds or withstrong sorbents. In this case, standard tubescontaining relatively large masses of solvent(possibly several milligrams) will be used tocalibrate sample tubes which are solvent-free.Conversely, whether loaded from the liquid orgas phase, standard tubes are generally dry.However, tubes used for field sampling maycontain several milligrams of water, depending

on the sorbent used and the prevailingatmospheric conditions. Similarly, it is often convenient to use differenttube sorbents for calibration to those used forfield monitoring. For example, CertifiedReference Standards on TD tubes and tubesprovided for quality assurance/proficiencyschemes are most commonly packed with 200mg Tenax TA™, whereas tubes used for fieldsampling may be packed with 500 mg carbonblack or two or three different sorbentsdepending on the application (see TDTS# 5Advice on sorbent selection, tube conditioning,tube storage and air sampling, and TDTS# 20Confirming sorbent tube retention volumes andchecking for analyte breakthrough). DifferentTD sorbents have very different densities andare available in a range of particle sizes from20-40 to 60-80 mesh. Some sorbents are alsomore prone than others to the formation of‘fines’. (Fines are small, dust-like particles ofsorbent that fill the voids in the sorbent bedand cause blockages. They are most commonlycaused by rough handling e.g. dropping thetubes or repeated insertion of syringe needlesto introduce standard.) All of these factors canaffect the back pressure of a tube and itsimpedance to gas flow.This technical note describes the issues whichmay arise as a result of differences betweensamples and calibrants for thermal desorptionand the operating conditions under which theeffects may be significant. Guidance forminimising the impact of such differences andfor validating that a given method is notsubject to these effects is also presentedtogether with advice on routine monitoring oftube impedance.

Thermal Desorption Technical SupportNote 75: Liquid standard injection, tube impedance and

other factors which may cause discrimination during thecalibration of thermal desorption methods

KeywordsTD Operation; Calibration; Impedance; Discrimination

www.markes.com

T D T S

TDTS 75_2 September 2009 of 13

www.markes.com

Potential analytical impact of varying tube impedancesTube impedance – normal rangeTubes manufactured by Markes Internationalare all prepared using weighed masses ofsorbents and each individual tube is flow testedunder a fixed supply pressure (~0.5 psig) toensure it exceeds a minimum threshold flow of200 ml/min. Tubes which fail this test arerejected and repacked. However, not all tube manufacturers use thislevel of care and the variation in impedance ofdifferent brands of sorbent tube can be verysignificant. The back pressure of a range ofcommercial pre-packed and self-packedstandard sorbent tubes [3.5-inch long x 6.4mm O.D. x ~5 mm I.D. (stainless steel) or ~4mm I.D. (glass)] was evaluated by UK Healthand Safety Laboratory with a flow of 50 ml/minN2 as described in their report OMS/2002/15 –See Table 1.

Preliminary observations from this data arethat standard glass tubes containing carbonblack or carbon molecular sieve sorbents,particularly tubes packed with multiple beds ofthese sorbents, are prone to high impedanceeven at moderate (50 mL/min) flow rates. Suchsample tubes are therefore most at risk ofdiscrimination due to impedance variation iftubes with more normal back-pressure are usedfor calibration or quality assurance.

The potential impact of highimpedance tubes/traps duringsamplingIn reality, the biggest practical issues causedby high impedance sorbent tubes (or focusingtraps) are likely to be observed duringsampling. Both on- and off-line monitoringmethods involving active sampling (pumped orpressure controlled) are subject to error iftubes or traps become ‘blocked’ i.e. if the back-pressure increases to such an extent that thesampling pumps or delivery mechanism cannotreach the required sampling flow and eithercut-out completely or deliver lower samplingflows/volumes than those expected. Pumped tube sampling All personal monitoring pumps have a tubeimpedance/flow limit and will struggle todeliver the required sampling rate if tubeimpedance increases above a given level. Thislevel varies from pump to pump depending onmake and model. However, standard methodsfor air/vapour monitoring with pumped sorbenttubes invariably require calibration of eachsample train (i.e. each specific pump and tubecombination) before field monitoring. Thismeans that pump/tube combinations whichstruggle to reach the desired flow can beeliminated before field monitoring begins.Furthermore, routine checks of tube impedanceshould be carried out by users, say every 20 orso uses. In either case it means tubes withhigh back pressure can be excluded from fieldmonitoring exercises so that no samples arecompromised. Online monitoringIn the case of TD-based on-line air monitoringsystems, where air or gas is drawn directly intothe focusing trap of the thermal desorber underelectronic mass flow control; ‘blocked’ or high-

Sorbent type psi kPaTenax TA 35-60(Chrompack) 0.145±0.04 1.00±0.3Chromosorb 106(Chrompack) 0.160±0.03 1.10±0.2Carbograph 1 TD(Alltech) 0.087±0.01 0.60±0.1Carbograph 5 TD(Markes) 0.029±0.01 0.20±0.1

Carbopack B (glass)(Supelco) 0.493±0.04 3.40±0.3Carbopack X (Supelco) 0.131±0.03 0.90±0.2Carboxen 1000 (glass)(Supelco) 0.377±0.04 2.60±0.3

Carbopack B /Carboxen 1003 (glass)(Supelco) 0.232±0.03 1.60±0.2

Carbopack B /Tenax GR (glass)(Supelco) 0.305±0.03 2.10±0.2Carbotrap 300 (glass)(Supelco) 1.306±0.58 9.00±4.0

Molecular Sieve 5Å(Chrompack) 0.145±0.06 1.00±0.4

Table 1: Back-pressure of a range of industrystandard stainless steel and glass tubes - N2

at 50 mL/min

T D T S


www.markes.com

impedance focusing-traps can impact the flowrate and volume of gas sampled. However, asthe same trap is used for collecting eachsample in the sequence, back-pressure typicallyincreases only gradually – for example; due tothe collection of fine particles from the samplestream or fragmentation of the trap sorbentover extended periods (months). This is,therefore, unlikely to cause run-to-runvariation. Markes on-line TD monitoringequipment incorporates filters in the samplestream (to prevent the ingress of fine particles)and uses flow sensors to log actualflows/volumes during sampling. This ensuresthat quantitative monitoring data can continueto be obtained and alerts the user wheneverthe trap needs changing.

The potential impact of variabletube impedance during TD analysisOlder thermal desorption technology, withforward flow desorption and no valve to isolatethe primary sorbent tube from the GC columnduring analysis (see Figure 1) is/was inherentlysubject to significant error due to tubeimpedance variation.

In this type of flow path configuration, anysignificant change in tube impedance causesthe carrier gas pressure downstream of thesorbent tube to vary; thus affecting both splitflow (if applicable) and chromatographicretention time. This effect is observed whetherthe supply of carrier gas to the thermaldesorber is manually or electronically controlledand impacts both quantitative and qualitativeanalysis. Modern configurations of two-stage TD aremuch less susceptible to issues related to tubeimpedance because the primary sorbent tube is

isolated from the focusing trap and GC(/MS)during secondary (trap) desorption/initiation ofthe GC run (figure 2b).

Note that the impedance of the primarysample tube cannot impact the column headpressure or split flow during secondary (trap)desorption & initiation of the GC(/MS) analysis,because the tube is not in the carrier gas flowpath at this time (Figure 2a).However, if the TD method requires an ‘inlet’split i.e. a split during primary (tube)desorption (figure 2a), variation in tubeimpedance can have an effect on the split ratio,particularly if the system is being operated atlow carrier gas pressures. Any variation inselected desorption temperatures can alsoexacerbate this effect. An example of thediscrimination that can be caused by variationin tube impedance, under extreme inlet splitconditions (high (140 mL/min) inlet split flowand low (8 psi) column head pressure) isshown by UK HSL in their report OMS/2002/15– see Table 2. Under these high flow/low pressure conditions,sorbent tubes with high back-pressure simplycause so much impedance that the pressure ofgas immediately downstream of the sampletube is not sufficient to drive the gas throughthe focusing trap or split vent at the ratesprogrammed. This remains true whether thesplit and desorb flows are manually orelectronically controlled. In this situation, where the pressuredownstream of the sample tube is too low tosupply the flows required, gas tends to takethe path of least resistance, i.e. it passespreferentially through the split vent withproportionally less of the sample streampassing through the comparatively highimpedance focusing trap. The discriminationobserved is thus a result of the split ratio beingeffectively increased for high impedance tubes(relative to those of lower impedance) thuscausing lower apparent recovery from the highimpedance tubes. At higher operating pressuresand more moderate gas flow rates, tubeimpedance is unlikely to cause discrimination,even when operating with inlet split – see Table3 overleaf. (Reproduced from ReportOMS/2002/15 with the kind permission of UKHSL).

T D T S

Figure 1: Flow schematic showing forwardflow TD system with no valve


www.markes.com

Figure 2a: Markes UNITY TD flow path duringprimary (tube) desorption

Figure 2b: Markes UNITY TD flow path duringsecondary (trap) desorption

Carriergas flow Desorbflow Outletsplit flow

Heated valve

Sorbe

nt tub

e

Cold

trap

Re-co

llectio

n tub

eCarriergas flow Desorbflow Inletsplit flow

Heated valve

Sorbe

nt tub

e

Cold

trap

Re-co

llectio

n tub

e

T D T S

Minimising the impact ofimpedance variation duringdesorptionSorbent tubesAs described above, the impact of varying tubeimpedance in an automatic desorptionsequence, is unlikely to be significant exceptwhen using a very high inlet split flow withrelatively low carrier gas pressure – a rare andusually ill-advised combination. Nevertheless,Markes TD systems facilitate monitoring of tubeand trap impedance, during the thermaldesorption process, to detect and minimise anyrisk of discrimination.At one extreme there is the issue of completelyblocked tubes. Blocked tubes are detected

during the ambient temperature leak test atthe beginning of the run and are not analysed.After completion of a sequence, the error logreports any such failures and allows the user toinvestigate samples which have failed the leaktest for this reason. More moderate variation in tube impedance isdetected on Markes thermal desorbers bycontinuous monitoring of carrier gas pressuredownstream of the sorbent tube/trap at allstages of system operation, e.g. standby, pre-purge, primary (tube) desorption andsecondary (trap) desorption.


www.markes.com

Markes International Ltd. Tel: +44 (0) 1443 230935 Fax: +44 (0) 1443 231531 email: [email protected]

TubeC 106mass(mg)

Back-pressure (psi) Flow duringconditioning

(mL/min)

TD split flows(mL/min) Toluene

peak areaµV.s-1(TIC)

Beforeconditioning

2 hconditioning

17 hconditioning Inlet Desorb Outlet

A 100 0.06 0.07 0.06 98 143 10.2 57.8 72657B 100 0.03 0.04 0.03 110 144 10.6 56.9 69882C 104 0.06 0.06 0.04 95 144 10.6 ~57 65221D 101 0.04 0.06 0.03 92 145 11.5 ~57 68305E 103 0.09 0.09 0.07 94 145 10.5 ~57 64394F 100 0.03 0.04 0.03 105 146 10.7 ~57 56383

66140±5660(mean±s.d.)

G 531 0.25 0.26 0.25 30 136 9.9 57.0 61566H 528 0.22 0.25 0.23 31 137 9.7 ~57 50546I 529 0.22 0.25 0.23 31 137 10.4 ~57 54540J 526 0.22 0.25 0.23 31 137 9.7 ~57 54033K 527 0.23 0.25 0.23 30 137 10.1 ~57 61140L 527 0.23 0.25 0.23 30 138 10.2 ~57 57576

56567±4327(mean±s.d.)

Table 2: Recovery of ~17 µg toluene from TD-GC system, operating with high inlet split and at 8 psipressure. 14% lower recovery is observed with high impedance tubes (Toluene loaded from gas

phase)

T D T S

It is normal for the carrier gas pressure downstream of the sorbent tube or focusing trap tovary slightly during the various stages of TDoperation and as the temperature of the tubeor trap increases during the desorption process.However, exceptional changes of carrierpressure downstream of the tube duringprimary desorption, is a clear indication that aparticular tube has a different impedance to theothers in a sequence. Markes thermaldesorbers incorporate a user-settable ‘minimumpressure’ parameter which will prevent analysisof any tube exhibiting unusually high backpressure.Focusing trapWhile focusing trap impedance is unlikely tochange dramatically from run-to-run undernormal operating conditions, a step change in

focusing-trap impedance could be expectedwhen a trap is replaced or if a differentsecondary desorption temperature is selected.In either case, back-pressure regulatedelectronic carrier gas control (ECC) of thecarrier gas stabilises the carrier gas pressure atthe head of the analytical column/outlet splitpoint, ensuring that column and outlet splitflows remain constant independent of trapparameters – impedance, split flow, trapdesorption temperature, etc. (see Figures 3aand 3b).


www.markes.com


Table 3: Recovery of ~17 µg toluene from TD-GC system, operating with inlet split and at 46.1 psipressure. No significant difference in recovery is observed between low and high impedance

tubes. (Toluene loaded from gas phase)

TubeC 106mass(mg)


(mL/min)

TD split flows(mL/min)

Toluenepeak areaµV.s x 10-3

(FID)Beforeconditioning

2 hconditioning

17 hconditioning Inlet Desorb Outlet

A 100 0.06 0.07 0.06 98 63.0 21.4 80.3 104.6B 100 0.03 0.04 0.03 110 62.8 21.3 81.0 85.2C 104 0.06 0.06 0.04 95 62.9 21.4 80.6 107.7D 101 0.04 0.06 0.03 92 62.8 21.7 82.0 99.6E 103 0.09 0.09 0.07 94 62.6 21.4 81.9 108.7F 100 0.03 0.04 0.03 105 62.3 21.4 81.8 101.8

101.3 ± 9(mean±s.d.)

G 531 0.25 0.26 0.25 30 62.5 21.3 80.7 93H 528 0.22 0.25 0.23 31 62.7 21.6 80.8 102.6I 529 0.22 0.25 0.23 31 62.4 21.3 81.4 99.6J 526 0.22 0.25 0.23 31 62.2 21.3 81.9 90.2K 527 0.23 0.25 0.23 30 62.2 21.3 82.2 104.9L 527 0.23 0.25 0.23 30 62.6 21.3 81.9 94.5

97.5 ± 6(mean±s.d.)

T D T S

Back-pressure regulated ECC (Figure 3) is therecommended carrier gas control configurationfor Markes thermal desorbers and has theadded advantage of being integrated with theGC/MS software. This allows access to the fullrange of ECC-based software enhancementsavailable from leading GC/MS suppliers. (SeeTDTS 47 The Analysis of Landfill GasCompounds using Thermal Desorption GC/MSand a Retention Time Locked Database andTDTS 66 Improving the identification andmeasurement of trace odorous and toxiccomponents during materials emissionstesting.)

Potential analytical impact ofvarying tube humidity/solventlevelsResidual solvent or exceptionally high humidity,in standard or sample tubes respectively, ismuch more likely to cause significant analyticaldiscrimination than impedance variation. It isone of the most common causes of quantitativeerrors during measurement with thermaldesorption – GC(/MS).The issue can arise during inlet, outlet ordouble split methods. Even splitless desorptionmethods, though immune to splitdiscrimination, can suffer from adversechromatographic effects due to high solvent orwater content, which can affect quantitation.


www.markes.com


ECCCarriergas Desorbflow Inlet splitflow

Heated valve

Sorbe

nt tub

e

Cold

trap

Re-co

llectio

n tub

e

Link toECCseptumpurge

ECCCarriergas Desorbflow Outletsplit flow

Heated valve

Sorbe

nt tub

e

Cold

trap

Re-co

llectio

n tub

e


Figure 3a: UNITY TD system with ECC control,flow path during primary (tube) desorption

Figure 3b: UNITY TD system with ECC control,flow path during secondary (trap) desorption

T D T S

An example of discrimination due to variablesolvent content is shown in Table 4 (reproducedfrom Report OMS/2002/15 with the kindpermission of UK HSL.) In this case, 5.0 µL of toluene solution inmethanol (3.4 mg/mL) was introduced to tubespacked with two different masses (100 mg and530 mg) of Chromosorb 106 sorbent, in astream of carrier gas (50 mL/min). Theanalytical conditions used were the same asthose used for analysis of the gas standard(see Table 3) in which no discrimination due toimpedance variation was observed. The 26%reduction in data between tubes packed with100 mg and those packed with 530 mg ofChromosorb 106 was thus purely due to the

excess solvent retained by the tubes with thelarger mass of sorbent. The difference in thetwo measurements is due to split discriminationcaused by flash vaporisation of the retainedsolvent during desorption. As the solventvaporises and expands this raises the gaspressure inside the TD flow path, temporarilyincreasing the split flow relative to the desorband/or column flow. Analytes desorbing/eluting from the sorbent at the same time asthe solvent will thus be subjected to a highersplit flow/ratio than later eluting compounds.


www.markes.com


Table 4: Recovery of ~20 µg toluene from TD-GC system, operating with inlet split and at 46.1 psipressure. 26% lower recovery is observed for the higher impedance tubes due to solvent

retention. (Toluene loaded from the liquid phase)

Tube C 106mass(mg)


(mL/min)

TD split flows(mL/min)

Toluenepeak area

µV.s-1(TIC)Before

conditioning2 h

conditioning17 h

conditioning Inlet Desorb Outlet

A 100 0.06 0.07 0.06 98 63.0 21.4 80.3 1086B 100 0.03 0.04 0.03 110 62.8 21.3 81.0 1081C 104 0.06 0.06 0.04 95 62.9 21.4 80.6 1084D 101 0.04 0.06 0.03 92 62.8 21.7 82.0 1090E 103 0.09 0.09 0.07 94 62.6 21.4 81.9 1087F 100 0.03 0.04 0.03 105 62.3 21.4 81.8 1090

1086 ± 4mean±s.d.

G 531 0.25 0.26 0.25 30 62.5 21.3 80.7 822H 528 0.22 0.25 0.23 31 62.7 21.6 80.8 842I 529 0.22 0.25 0.23 31 62.4 21.3 81.4 792J 526 0.22 0.25 0.23 31 62.2 21.3 81.9 792K 527 0.23 0.25 0.23 30 62.2 21.3 82.2 784L 527 0.23 0.25 0.23 30 62.6 21.3 81.9 770

801 ± 27mean±s.d.

T D T S

Split discrimination of this kind has beenextensively reported for all forms of flashvaporising GC injection. The optimum solutionsare to programme or slow down the heatingrate of the sorbent tube/trap (such that thesolvent or water vaporises more slowly andwithout causing a pressure surge) and tominimise the mass of solvent/water retained.Markes thermal desorption systems areuniquely designed to minimise splitdiscrimination due to differences in humidity/solvent content between standards and sampletubes. Relevant features include: • Automated on-and off-line options for dry

purging of sample tubes, in the samplingdirection. This helps reduce anydifferences in humidity content betweenthe sorbent tubes in a sequence.

• The primary tube desorption oven, heatsfrom ambient at ~3°C/sec at the start oftube desorption, to ensure gradual solvent/water vaporisation and to eliminate anyrisk of a pressure surge which mightchange the inlet split ratio. Note that thedata shown in Table 4 was generatedusing a pre-heated tube desorption ovenwhich is more prone to discrimination dueto flash vaporisation.

• The option of inlet splitting, withquantitative re-collection for validation. Incases where analyte concentrations aresufficiently high to allow a moderate tohigh split (>20:1), the mass of water orsolvent reaching the focusing trap can besignificantly reduced by introducing a spliton the inlet to the cold trap i.e. duringprimary (tube) desorption.

• The internal focusing trap of Markesthermal desorbers has been optimised forselective elimination of water/solvent tominimise risk of discrimination. Relativeto older TD technology, the Markes trapcontains an extended (60 mm) sorbentbed ensuring quantitative retention oftarget analytes without sub-ambientcooling. The 60 mm bed length of sorbentused in every Markes thermal desorbermatches the maximum sorbent bed lengthused in industry standard sample tubesand allows the type and bed length ofsorbents used in the focusing trap toreplicate those used for air monitoring.

While the I.D. of the focusing trap is muchsmaller than that of a sample tube (toensure optimum desorption efficiencyduring capillary GC analysis) the volumeof gas passing from the sample tube tothe focusing trap during primary (tube)desorption (typically <500 mL) isinvariably lower than that used for tubesampling (typically >5L). Therefore, anycompound which can be quantitativelyretained by the sample tube during airmonitoring at ambient temperatures, canalso be quantitatively retained by thefocusing trap at ambient temperature.The advantage of this is that most volatilesolvents and water can be selectively, butnot completely, purged from sorbenttubes/traps at ambient temperature. Ineffect the extended sorbent bed of theMarkes focusing trap thus introduces asecond automated dry purge/solventpurge step allowing the mass of water/solvent retained by the system to bereduced, typically by another order ofmagnitude.

This combination of TD features minimizes therisk of analytical discrimination due to variablemasses of water/solvent in sample/standardtubes. Moreover this aspect of system design iscomplemented by the SecureTD-Q facilityoffered by Markes thermal desorbers.SecureTD-Q allows quantitative re-collection ofboth inlet and outlet split (manual orautomated (100-tube) configurations available)allowing users to repeat analyses and validatethe analytical data. Application of SecureTD-Qfor thermal desorption method and datavalidation is described below.


www.markes.com


T D T S

Using SecureTD-Q to validatethermal desorption dataPioneered by Markes International in 1998,quantitative re-collection of TD samples(SecureTD-Q) has set a new benchmark foranalytical thermal desorption technology.SecureTD-Q is available on all Markes thermaldesorption systems and allows repeat analysisand validation of TD methods/data (e.g. asreferenced in international standard methodssuch as ASTM D6196-03). By facilitatingquantitative re-collection of both inlet andoutlet split flow, SecureTD-Q is ideal forevaluating thermal desorption analytical datafor possible discrimination due to varyingimpedance or solvent/water levels. [Applicationof SecureTD-Q for TD method/data validation isexplained in the associated Markes brochure.] Quantitative re-collection of inlet split is

essential as a test for discrimination due totube impedance variation as this can only bean issue for inlet split methods (see above).Inlet split methods are also most susceptible todiscrimination due to variable solvent/waterlevels. Application of SecureTD-Q as avalidation tool in both cases is illustrated belowwith examples: Example 1:Using low impedance Certified ReferenceStandard (CRS) tubes packed with 200 mgTenax for quality assurance of various types offield monitoring tube under the following TDanalytical conditions:

Carrier pressure: 8 psiInlet split flow: Set to 100 mL/minDesorb flow (trap flow during primary(tube) desorption): Set to 25 mL/min


www.markes.com


25mL/min

100mL/min

Heated valve

Sorbe

nt tub

e

Cold

trap

Re-co

llectio

n tub

e


1 mL/min

49mL/min

Heated valve

Sorbe

nt tub

e

Cold

trap

Re-co

llectio

n tub

e


Figure 4a: Primary (tube) desorption showinggas flows for example 1

Figure 4b: Secondary (trap) desorption showinggas flows for example 1

T D T S

Outlet split flow: Set to 49 mL/minColumn flow: 1 mL/minRe-collection tube: 200 mg of conditionedTenax

The desorption process is illustrated in Figure4. Split ratios:

Inlet split: 25/125 = 1/5Outlet split: 1/50Total split: 1/250

Primary and repeat analyses of a lowimpedance (Tenax) tube containing 10 µgeach of toluene and n-C12

Primary analysis:

Repeat analysis (i.e. analysis of the re-collected sample):

Note that analytical data from the re-collected sample is almost identical(99.6%) to data from the primary analysisfor both compounds reflecting theexpected split ratios and indicating nodiscrimination/loss.

However, if the analysis of high impedance,multi-sorbent tubes, using these analyticalconditions, gave an effective inlet split ratio of1/5.5 rather than the expected 1/5, theresults of the primary (original sample) andsecondary (re-collected sample) analysis, wouldbe as follows:Primary and repeat analyses of a highimpedance tube containing 10 µg each oftoluene and n-C12

Primary analysis:

Repeat analysis (using a low impedanceTenax re-collection tube):

Note that, in this case, results from theprimary analysis (with 36.4 ng of eachcompound reaching the column anddetector) are lower than those from therepeat analysis (with 39.9 ng of eachcompound reaching the column anddetector), thus clearly indicating thatdiscrimination had occured in the primaryanalysis. Moreover, the uniformity of thediscrimination across the volatility rangepoints to impedance rather than flashvaporisation of water/solvent as theprimary cause.


www.markes.com


Amount of toluene & n-C12 reachingfocusing trap2 µg

(1/5 inletsplit)Amount of toluene & n-C12 re-collectedduring tube desorption 8 µg

Amount of toluene & n-C12 reachinganalytical column & detector40 ng

(1/50 outletsplit)Amount of toluene & n-C12 re-collectedduring trap desorption 1.96 µg

Total amount of each compoundre-collected 9.96 µg

Amount of toluene & n-C12 reachingfocusing trap1.99 µg(1/5 inletsplit)

Amount of toluene & n-C12 re-collectedduring tube desorption 7.97 µgAmount of toluene & n-C12 reachinganalytical column & detector

39.8 ng(1/50 outletsplit)

Amount of toluene & n-C12 re-collectedduring trap desorption 1.95 µg


Amount of toluene & n-C12 reachingfocusing trap1.82 µg(1/5.5 inletsplit)

Amount of toluene & n-C12 re-collectedduring tube desorption 8.18 µg

Amount of toluene & n-C12 reachinganalytical column & detector36.4 ng(1/50 outletsplit)



Amount of toluene & n-C12 reachingfocusing trap1.99 µg(1/5 inletsplit)

Amount of toluene & n-C12 re-collectedduring tube desorption 7.97 µg

Amount of toluene & n-C12 reachinganalytical column & detector39.9 ng

(1/50 outletsplit)Amount of toluene & n-C12 re-collectedduring trap desorption 1.95 µg


T D T S

Example 2: Using a dry (and solvent-free) standard tocalibrate a field sample tube containing 1 mg ofwater under the following TD analyticalconditions:

Carrier pressure: 15 psi, Inlet split: Flow offDesorb flow (trap flow during primary (tube)desorption): Set to 40 mL/minFocusing trap temperature: Set below 0°COutlet split flow: Set to 19 mL/minColumn flow: 1 mL/minRe-collection tube: 200 mg of conditionedTenax

The desorption process is illustrated in Figure5.

Split ratios:Outlet split: 1/20Total split: 1/20

Primary and repeat analyses of a drystandard tube containing 1 µg each of tolueneand n-C12Primary analysis:


www.markes.com


40mL/min

OFF

Heated valve

Sorbe

nt tub

e

Cold

trap

Re-co

llectio

n tub

e


1 mL/min

19mL/min

Heated valve

Sorbe

nt tub

e

Cold

trap

Re-co

llectio

n tub

e


Figure 5a: Primary (tube) desorption showinggas flows for example 2

Figure 5b: Secondary (trap) desorption showinggas flows for example 2

T D T S


(No inletsplit)Amount of toluene & n-C12 reachinganalytical column & detector

50 ng(1/20 outletsplit)


Total amount of each compoundre-collected 950 ng

Repeat analysis (i.e. analysis of the re-collected sample):

Note that, as expected, analytical datafrom the re-collected sample is 5% lowerthan the primary analysis for bothcompounds reflecting the expected splitratio and indicating no discrimination/loss.

However, if the analysis of tubes containing 1mg of water produced an immediate pressuresurge, under these analytical conditions, givingan effective split ratio of 1/28 rather than theexpected 1/20 during the early stages ofsecondary (focusing trap) desorption, theresults of the primary (original sample) andsecondary (re-collected sample) analyses,would be as follows:Primary and repeat analyses of a tubecontaining 1 mg of water and 1 µg each oftoluene and n-C12Primary analysis:

Repeat analysis (N.B. water will be purgedfrom the re-collection tube during re-collection thus the re-collected sample willbe dry):

Note that, in this case, results from theprimary analysis of toluene (35.7 ng) werelower than the repeat (48.2 ng) indicatingdiscrimination, but that primary andrepeat analysis data for n-C12 (50 & 47.5ng respectively) were as expected. Thisindicates that the discrimination was notuniform across the volatility range. Datashowing lower than expected primaryanalysis results for more volatilecompounds, but with expected recovery ofless volatile compounds indicatesdiscrimination caused by the pressuresurge of flash vaporisation ofsolvent/water. This would be most simplyaddressed by resetting the focusing traptemperature to +30°C, allowing water tobe selectively purged from the systemduring primary (tube) desorption.Note that if an analysis, such as that describedin example 2, was carried out with higheranalyte levels (i.e. requiring both inlet andoutlet split) and using a desorber that operatedwith a pre-heated oven; risk of flashvaporisation could effect BOTH the inlet andoutlet split. The precautionary steps describedabove for minimising residual water/solventlevels should be implemented and quantitative,re-collection of both inlet and outlet split wouldbe required to demonstrate that nodiscrimination had occured in either phase ofTD operation.

Applications were performed using the stated analytical conditions.Operation under different conditions, or with incompatible samplematrices, may impact the performance shown.


www.markes.com


Amount of toluene & n-C12 reachingfocusing trap950 ng(No inletsplit)

Amount of toluene & n-C12 reachinganalytical column & detector47.5 ng(1/50 outletsplit)

Amount of toluene & n-C12 re-collectedduring trap desorption 902.5ng

Total amount of each compoundre-collected 902.5ng


(No inletsplit)Amount of toluene reaching analyticalcolumn & detector


Amount of n-C12 reaching analyticalcolumn & detector 50 ng

(1/20 outletsplit)Amount of toluene re-collected duringtrap desorption 964.3

ngAmount of n-C12 re-collected duringtrap desorption 950 ng

Amount of toluene reaching focusingtrap964.3 ng

(No inletsplit)Amount of n-C12 reaching focusing trap 950 ng

(No inletsplit)Amount of toluene reaching analyticalcolumn & detector


Amount of toluene re-collected duringtrap desorption 916.1 ng

Amount of n-C12 reaching analyticalcolumn & detector 47.5 ng(1/20 outletsplit)

Amount of n-C12 re-collected during trapdesorption 902.5 ng

T D T S

Thermal Desorption Technical Support€¦ · tube sorbents for calibration to those used for field...

Documents

Transcript of Thermal Desorption Technical Support€¦ · tube sorbents for calibration to those used for field...