; EPA/600/A-92/251
EVALUATION OF A SORBENT-BASED
PRECONCENTRATOR FOR ANALYSIS OF VOCS IN AIR
USING GAS CHROMATOGRAPHY - ATOMIC EMISSION
DETECTION
Karen D. Oliver and E. Hunter Daughtrey, Jr.
ManTech Environmental Technology, Inc.
P. O. Box 12313
Research Triangle Park, NC 27709
William A. McClenny
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
ABSTRACT
A Hewlett-Packard 5890 gas chromatograph and 5921A atomic emission detector (AED)
were used to determine volatile organic compounds (VOCs) at part-per-billion-by-volume levels
in ambient air samples which were preconcentrated by using the Dynatherm ACEM 900 sorbent-
based preconcentrator. Several combinations of multisorbent sampling tubes and focusing tubes
were tested. Mixtures of 51 VOCs including 10 polar compounds were prepared in humidified
scientific-grade air and were used to evaluate the system with regard to compound recoveries,
linearity of compound concentration with varying sample volume, and the optimum volume of
purge gas needed to remove water from the sorbent before thermal desorption. The automated,
unattended operation of the system was also evaluated by allowing the instrument to sample indoor
air at intervals of approximately 1 h over a 24-h period.
Because individual elements are detected by the AED, the hydrogen response due to water
may be monitored concurrently with the response of other elements. This allowed a thorough
investigation of the effect of water vapor on the carbon, chlorine, and bromine response for those
compounds in the standard mixtures which coelute with water. Also, the relative amount of water
vapor still present in the sample after various drying techniques were employed was easily
monitored. These results and the results of the experiments mentioned above are discussed in this
paper.
This paper has been reviewed in accordance with the U.S. Environmental Protection
Agency's peer and administrative review policies and approved for presentation and publication.
Mention of trade names or commercial products does not constitute endorsement or
recommendation for use.
INTRODUCTION
Currently, the U.S. Environmental Protection Agency is evaluating automated gas
chromatographic systems (autoGCs) for use in network monitoring stations.1 Desirable features
of these systems include (1) the need for little or no liquid nitrogen, (2) the capability for
unattended, continuous operation, (3) the capability for drying the sample stream without removing
polar volatile organic compounds (VOCs) and (4) easy deployment in the field. The autoGC
system being evaluated in our laboratory utilizes a Dynatherm Automated Continuous
Environmental Monitor (ACEM) Model 900 sorbent-based sample preconcentrator and a Hewlett-
Packard 5890 GC and 5921A Atomic Emission Detector (AED). The AED has been a useful and
interesting detector for the laboratory evaluation of the svstern but has never been considered.
1
-------
suitable for field deployment because of the fragility of the GC-AED interface and support gas
requirements.
EXPERIMENTAL
A Dynatherm Analytical Instruments, Inc. (Kelton, PA) ACEM 900 for sample
preconcentration and thermal desorption is interfaced to a Hewlett-Packard (HP, Avondale, PA)
5890 GC which is equipped with an HP 5921A AED. The ACEM 900 is a sorbent-based system
which employs two tubes; one sorbent tube collects sample and a second, narrower sorbent tube
focuses the sample prior to thermal desorption onto the capillary column. A Dynatherm External
Sampling Module is used to load sample onto the collection tube from a canister or to pull ambient
air through the tube by using a vacuum pump. Helium may be used to purge water from the
collection tube prior to desorption of the sample onto the focusing tube. The 1-m x 0.20-mm
deactivated fused-siliea transfer line which connects the ACEM 900 to the GC column was heated
to 200 "C. A 60-m x 0.32-mm x 5-^m DB-1 capillary column (J & W Scientific, Inc., Rancho
Cordova, CA) was used for the experiments discussed here, and the GC oven temperature was
programmed as follows: 6 min at 30 *C, an 8 "C/min ramp to 240 *C, and a 10 min hold at
240 * C. For the analyses, the AED transfer line and cavity block were heated to 250 * C, and the
AED was programmed to monitor responses of emission lines of carbon at 496 nm, hydrogen at
486 nm, chlorine at 479 nm, and bromine at 478 nm.
Challenge gas mixtures for the experiments included 6-L canister samples of a mixture of
10 ppbv of the 41 VOCs on the EPA Compendium Method TO-14 target list2 in humidified air
at -50% RH. The canisters were prepared3 from a cylinder containing 1-2 pprn of each
compound in nitrogen (Alphagaz, Walnut Creek, CA). Also used were canister samples of a
mixture of 10—20 ppbv of 10 polar compounds (methanol, ethanol, isopropanol, butanol, acetone,
methyl ethyl ketone, acetonitrile, acrylonitrile, methyl methacrylate, and ethyl acrylate in humidified
air at -50% RH) which were prepared from cylinders containing 10 ppm of the compounds in
nitrogen (Scott Specialty Gases, Plumsteadville, PA). Mixtures of C^-C,^ compounds at
concentrations of 15—100 ppm in nitrogen (Scott Specialty Gases) were also used to spike the tubes.
This was accomplished by moving the collection rube from the ACEM 900 to a Dynatherm Model
10 tube conditioner and injecting the sample from a gaslight syringe into a stream of nitrogen
flowing at 50 cm3/min through the tube. The collection and focusing tube combinations
(Dynatherm Analytical Instruments, Inc.) tested are presented in Table I.
Initially, the effect of different helium purge volumes (used to remove water from the
collection tube) on the response of VOCs was investigated. In these experiments, a 480-cm3 sample
of the 41-compound mixture was collected on the sorbent tube from a 6-L canister by using a
0-500-sccm mass flow controller (MFC, Tylan General, Torrance, CA) set at 80 cm3/rnin. The
tube was then purged with 26 cm3/rnin of helium; the helium purge volumes used were 0, 50, 100,
250, 500, and 1000 cm3. The collection tube was normally held at 40 °C for these experiments,
although some experiments were repeated with the tube at 55 and 65 * C. Sample was desorbed
from the collection tube onto the focusing tube for 3 min at 200 or 300 "C, followed by 2 min in
cool mode (in which helium continues to flow through the tube in the desorb position while the
tube is cools down from the desorb temperature). Then, the focusing tube was heated to 350 * C,
as indicated by a thermocouple located outside the tube, for 3 min to desorb sample onto the GC
column.
The linearity of response of the 41 VOCs and the polar compound mixture was investigated
by collecting sample volumes of 250, 500, 1000, and 2000 cm3. Again, an MFC set at 80 cm3/min
was used to load the samples from 6-L canisters onto the collection tube, and the tube was purged
with 500 cm3 of helium to remove residual water. Tube heating and cooling parameters were the
same as those listed above.
-------
Table I. Tube combinations
Experiments
Vary Retention
Tubes (Sorbent* and Focusing) He Purge of Q 24-h
Volume linearity Compounds Monitoring
Tenax-TA/Ambersorb XE-340/CharcoalX X ' X X
+ Tenax-TA/Silica gel/Ambersorb XE-340/
Charcoal
Carbotrap C/Carbotrap B/Carboxen 1000 X X X X
+ Tenax-TA/Silica gel/Ambersorb XE-340/
Charcoal
Tenax-TA/Carboxen 1000 X X
+ Carbotrap B/Carboxen 1000
Tenax-TA/Ambersorb XE-340/Charcoal X X
+ Carbotrap B/Carboxen 1000
*6-mtn-o.d. sorbent tubes.
The retention of ethane, ethylene, and acetylene on different collection and focusing tube
combinations was investigated by spiking 6-mm-o.d. and 10-mm-o.d. tubes with 1-10 cm3 of the
Q-Cfi gas mixtures as discussed above. In addition to the tube combinations in Table I, nine tube
combinations were evaluated by using one of each of the following 10-mm-o.d, collection tubes:
Tenax-GR/Carboxen 1000, Tenax-GR/Carboxen 1000/Spherocarb, or Tenax-GR/Carboxen
1000/Carbosieve SUI; the collection tube was used with one of each of the following focusing tubes:
Carbotrap B/Carboxen 1000, Tenax-GR/Spherocarb, or Tenax-GR/Carbosieve SHI. In these
experiments, the collection tube was held at 40 °C and desorbed at 300, 325, or 350 *C. The heat
and cool mode times were varied from 1 to 3 min and 0 to 2 min, respectively, to determine the
optimum operating conditions for retention of Cj compounds. The focusing tube was then
desorbed for 3 min at 300 or 350 "C.
To test the unattended, repetitive operation of the ACEM 900, the unit was set to collect
~20 samples of ambient indoor air during a 24-h period. A sampling pump was used to pull
sample across the sorbent tube by using the external sampling module. Sample volumes of 250 cm3
of air were collected, and the tube was flushed with 500 cm3 of helium. The collection tube was
heated to 200 or 300 * C for 3 min and then cooled for 2 min prior to desorbing the focusing tube
for 3 min at 350 *C.
RESULTS AND DISCUSSION
Because the GC-AED is capable of monitoring the responses of individual elements, the
effect of water vapor on the responses of Cl, Br, and C for the 41 VOCs could be easily
investigated. Of the four tube combinations evaluated for recovery of the 41 VOCs as a function
of varying the helium purge volume, two combinations worked satisfactorily. When the
Tenax/silica gel/Arnbersorb/charcoal focusing tube was used in combination with either the
Tenax/Ambersorb/charcoal or the Carbotrap C/Carbotrap B/Carboxen 1000 sorbent tubes, the
large amount of water left on the sorbent tube at low-helium purge volumes resulted in decreased
responses for lighter compounds eluting simultaneously with the broad water peak. The responses
of compounds which eluted after the water had eluted were not affected. Examples of this are
presented in Figures la and Ib. When the Carbotrap B/Carboxen 1000 focusing tube was used in
combination with either the Tenax/Carboxen 1000. or Tenax/Ambersorb/charcpal sorbent tube,
-------
substantially less water was retained, and the response of the lighter compounds was not suppressed
at lower helium purge volumes. This is illustrated in Figure Ic.
The results of the linearity tests for both the 41-compound mixture and the polar mixture
showed good linearity for most compounds for up to 1 L of sample collected. Some compounds,
such as benzyl chloride, m-, p-, and o-dichlorobenzene, cbJorobenzene, and 1,1,2,2-
tetrachloroethane collected on the Tenax/Ambersorb/charcoal tube and the polar compounds
collected on the Carbotrap C/Carbotrap B/Carboxen 1000 tube, were observed to be linear for up
to 2 L of sample collected. Bromomethane became nonlinear when more than 500 cm3 of sample
was collected on the Tenax/Ambersorb/charcoal tube, possibly because at higher sample volumes
the bromomethane travels into a sorbent layer from which it is not easily desorbed. The linearity
results for two representative compounds are presented in Figure 2.
For retaining the Q compounds, the optimum operating parameters were determined to be
heating the collection tube 2 rnin and cooling 0 min. When the tube was heated and cooled for
3 min and 2 min, respectively, as for the 41-compound mixture, the ethene and acetylene were not
retained. To obtain better separation of the light compounds for the determination of recoveries,
the GC oven was programmed as follows: -50 *C for 2 min, 8 °C/min to 150 *C, 150 'C for
3 min. The best tube combinations for recovering the C-/, compounds were Tenax-GR/Carboxen
1000/Carbosieve S-III or Tenax-GR/Carboxen 1000/Spherocarb sorbent tubes coupled with a
Tenax-GR/Carbosieve S-H1 focusing tube. With these tube combinations, recoveries were
estimated to be ~ 100% for ethane, -70% for ethene, and -30% for acetylene when a 500-cm3
helium purge volume was used. Other tube combinations tested retained less, if any, of the
ethylene and acetylene when purged with 500 cm3 of helium. A sorbent tube combination that will
retain 100% of acetylene and ethene has not yet been identified.
The ACEM 900 was easily programmed for unattended, continuous operation, and the
system ran without fail for the two 24-h experiments. Figure 3 is a plot of the concentration of
dichloromethane observed in the laboratory air vs. time of day for the experiment using the
Tenax/Ambersorb/charcoal sorbent tube and the Tenax/silica gel/Ambersorb/charcoal focusing
tube combination.
CONCLUSIONS
The Dynatherm ACEM 900 preconcentrator offers several attractive features. It is reliable
and easy to operate. The instrument requires no cryogenic liquids and may be operated
unattended, and the helium purge option allows water to be removed from the sample without
removing polar VOCs. These features contribute to the ACEM 900's promise as a preconcentrator
for use in an autoGC network for monitoring polar and nonpolar VOCs. However, the instrument
must be further evaluated to compare the results of these experiments with those of a cryogenic
preconcentrator and to challenge the system with the very low (low-part-per-billion-by-volume)
levels of VOCs found in ambient air.
REFERENCES
1. W.A, McClenny, G.F. Evans, KJX Oliver et al., "Status of VOC methods development to meet
monitoring requirements for the Clean Air Act Amendments of 1990," in Proceedings of the 1991
U.S. EPA/A&WMA International Symposium on Measurement of Toxic and Related Air
Pollutants." V1P-21, Air & Waste Management Association, Pittsburgh, 1991, pp 367-374.
2. W.T. Winberry, Jr., N.T. Murphy and R.M. Riggin, Compendium of Methods for the
Determination of Toxic Organic Compounds in Ambient Air. EPA-600-4-84-041, U.S.
Environmental Protection Agency, Research Triangle Park, 1988.
3. K.D. Oliver and J.D. Pleil, Automated Cryogenic Sampling and Gas Chromatographic Analysis
of Ambient Vapor-Phase Organic Compounds: Procedures and Comparison Tests. TN-4120-85-02,
EPA contract 68-02-4035, Northrop Services, Inc., Research Triangle Park, NC, 1985.
-------
A 2000
1600
g 1200
o
| 800
400
B
0
2000
1600
a
g 1200
o
u
R
o
R
O»
Vinyl Chloride
800
400
0
2000
1600
1200
800
400
Benzyl Chloride
Vinyl Chloride
0 200 400 600 800
Purge Volume (cm3)
1000
Figure 1. Effect of purge volume on Cl response with different sorbent tube - focusing tube
combinations: Tenax/Ambersorb/charcoal - Tenax/silica gel/Ambersorb/charcoal (A)
and (B), and Tenax/Carboxen 1000 - Carbotrap B/Carboxen 1000 (C).
-------
B
£
£
40xl(P
20 x IIP
10x103
I
1,2-Dibromoethane
SOO-cm3 Purge
40xl(P
30xl(P
20xl(P
10x103
400 800 1200 1600 2000
Sample Volume (cm3)
Methyl MethacryUte
500-on3 Purge
400 800 1200 1600
Sample Volume (cm3)
2000
Figure 2. Linearity test with (A) Tenax/Ambersorb/charcoal - Tenax/silica gel/Ambersorb/
charcoal tubes and (B) Carbotrap C/Carbotrap B/Carboxen 1000 - Tenax/silica
gel/Ambersorb/charcoal tubes.
5.0
4.0
D.
0,
£ 3.0
I
o
U
2.0
1,0
0.0
Dichloromethane
2:00 pm
July 2
12:00 am
Time
Release of
Dichloromethane |
in Nearby
Hood
July 3
2:00 pm
Figure 3. Diurnal variation in concentration with the ACEM 900 and the Tenax/Ambersorb/
charcoal sorbent tube — Tenax/silica gel/Ambersorb/charcoal focusing tube
combination.
-------
TECHNICAL REPORT DATA
1. REPORT NO,
EPA/600/A-92/251
3.
PB93-121G77
4. TITLE AND SUBTrTLE
Evaluation of a Sorbent-Based Preconcentrated for
Analysis of VOCs in Air Using Gas Chromatography-
Atomic Emission Detection
5.REPORT DATE
April 15, 1992
6.PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
K. D. Oliver, 1. H. Daughtrey, Jr. and W. A.
McClenny
8.PERFORMING ORGANIZATION REPORT
NO.
9 PERFORMING ORGANIZATION NAME AND ADDRESS
ManTech Environmental Services
Research Triangle Park, NC 27711
10.PROGRAM ELEMENT NO.
A101C-C18-01
11. CONTRACT/GRANT NO,
68-DO-0106
12. SPONSORING AGENCY NAME AND ADDRESS
Atmospheric Research and Exposure Assessment Lab
RTF
Office of Research and Development
U.S. Environmental Protection Agency
Research" "Triangle Park, NC 27711
13 .TYPE OF REPORT AND PERIOD COVERED
Proceedings - 4/91 - 4/92
14. SPONSORING AGENCY CODE
EPA/600/09
15. SUPPLEMENTARY NOTES
f3 .
This- work establishes design parameters for the use of carbon-based multisorbent
tubes used as preconcentrators of volatile organic compounds when sampling ambient
air. The multisorbents were tested in a two stage configuration with a primary
tube for collection and a secondary tube for focusing of the VOCs collected in the
primary tube. A system consisting of a Dynatherm ACEM 900 preconcentrator,
Hewlett-Packard 5890 gas chromatograph and 5921A atomic emission detector was used
to establish VOC trapping and release efficiency and the optimum purge gas volume
to remove water from the primary tube. Results will be used to design an automated
gas chromatograph to meet the VOC monitoring requirements of Title I and Title III
of the CAAA of 1990.
16. ABSTRACT
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.IDENTIFIERS/ OPEN ENDED
TERMS
c.COSATI
18. DISTRIBUTION STATEMENT
RELEASE TO PUBLIC
19. SECURITY CLASS (This Report)
UNCLASSIFIED
21.NO. OF PAGES
7
20. SECURITY CLASS (This Page)
UNCLASSIFIED
22. PRICE
------- |