The analytical method described below is by gas chromatography and FID.
NILU had several years of experience with a hand-made, quite time consuming, non-automated method. This method was based on a two-step concentration of up to one litre of air, dried with K2CO3 and NaOH on support (revocation of CO2), a gas chromatographic separation on Al2O3 PLOT column and FID detection (Schmidbauer and Oehme, 1985, 1986).
Considering the high number of samples which needed to be analysed within a very short time after sampling, automated instrumentation was needed. A prototype of a new instrument was presented by Jack Mowrer (IVL, Gothenburg, Sweden) at the Lindau workshop (EMEP/CCC-Report 3/90). This was an instrument for continuous unattended measurements of C2-C5 hydrocarbons at background levels. The instrument was designed and built together with Chrompack (Middelburg, The Netherlands). In the years after Lindau this instrument was further modified by Chrompack in cooperation with H. Bloemen (RIVM, Bilthoven, The Netherlands). More and more laboratories, especially within the EUROTRAC/TOR community, have since then based their on-line measurements of hydrocarbons on a commercial available type of this instrument. The last aspect the increasing number of users was one of the reasons for our decision to do the analyses for EMEP with this instrument. The instrument, as it comes from the factory, has to be modified to some extent to act as proper as some of the various homemade devices. The users have introduced some good improvements.
A brief description of the set-up and procedures are given
below. The analysis is complicated and should preferably be learned by
training.
Figure 4.15.1 shows the set-up with the original Nafion dryer (not used by NILU) and the 10 canister stream-selector.
Figure 4.15.2 shows the set-up with the dryer as it is used at NILU.
A drying-tube with backflush and heating option has been added between valve V3 and V4. (10 cm teflon-tube 1/4 with 20 micron steel-sinters on both ends, filled with 3 cm of K2CO3 on both ends and 4 cm NaOH on support in the middle.)
The first trap is a 1/4 glass-tube packed with Carbosieve, Carbotrap and Carbotrap C - the refocussing trap a 10 cm piece of coat fused silica (Poraplot U).
Figures 4.15.3 and 4.15.4 show the flow-schemes of the seven different steps in an analytical cycle of the thermodesorption unit.
Al2O3/KCl PLOT column, 50 m, 0.32 i.d. (Chrompack).
VG Minichrom PC based chromatographic data handling system.
All gases are grade 4 or better. They are further cleaned by
passage through two different 200 ml cylinders filled with activated charcoal
and molecular sieve. This is sufficient for the FID gases, but not for the
helium as carrier gas. Helium is further cleaned in a 1/4 steel trap filled
with molecular sieve at liquid nitrogen temperature. All tubes which are in
contact with the sample are premium grade stainless steel or teflon (drying
tube).
The analytical procedure for this type of equipment is given
in Figures 4.15.3 and 4.15.4.
Table 4.15.1: Settings of temperatures, gases, flow-directions and gas-flows in the different steps of an analytical cycle of the thermodesorption unit.
Step |
Trap 1 |
Gas |
Direction |
Flow |
Trap 2 |
Gas |
Time |
I |
270°C |
He |
Back |
20 ml |
120°C |
He |
26 min |
II |
Ambient |
He |
Back |
20 ml |
Ambient |
He |
60 min |
III |
-30 |
He |
Back |
20 ml |
Ambient |
He |
6 min |
IV |
-30 |
Sample |
Front |
25 ml |
Ambient |
He |
14 min |
V |
-30 |
Sample |
Front |
25 ml |
-180 |
He |
6 min |
VI |
-30 |
He |
Front |
5 ml |
-180 |
He |
1 min |
VII |
250 |
He + HC |
Back |
8 ml |
-180 |
He + HC |
8 min |
The consume of liquid nitrogen is about 2 litres per sample.
A chromatogram is given in Figure 4.15.5.
NILU is using calibration-gas-mixtures from NIST (USA) to determine the carbon-number-response for the FID.
The standard-gas-mixture is analysed with the same conditions as a sample (same pressure, flowrate, volume and time period) connected to one of the ten ports of the stream-selector. The absolute precision of the read-out of the massflow-controller is therefore not a critical point in calculating concentrations.
NILU is using the n-butane response calculated from a 10 ppb NIST standard as a basis for calculation of carbon-number-response.
The standard gas is connected to port 1 of the multistream-selector and therefore always the first and thereafter every tenth injected sample.
All hydrocarbon peaks in the chromatograms are identified and integrated by hand.
For identification purposes ppm standard-gas-mixtures from Scotty or self-made standards are injected via a home-made injection system.
The blank values of the carrier gas are checked by direct connection to one of the ports of the multiposition-valve. A good performance of carrier-gas cleaning is of fundamental importance for a reliable analysis of the very volatile hydrocarbons.
The blank values of the instrument especially of the traps have to be controlled quite often. High boiling compounds on the traps may decompose and give higher background noise of several compounds.
A need for change of the drying-agents is indicated by bad chromatographic performance. On a routine basis the tube is renewed every week.
A record of the raw-area of the standard-runs is very useful to judge the instruments behaviour over longer time.
All gas-flows need to be checked quite often.
Schmidbauer, N. and Oehme, M. (1985) Analysis of light hydrocarbons (C2C6) at ppt levels by high resolution gas chromatography. J. High Res. Chrom. & Chrom. Commun., 8, 404-406.
Schmidbauer, N. and Oehme, M. (1986) Improvement of a cryogenic preconcentration unit for C2C6 hydrocarbons in ambient air at ppt levels. J. High Res. Chrom. & Chrom. Commun., 9, 502-505.
EMEP (1990) EMEP Workshop on measurement of hydrocarbons/VOC. Lindau, Federal Rep. of Germany, November, 69, 1989. Lillestrøm, Norwegian Institute for Air Research (EMEP/CCC-Report 3/90).
Figure 4.15.1: Original set-up with Nafion dryer (not used by NILU) and the 10 canister stream.
Figure 4.15.2: Instrument set-up with dryer as used at NILU.
0 Run
The
system is put into the Run mode by pressing the <Standby>
key, turning off the LED.
See B: Run mode, above.
NOTE From software revision SWL 1.03L onwards,
the system cannot be started from any other stage.
I Backflushlinjection
The
cold trap is flash heated and the sample, which was cryo-focussed
in stage 7. is injected into the capillary column. At the same time the
adsorption tube is heated and cleaned by a backflush gas flow to prepare it
for the next sample collection.
II Backflush/Stop-wait
This
stage allows the user to synchronize the auto-TCT with the GC temperature
program or his daily sample frequency (see Auxiliary run functions: Stop).
III Backflush/Oven precool
The sample
tube is cooled down to between -170°C and approximately 28°C to prepare
it for the next sample collection.
Figure 4.15.3: Analytical procedure - step 0-III.
IV Sample collection
The air sample is sucked through the adsorption tube by
means of a vacuum or pushed through by means of pressure.
V Sample collection/Trap precool
While sample
collection is still in progress, the cold trap is cooling down to between -100°C
and
-150°C to prepare it for cryo-focussing of the collected
components onto the fused silica trap.
VII Preflush
If
necessary, water is removed from the adsorption tube by a dry carrier gas flow
before the actual
desorption starts and the components are transferred to the
cold trap.
VII
Desorption/trapping
The adsorption tube is heated and the
components are transferred to the cold trap by the carrier gas flow and
ayofocussed.
Figure 4.15.4: Analytical procedure - step IV-VII.
Figure 4.15.5: Chromatogram of light hydrocarbons by GC.