the use of photoionization,flameionization and electron capture detector inseries for low mw trace...

8/4/2019 The Use of Photoionization,Flameionization and Electron Capture Detector Inseries for Low MW Trace Components in Non Urban Atmospheres

1/6

ematic.,rder fit:ground:hieved,led thaty in the;. hydro

n-London-

lams, Anal.

\. ."

The Use of P h o t o i o n i ' ~ a t i o n ,Flameionization and ElectronCa ptu re Detectors in Series forthe Determination of LowMol.ecular Weight TraceComponents in the Non UrbanAtmospheretJ. RUDOLPH AND C. JEBSENI ns ti tu t fu r Chemie 3: Atmosphiirische Chemie der Kernforschungsanlage


2/6

INTRODUCTIONThe spectrum of organic compounds in the atmosphere-as well inpolluted urban and industrial areas as in non urban regions-is verycomplex and their abundance may vary considerably, especially for themore reactive species with short atmospheric residence time. In additionthe mixing ratios of many important trace gases are low, a few ppb (10-9)to a fraction of a ppb. The most widely used technique for themeasurement of organic species in the atmosphere is gaschromatographycombined with ionizat ion detectors or mass spectrometry. Althoughgaschromatography combined with mass spectrometry---especially in thespecific ion monitoring mode-is a very selective and sensitive analyticalmethod, the expense of a GCMS system is i n a d e q u a ~ high for a routineenvironmental monitoring instrument. Very often a ~ p e c i f i c


3/6

FIGURE I Schematic diagram of gas chromatograph and inlet system.

earriergas I vonl smaller a ir samples the drying tube is installed between enrichment andseparation column. Larger air samples are dried previous to theenrichment step. After sample concentration the rotary valve is switchedinto the "inject ion" posit ion and the temperature of the precolumn israised-by direct resistance heating-to about 530K. During sampledesorption the separat ion column is kept at 170 K to preventpeakbroadening during injection. The column is heated programmed from170K to 320K with 25Kmin- 1 and from 320K to 400K with 2Kmin- 1and finally kept at 400 K for up tp 30 minutes if a completechromatogram is run.Qualitative peak identification is made according to the retention timesand by co-chromatog raphy. In addition, the peak identity can beconfirmed by the signal ratios for the different detectors (see below). Forquantitative evaluations, the peak heights (or sometimes the peak areas)are compared with those from standards of known trace gas mixing ratios.Appropriate standards are prepared by dilut ion of the pure compoundswith purified synthetic air in a static dilution system (two or three dilutionsteps).The air samples are collected in 2dm3 stainless steel containers (in

special cases also containers of 10 dm 3 volume are used) and transferredinto the laboratory for analysis. The design and conditioning of suchsample containers have already been described in detail. 11

RESULTS AND DISCUSSIONAn example for a chromatogram from an air sample collected in a semirural area in central Europe (some km outside the small town of Jiilich inWestern Germany) is shown in Figure 2 a-c. The FID trace is shown in2a, ECD and PI D (117. eV photonenergy) trace in 2b and 2c. In Figure2d the trace ofthe Pl O with a lO.2eV lamp from a different run is shown.The sample and the sample volume are the same as for Figure 2a-c. Asexpected, the possibilities to detect and identify different trace species aresignificantly increased by the use ofa multidetector system, compared to asingle detector.The primary question is, whether such a multidetector system meets therequirements for the analysis of light trace species in the non urbanatmosphere (e.g. linearity, reproducibility, detection limits). Table I liststhe average and the standard deviation of the peakheights from sevenrepetitive measurements of the same air sample (0.5 dm sample volume).These seven measurements were made within three days, thus the standarddeviation includes also any variation due to instrument drifts. For most

-'

Co/';;C'"

vent

make uP9as m~ h y d r 0 9 . n~ Q l r

column "".nlseparation calUI1'l\IAll AAAAAILJ.

1 I . . .m1I 11 NNJ

11 bac:ktluoll col""" Ie n r i c h m ~ t pump carrierprteolumn 9as 1I

sample

- - - sampling/bation of a variety of low molecular weight compounds and thepossibil ity to operate this colume at subambient temperatures down to170 K. The separation column consists of two parts, one of 6m length andone o f 3m length. The 3m piece is used for backilush in order to avoidthe accumulat ion of species of low volati li ty on the separat ion columnwhich would result in increased baseline noise and drift. Fo r the backflushsystem a ten port switching valve-which simultaneously serves as sampleinjection valve-is used. Since-with the exception of a few speciessample volumes of some cm 3 are not sufficient for the determination oftrace species in the non urban atmosphere, the sample loop of the 10 portvalve is substituted by an enrichment precolumn. Various adsorbents aredescribed in the l ite rature for the enri chment of trace gases from airsamples.6 . S - IO We decided to use porous glass beads (>:::60 mesh) in a10 cm long 1/8" stainless steel precolumn at >::: 100 K for samplepreconcentration. Since the packing of the separation column is sensitiveto moisture and in order to prevent the separat ion column to be cloggedby water, the air samples are dried by means of a tube packed withmagnesiumperchlorate. The measurements of light halocarbons andhydrocarbons are not adversely affected by this drying procedure. For

'- , ~ '\.....t


4/6

0>,5 IS: 1'f ."10

TABLE 1Reproducibility of peakheights for some selected trace gasest

FIGURE 2 Chromatograms oblained from a 1.7dm3 air sample, separation conditions seetext. a: FIO trace, b: ECO trace, c: PID (11.7 eV) trace, d: second separation, PlO with10.2eV, same sample and sample size as a-c.

tThe data represent the average o( seven measurtmems within Ihree days and their relative standard deviation. Samplevolume was O.Sdm.1.ttErrors or the absolute ~ l i b r a l i o ( \ are not in


5/6

-'1:.'

-IT:

t 10-1>et:x~ ' O - 2 1 ""1111 1 II . l11111 Iltl",,' "I10 ., __ ":I . ~

Amount of substance, pg -CH3C1 0CC14


6/6

sample with a PID at 1O.2eV photonenergy corresponds to ~ O . 7 p gbenzene absolut. This is slightly less than the value of 0.3 ppb with 1cm3air sample and a signal to noise ratio of 2: 1 reported by Hester andMeyer 12 for measurements with a 1/8" separation column underisothermal conditions in combination with a photoionization detector ofident ical type. This improvement is probably due to the use of aseparation column with only 0.8 mm i.d. and of a sta tionary phase with anextremely low bleeding.The chromatograms in Figure 2 clearly demonst ra te the differentresponse of the various detectors for the same compound. For aconsiderable number of species valuable qualitative informations on thetype of compound (e.g. halocarbon, unsaturated hydrocarbon elc.) can bedirectly obt ai ned from a comparison o f the signals o f the differentdetectors. The differences in the signal rat ios are'jot o n ~ , valuable inorder to identify the individual compounds, but can also b ~ s e d to test ifthere are any significant interferences from other species, 'provided theirresponse ratios are sufficiently different. Since the errors of the peakheightsare only a few percent, the presence of 10% of an interfering substance ina chromatographic peak could already be recognized if the response ratiosfor two detectors of the two compounds differ by more than a factor of 2.For most compounds the differences in the response ratios for the variousdetectors are even larger, as can be seen from the ch romatograms inFigure 2.The main problem with the use of several detectors in series is thepossible decomposi tion of par t of the compounds in one of the detectors ,except the last one. For most of the trace species we tes ted in this work(that is mainly hydroca rbons and halocarbons) no such effects wereobserved. However, there is one surprising exception, cthyne-also presentand detectable in the PID (J 1.7 eV lamp)-

the use of photoionization,flameionization and electron capture detector inseries for low mw trace...

Documents