United States Industrial Environmental Research EPA-600/7-79-095
Environmental Protection Laboratory April 1979
Agency Research Triangle Park NC 2771 1
vvEPA Gas Sample Storage
Interagency
Energy/Environment
R&D Program Report
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditiona! grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
8. “Special” Reports
9. Miscellaneous Reports
This report has been assigned to the INTERAGENCY ENERGY-ENVIRONMENT
RESEARCH AND DEVELOPMENT series. Reports n this series result from the
effort funded under the 17-agency Federal Energy.Environment Research and
Development Program. These studies relate to EPAs mission to protect the public
health and welfare from adverse effects of pollutants associated with energy sys-
tems. The goal of the Program is to assure the rapid development of domestic
energy supplies in an environmentally-compatib !e manner by providing the nec-
essary environmental data and control technology. Investigations include analy-
ses of the transport of energy-related pollutants and their health and ecological
effects; assessments of, and development of, control technologies for energy
systems; and integrated assessments of a wide’range of energy-related environ-
mental issues.
EPA REVIEW NOTICE
This report has been reviewed by the participating Federal Agencies, and approved
for publication. Approval does not signify that the contents necessarily reflect
the views and policies of the Government, nor does mention of trade names or
commercial products constitute endorsement or recommendation for use.
This document is available to the public through the National Technical Informa-
tion Service. Springfield, Virginia 22161.
-------
EPA-600/7-79-095
April 1979
Gas Sample Storage
by
K.E. Thrun, J.C. Harris, and K. Beltis
Arthur D. Little, Inc.
Acorn Park
Cambridge, Massachusetts 02140
Contract No. 68-02-2150
T.D. 10702
Program Element No. INE624
EPA Project Officer: Larry D. Johnson
Industrial Environmental Research Laboratory
Office of Energy, Minerals, and Industry
Research Triangle Park, NC 27711
Prepared for
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Research and Development
Washington, DC 20460
-------
ABSTRACT
A laboratory evaluation has been conducted to compare the storage
stability of selected gases covering a range of compound categories, in
three types of containers including glass bulbs and two different poly-
meric sample bags.
The studies indicated that glass bulbs are the best overall choice,
with no significant losses of gases. Reactive and highly polar gases
were preferentially lost from samples stored in the polymeric bags.
Some contaminants were detected in the samples taken from the polymeric
sample bags, presumably the result of outgassing from the materials used
to construct the bags.
11
-------
TABLE OF CONTENTS
P age
I. SUNNARY 1
II. INTRODUCTION 2
III. BACKGROUND 4
IV. EXPERIMENTAL 8
A. Approach 8
B. Sample Preparation and Analysis 11
V. RESULTS AND DISCUSSION 17
A. 5-Layered Bags 17
B. Tedlar Bags 23
C. Glass Bulbs 26
D. Comparison of Analytical Data from the Three Containers 29
E. Nitrogen Blanks in the 5—Layered and Tedlar Bags 29
F. Methanol Preconditioning 36
VI. CONCLUSIONS AND RECOMMENDATIONS 38
VII. REFERENCES 40
1] 1
-------
LIST OF FIGURES
1 Gas Chroniatogram of Gas Mixture 16
2 Ethane Sample Stability 31
3 Propylene Sample Stability 32
4 Methyl Chloride Sample Stability
5 Methanol Sample Stability
6 Ethylene Oxide Sample Stability
LIST OF TABLES
Page
1 Test Gas Mixture 9
2 Gas Sampling Containers 10
3 Analytical Data: 5-Layered Bags (A) 18
4 Analytical Data: 5-Layered Bags (A’) 20
5 Ethylene Oxide Analytical Data: 5—Layered Bags (A and A’) 22
6 Analytical Data: Tedlar Bags (B) 24
7 Analytical Data: Glass Bulbs (C) 27
8 Suniniary of Analytical Data 30
9 Methanol Analytical Data: 5—Layer and Tedlar Bags 37
iv
-------
I. SUMMARY
The suitability of several types of containers for the long—term
storage of gas samples was compared in a laboratory study. The perfor-
mances of glass bulbs and polymeric bags from two vendors were evaluated.
A test gas mixture consisting of ethane, propylene, methyl chloride,
methanol, ethylene oxide and benzene was prepared in order that the
behavior of compounds of differing molecular weights and polarity
could be studied. The test gas mixture was introduced into each of
the three containers and the concentrations of the individual gaseous
compounds were measured as a function of storage time, using gas
chromatography.
The studies indicated that glass bulbs are the best overall choice,
with no significant losses of the gases. Reactive and highly polar
gases were preferentially lost in the polymeric sample bags. Relatively
simple gases were stable in these sample bags. Contaminants were
detected in the samples taken from the polymeric bags, presumably the
result of outgassing from the construction materials of the bags.
These contaminants were found at very low ppb levels, but they could
nevertheless interfere with ultratrace analyses of collected gas samples.
Therefore; the polymeric bags are recommended for the storage of
relatively simple gases at high concentration levels, i.e. ppm or
greater. The calibrated Instrument’s 5—layered bags were superior to
the Dupont Tedlar bags, with fewer interferences and less sample loss.
1
-------
II. JNTRODUCTION
Environmental assessment sampling and analysis studies frequently
necessitate the detection, identification and quantification of very
small concentrations of gaseous organic compounds. Gases are defined,
in this study, to be those species with an approximate boiling point
range similar to those of the Cl—C7 hydrocarbons, i.e., —160°C to 100°C.
The analysis of gaseous organic components requires that suitable
sampling methods be used. Representative samples must be collected, and
their integrity preserved between the sampling and analysis interval.
Several factors contribute to sample loss and contamination when collec-
ting, storing and analyzing gas samples. Major problems that have been
reported in the literature include sample container permeability, reac-
tivity, adsorption, condensation, “memory effect” from reusing con-
tainers, and contaminants in the container.
The objective of this study was to demonstrate what kinds of con-
tainer materials are most suitable for storing organic compounds of
environmental interest, i.e., containers demonstrating the least amount
of container/compound interactions.
Stability was demonstrated by a laboratory evaluation of gas sample
storage as a function of time and container composition, keeping the
other variables as constant as possible. Gas storage stability in two
types of polymeric bags and glass bulbs were compared. The containers
chosen for the study are described in Section IV. The Dupont Tedlar and
Calibrated Instruments 5—layered bags were chosen as being constructed
of materials representative of the types of polymeric materials cur-
rently (1978) in use, and showing relatively high stablity for gas sample
storage. Ace Thred glass containers, which are frequently used for col-
lecting gas samples, were also evaluated.
2
-------
Studies have been completed for six compounds——ethane, propylene,
methyl chloride, methanol, ethylene oxide and benzene——in the three
containers. This mixture was chosen to cover a range of compound cate-
gories. The preparation of the mixture is described in Section IV.
3
-------
III. BACKGROUND
Literature Review
A number of papers have been published discussing methods for sam-
pling and analysis of gases. These have been reviewed to identify im-
portant factors contributing to sample loss and contamination when col-
lecting, storing and analyzing gas samples. The variables identified
were container composition, size and preconditioning, storage time,
temperature, relative humidity of sample air, lighting conditions, sam—
pie composition and concentration, and methods of sample collection.
Baker( 1 ) compared storage of various gases in steel cylinders,
plastic bags and glass flasks. Compression storage in steel cylinders
was adequate for hydrocarbons but resulted in high losses of sulfur
dioxide and nitrogen dioxide. Baker evaluated Mylar (polyester film),
aluminized Mylar Saran wrap type 12 (polyvinyidene chloride film),
Saran wrap drum liner, Scotchpak 2OA5 (polyester film) and aluminized
Scotchpak 20A5 as storage containers for gases at atmospheric pressure.
Permeability of gases and vapors through polymeric films is a function
of temperature, polymeric sidechains, plasticizers and H 2 O.( 2 )
Moisture was a major factor in the polymeric bags; mineral oxides
disappeared at high relative humidity. Hydrocarbon loss was controlled
by diffusional changes alone and showed only minor changes. Acetone
and benzene showed little change, and 2—pentene showed minor losses.
The two—layer bags (aluminized Scotchpak and Mylar) reduced the
moisture problem. A sulfur dioxide, nitrogen dioxide, and 2—pentene
mixture stored in stainless steel containers at atmospheric pressure
showed no significant loss. Comparable storage in glass flasks showed
high loss of sulfur and nitrogen dioxides.
Cleinens( 3 ) further found Mylar to show no systematic changes in
paraffinic hydrocarbon concentration over a period of ten days. A
major portion of the aromatic hydrocarbons were lost after a few days.
4
-------
Some of the aromatic hydrocarbons later reappeared in the vapor space
of the container.
Conner used Mylar and Teflon containers to study the storage of
N0 2 ,S0 2 , 03 and hydrocarbons. Teflon bags were found to be inferior to
Mylar for containing SO 2 , NO 2 , and 03, and offered no noticeable
advantage for storing hydrocarbons. Preconditioning Mylar bags with
SO 2 and 03 improved their storage, but Mylar could not be conditioned
to NO 2 and hydrocarbons.
Glass containers are known for major losses of sulfur compounds,
apparently due primarily to surface adsorption. Wright’ investigated
Teflon and Tedlar bags for sulfur compound storage. Initial losses
in Tedlar and Teflon bags were found to be similar. However, after 160
hours the Tedlar bags retained 75% of the sulfur compounds, while
Teflon retained only 25%.
Schuetzle used hydrocarbon and odor measurements to evaluate
sampling procedures. Polyethylene, Saran and Mylar bags were found to
be unsuitable for storing organic emissions, whereas standards and
samples stored in Tedlar bags have shown less than 10% average hydrocar-
bon loss over a period of several days. It was found that most errors
affecting odor and hydrocarbon concentration values of gas samples in
Tedlar bags occurred when collecting warm (>30°C), highly concentrated,
or very polar organic gases. Errors were reduced by 1). minimizing
adsorptive effects by increasing bag size and conditioning in evacuated
containers, 2). pressurizing with a non—odorous gas, and prediluting the
sample, 3). dynamically diluting warm (>30°C) stack gases and 4). removing
container contaminants by flushing with dry air at 80°C for approximately
2 hours.
Levime compared vinyl chloride monomer (VCM) gas storage in
Teflon and aluminized Scotchpak 3—layered bags (Calibrated Instruments).
The Teflon bag showed a loss of VCM in the range of 20% per day. There
5
-------
was no detectable loss of V M in the aluminized Scotchpak, 3—layered bag
for a period of one week.
Polasek(8) evaluated a bag sequential sampling technique for ambient
air analysis. To determine the influence of bag materials on samples,
calibration gas containing carbon monoxide, non—methane hydrocarbons and
methane was placed in the bags and then analyzed at 0, 24, 48 and 100
hours. The results indicated that polyvinyl chloride bags were satisfac-
tory for sample storage (up to about 15 hours), whereas Tedlar bags were,
in general, unsatisfactory for sample storage. Polasek stated that the
5—layered bags were excellent for long—term storage of carbon monoxide,
but completely unacceptable for non—methane hydrocarbons. Aluminized
polyester bags, consisting of a layer of polyester on both sides of an
aluminum layer, were excellent for long—term storage of both carbon
monoxide and non—methane hydrocarbons.
Vanllaam( 9 ) investigated possible interferences from low molecular
weight additives in polyvinyl fluoride (PVF: “Tedlar”) used for the
construction of smog chambers. Organic compounds released from Tedlar
interfered with (total) hydrocarbon concentration, NOx response and
hydrocarbon/NO ratio measurements. It was also observed that ozone
and peroxyacetyl nitrate were formed in the bags. The interference
decreased with increasing surface to volume ratio.
Dimitriades(1 0 ) used Tedlar (2 sill thick) bags to investigate the
development of procedures for routine analysis of automotive fuels and
the hydrocarbon of exhaust emissions. Tedlar was found to lead to
losses of individual hydrocarbons on or through bag walls, of 5 to 6%.
Acetone was observed to be lost by permeation through the Tedlar walls.
Acetaldehyde and butyraldehyde levels did not change significantly,
whereas benzaldehyde disappeared completely within 0.5 hr. Epoxides
of ethylene and 1—butene were somewhat unstable. It was also observed
that adsorption and accumulation of adsorbable material occured on
the gas chromatography (GC) sample loop, used for sample
introduction. Errors in analyzing gas mixtures were reduced by:
6
-------
1) increasing the container size, to minimize water condensation
and slow down reaction among sample components;
2) keeping the Tedlar bags in a dark cloth bag, to reduce light
induced reactions;
3) keeping sampling lines and proportional sampling device at 80°C,
to prevent water condensation and subsequent loss of water
soluble components; and
4) heating the loop to 110°C, to minimize sample adsorption and
build—up in the GC sampling loop.
Vinyl chloride monomer (VCM) gas storage in 5—layered bags was
studied by the Independent Research Division for Gases of the Linde
Corporation, for Calibrated ins truments(h1 ). Minor losses of analyte
were found after periods of 8 days and 8 weeks. Other tests performed
were: 1) bursting strength, 2) aging, 3) change of gas composition,
4) total gas losses at changing temperatures, 5) analytical measure-
ments of dilution losses of individual components, and 6) determination
of losses of gas components through permeation. The overall performance
of the 5—layered bag was found to be quite good.
Based upon the information available in 1978, the Tedlar bags,
5—layered bags and glass bulbs appeared to be suitable containers to
consider using for gas sampling. The literature review indicated that
contamination, absorption and permeation problems in the 5-layered bags
were somewhat less than in the Tedlar bags. The polymeric bags appeared
to offer significant advantages compared to glass bulbs in terms of
handling in the field and shipping gas samples to the laboratory for
analysis.
7
-------
IV. EXPERIMENTAL
A. Approach
This study of gas sampling and storage looked at sample storage
stability as a function of time and container composition. Light was
excluded from all of the containers, as much as possible, to eliminate
potential light reactions. The approach taken involved adding a known
concentration of a test gas mixture into a container and measuring the
gas concentration levels as a function of time. Replicate sample data
were used to calculate the mean measured value, standard deviation and
coefficient of variation, at predetermined times.
Experiments were conducted with a test gas mixture representative
of several categories of compounds. The components of the mixture are
described in Table 1.
The specific components of the mixture were chosen to represent a
range of chemical reactivities, and to be of environmental interest.
Where possible, compounds with low toxicity to the analyst were chosen
from the compound categories. The three types of containers studied,
described in Table 2, were the 5—layered bag, Tedlar bag and Ace—Thred
glass collecting bulbs. These containers are generally representative
of the various container compositions currently being used for storing
samples.
The literature review (Section III) indicated that surface adsorption
may be a problem with polar organic gases. Suggestions were made that
increasing the container size and preconditioning may to some degree
eliminate the problem. Some experiments were run to evaluate whether
or not preconditioning the polymeric bags with methanol would decrease
methanol concentration losses due to absorption into the walls of the
bags.
8
-------
TABLE 1
Test Gas Mixture
MATE (12) Retention Time
(c)
Compound i “ (a) (b) BP on Porapak Q
Category Compounds Composition Air\a/ Water MW °C Hinutes
Aliphatic Ethane C 2 H 6 6.1E6 30 —89 35
Hydrocarbon
Aliphatic Propylene C 3 H 6 8.6E6 1.3E8 42 47 18.5
Hydrocarbon
Alkyl Halide Methyl Chloride CH 3 C1 2.lE5 3.2E6 50 —24 22.5
Alcohols Methanol CH 3 OH 2.6E5 3.9E6 32 64 27
Epoxides Ethylene Oxide C 2 H O 44 11 50
Benzenes Benzene C 6 H 5 3.0E3 4.5E4 78 80 30 5 (d)
(a) Health effects basis (pg/rn 3 air and pg/L, water).
(b) Monoisotopic, lightest isotope of each atom.
Cc) 2m x 2mm I.D. Porapak Q (waters), 100/120 mesh, glass column, 65°C isothermal, He= 20 mLs min
(d) 2m x 2mm I.D. Porapak Q (waters), 100/120 mesh, glass column., 130°C isothermal, He= 20 mLs min .
-------
TABLE 2
Gas Sampling Containers
Calibrated
Instruments
From Innermost Layer
High Liensity Poly Ethylene —
75 i.im
Polyaxnide — 40 inn
Aluminum Foil — 12 inn
Polyvinylidene Chloride —
4 im
Polyethylene Terephthalate —
12 inn
Polyvinyl Fluoride —
51 tm
Fitted with
self—sealing
septum, on/off
valve hose/bib
combination,
shape snout—type
fitted with
on/off valve
hose/bib
combination.
fitted with
septum sampling
port and
Ace—Thred valve.
Container
Vendor
Material Composition
Volume
Surface
Area
Comments
C
5—Layered Bags
Tedlar Bags
Glass Collecting
Bulbs
Dupont
Ace—Thred
2L
4L
2L
1100 cm 2
1800 cm 2
950 cm 2
Glass
-------
B. Sample Preparation and a1ysis
1. Test Gas Mixture
The sample mixture purchased from Matheson Gas Products, consis-
ted of ethane, propylene, methyl chloride, benzene and methyl
alcohol in nitrogen at 10,000 ppm (v/v) each. Ethylene oxide,
also purchased from Matheson Gas Products, was available in a
separate gas cylinder and was added separately. The compo-
nents in the mixture were chosen to cover a range of compound
categories and boiling points, and were of generally low toxi-
city. The test gas mixture was added into the three types of
sample containers to a concentration level of 100 ppm. All the
components in the mixture, except benzene, could be analyzed in
a single gas chromatography (GC) experiment, as described in
Section B.7. Benzene was analyzed separately.
It was originally intended to use air as the diluting gas in
this study. Blanks were run on the 5—layered bag samples con-
taining 2 liters of air (Matheson air, zero gas). The sample
mixture was spiked into the 5—layered bags containing 2 liters
of air.
During the determination of gas chromatographic conditions, an
unacceptable baseline and several extraneous gas chromatographic
peaks, interfering with the sample mixture peaks, were observed.
These interfering peaks gradually increased with repetitive
sample and blank injections. Subsequent experiz nts indicated
that multiple injections of air on the Porapak Q column ap-
peared to cause the interfering peaks. A Durapak GC column
was tried as a substitute for the Porapak Q; however, methanol
was not eluted from the column.
When nitrogen gas (tapped from a liquid nitrogen cylinder) was
substituted for the air, no interfering peaks were observed on
11
-------
the Porapak Q column. Nitrogen gas, therefore, was chosen as
the diluting gas for the study. Nitrogen gas produced an
acceptable baseline on the gas chromátogram and no interfering
GC peaks.
Losses in analyte concentration values are limited to container!
compound Interactions, when nitrogen is used as the diluting
gas. Analyte and container reactions with air could introduce
additional problems.
2. 5—Layered Bags
Data were obtained by analyzing five replicate samples of
two sets of 5—layered bags (sample series A and A’) at
intervals up to 14 days. The 5—layered bags were purged with
nitrogen at least twice. Two liters of nitrogen were then
metered into the containers. The gas sample mixture was
added with a gas tight syringe to give an approximate
concentration of 100 ppm. Five sample mixtures were each
analyzed immediately after preparation (at 0 hours) and stored
in a dark, dry cabinet. The 5—layered bag samples were sub-
sequently analyzed at predetermined times. Sample Introduction
was made via a gas—sampling loop, directly from the bag.
Because benzene could not be analyzed under the CC conditions
used for the other five sample components (IV.B.7), the benzene
concentration was not determined at each of the predetermined
storage time intervals. Benzene stability in the sample mixture
was evaluated based on the concentration measured after 53 days
of sample storage.
3. Tedlar Bags
Data were obtained by analyzing five replicate samples in the
Tedlar bags (sample series B), at intervals up to 14 days.
Complete sample loss was observed with the 2—liter Tedlar bags
12
-------
originally obtained, and it was found that they leaked. Four—
liter Tedlar bags obtained from a different supplier were
satisfactory. The bags were wrapped with aluminum foil, to
exclude light as much as possible, and purged with nitrogen
at least twice. Two liters of nitrogen were then metered into
the containers. The gas sample mixture was added to the Tedlar
bags to an approximate concentration of 100 ppm, using a gas—
tight syringe. Five sample mixture replicates were each
analyzed immediately (at 0 hour) and then stored in a dark, dry
cabinet. The Tedlar bag samples were subsequently analyzed at
predetermined times. Sample introduction into the gas chroma-
tography column was made via a gas sampling loop.
Benzene data were obtained after the basic time study was
completed, i.e., after 51 days of storage (please refer to
Section IV—B—7).
4. Glass Bulbs
Data were obtained by analyzing four replicates of the
Ace—Thred glass bulb samples at intervals up to 14 days.
The glass bulbs were wrapped with aluminum foil to
exclude light, and purged with a stream of nitrogen. Two
liters of nitrogen were metered into the containers. The
gas sample mixture was added to an approximate concentra-
tion of 100 ppm, using a gas—tight syringe. Sample
withdrawal from the glass bulb was made with a gas—tight
syringe, and sample introduction onto the gas chromato—
graph column was done via a gas—sampling loop. Four
sample imixture replicates were each analyzed immediately
after spiking (at 0 hour) and stored In a dark, dry
cabinet. The glass bulb samples were subsequently
analyzed at predetermined times.
13
-------
Benzene data were obtained after the basic time study was
completed, i.e., after 15 days of storage (please refer to
Section IV—B— 7).
5. Nitrogen Blanks in the 5—Layered and Tedlar Bags
The 5—layered and aluminum foil covered Tedlar bags were
purged with nitrogen at least twice. Two liters of nitrogen
were then metered into the containers. Samples were intro-
duced onto the gas chromatograph column via a gas sampling
loop. The detector sensitivity was increased ten fold, com-
pared to that used for samples.
6. Analysis of Methanol In 5—Layered and Tedlar Bags
Preconditioned with Methanol
The 5-layered and aluminum foil covered Tedlar bags were
filled with nitrogen. Methanol was subsequently added to the
Tedlar and 5—layered bags and equilibrated. The containers
were then emptied. This procedure was followed three times,
in order to precondition the inner walls of the two types of
containers with methanol. After preconditioning, two liters
of nitrogen were metered into duplicate Tedlar and 5—layered
bags. Methanol was added to the containers to a concentra-
tion of 100 ppm. Data were obtained at intervals up to seven
days.
7. Analysis
Gas Chroinatographic Conditions
Sample introduction on the Varian 2700 CC was made via a
1.0 mL gas sampling loop and valve. Gas chromatographic
(GC) conditions were investigated using a 2m x 2mm I.D.,
Porapack Q (Waters), 100/200 mesh, glass column. The
14
-------
optimal conditions for chromatography of the gas sample
mixture were determined to be:
Column Temperature: 65°C, isothermal
Helium flow: 20 mLs/min
Injection temperature: 200°C
Detector temperature: 240°C
Under these conditions, the retention times for each of
the model compounds were: ethane — 3.5 minutes, propylene —
18.5 minutes, methyl chloride — 22.5 minutes, methanol —
27 minutes, and ethylene oxide — 50 minutes. A gas
chromatogram is presented in Figure 1. Benzene, however,
had an unacceptably long retention time under these
conditions. Temperature programming was investigated,
but this resulted in unacceptable baseline drift. Benzene
was quantified in separate experiments at an isothermal
temperature of 130°C. At this temperature, the retention
time for benzene was 30.5 minutes.
Fresh calibration standards were prepared daily at concen-
tration levels of 200 ppm (v/v), 100 ppm (v/v) and 50 ppm
(v/v). Standard curves were drawn for each of the model
compounds, and used to calculate the concentration values.
It was observed that the retention times of the sample GC
peaks gradually increased with repetitive sample injections.
The column temperature therefore had to be increased to 85°C
isothermal, so that the analysis time would remain rea onah1e.
i.e., less than 1 hour per sample. There seemed to be some
serious stability problems associated with the use of Porapak Q.
15
-------
Sample 100 ppm Test Gas Mixture in 5-Layered Bag
Column: 2mx 2mm l.D. Porapak Q (Waters), 100/120 Mesh, Glass Column
Column Temperature 65°C (isothermal)
Helium Flow: 20 mIs min 1
Inj. Temperature 200°C
Det. Temperature 240°C
m
Co
0 ’
0
V
CO
5 10 15 2530 35 40 4550
Time. Minutes
FIGURE 1 GAS CHROMATOGRAM OF GAS MIXTURE
-------
V. RESULTS AND DISCUSSION
A. 5—Layered Bags
] ta were obtained for two sets of bags spiked with the test
gas mixture, at intervals up to 14 days of storage. The thta
are presented in Tables 3, 4, and 5. The mean,(X), standard
deviation (a), and coefficient of variation (CV) were calcu—
lated for each set of five replicates.
Five of the six model compounds are apparently stable for up to
14 days of storage. Ethane, propylene, methyl chloride, ethylene
oxide and benzene show no apparent trend towards loss or gain.
The ethylene oxide data are presented in Table 5. The repli-
cate data at each time period have high coefficients of varia-
tion. The syringe used for adding the ethylene oxide may have
been plugged, causing this high variability. Therefore, the
ethylene oxide data are presented for each container. The coef—
ficient of variation for ethylene oxide concentration in any
given container is low, and there do not appear to be any signi-
ficant losses or gains. Ethylene oxide appears to be stable in
the Calibrated Instrument 5—layered bags.
The methanol data in both sets of bags have high coefficients
of variation, and apparent losses over 14 days. The overall
methanol concentration loss in the replicate bags is 50%. The
loss appears to occur within the first 24 hours. The first set
of 5—layered bags (A) has an 86% loss at 3 days. This loss is
not replicated and may be due to experimental error. The
methanol concentration losses may be due to permeation through
or absorption on the walls of the 5—layered bags.
Benzene was quantified in duplicate 5—layered bags, after 53
days of storage. The calculated mean concentration at 53 days
17
-------
TABLE 3
Gas Sampling Study Analytical Data
S p1e Container : Calibrated Instrument’s
5-Layered Bags (A)
COMPOUND CONCENTRATION, ppm* (v/v)
Sample _-
Time Ethane
Methyl
Propylene Chloride
Methanol
1A /0
hrs
110
109
100
110
2A
115
115
115
110
3A
104
130
115
93
4A
109
110
90
83
5A
104
105
95
93
X = 109
0= 5
CV 4%
= 114
0 9
CV = 8.5%
X = 103
a= 12
cv = 11.2%
X = 98
o=12
cv = 12.1%
1A/ 1
day
104
106
100
50
2A/
105
110
115
39
3A/
105
105
100
50
4A/
117
110
96
65
5A/
103
108
100
57
X = 107
a6
CV = 4%
X = 108
a= 2
CV = 2.2%
X = 102
a= 7
CV 7.2%
= 52
a=lO
CV = 18.4%
1A/ 3
days
104
113
103
10
2A/
108
108
103
10
3A/
115
104
103
21
4A/.
109
110
95
10
5A/
104
106
103
21
x= 107
0= 5
CV = 4.5%
x = 108
0= 3
CV = 3.2%
X = 101
0= 3
CV = 3.5%
= 14
0= 6
CV = 41.8%
*pm — Calculated from a standard curve. (Concentration vs. Peak Area)
Mean =
Standard Deviation = a —
Coefficient of Variation — a/x = CV
0,
-------
TABLE 3 (Cont’d)
Gas Sampling Study Analytical Data
Sample Container: Calibrated Instrument ’s 5—Layered Bags (A )
*
(v/v)
COMPOUND CONCENTRATION, ppm
¼0
eT.
Ethane
Propylene
Methyl
Chloride
Methanol
1A/14
Days
118
111
95
36
2A/
118
111
107
36
3A/
ill
104
116
40
4A/
111
107
98
32
5A/
111
107
103
28
X= 114
0= 4
CV = 3.4%
3 = 108
0 3
CV = 2.7%
X= 109
Y= 8
CV = 7.4%
X= 34
0=5
CV = 13.2%
*
ppm = Calculated from a standard curve.
Mean =
Standard Deviation = 0
Coefficient of Variation = a/x CV
(Concentration vs. Peak Area)
-------
TABLE 4
Gas Sampling and Storage Study Analytical Data
( Sampling Container: Repeat Calibrated Instrument’s 5—Layered Bag (A’ )
COMPOUND CONCENTRATION, ppm* (v/v)
Sa
mple
Time
Ethane
Propylene
Methyl
Chloride
Methanol
1A’/O
2A’
3A’
4A’
5A’
hr
1.05
107
117
117
122
109
115
115
11.1.
112
92
98
115
106
106
84
93
93
73
111
x = 114
X = 112
X = 103
X = 91
0= 7
0= 3
0= 9
ci=14
CV = 6.4%
CV = 2.3%
CV = 8.4%
CV = 15%
1A’/
2A’/
3A’/
4A’/
5A’ /
1 day
105
108
113
105
110
X = 108
100
102
105
107
102
X = 103
103
95
90
95
100
X = 97
62
50
32
40
10
X = 39
a= 3
0= 3
0= 5
=2O
CV = 3.2%
CV = 2.7%
CV = 5.2%
= 39%
0
*ppm = Calculated from a standard curve. (Concentration vs. Peak Area)
Mean = X
Standard Deviation = —
Coefficient of Variation = a/X = CV
-------
TABLE 4 (Cont’d)
*
Gas Sampling and Storage Study Analytical Data
Sampling Container: Repeat Calibrated Instrument’s 5—Layered Bag (A’ )
COMPOUND CONCENTRATION, ppm* (v/v)
ppm = Calculated from a standard curve.
Mean =
Standard Deviation = a
Coefficient of Variation = aff = CV
Ni
—
Time
Ethane
Propylene
Methyl
Chloride
Methanol
1A’/8 days
97
102
95
68
2A’
93
96
83
20
3A’
49
100
95
29
4A’
101
103
100
32
5A’
99
95
83
50
X=98
= 3
CV = 3.1%
x=99
0=4
CV = 3.6%
x=9l
a = 8
CV = 8.5%
x=4O
a=l9
CV = 48%
1A’/14 days
95
97
92
57
2A’
95
98
88
42
3A’
95
97
ill
45
4A ’
95
99
92
47
5A’
102
97
92
62
X=96
a = 3
CV = 3.2%
x=98
0= 1
CV = .9%
X=95
0= 9
CV = 9.5%
X=51
c 9
CV = 16.8%
(Concentration vs. Peak Area)
-------
TABLE 5
Gas Sampling Study Analytical Data
Sample Coritairter: Calibrated Instruments 5—layered Bags (A and A’ )
Ethylene Oxide Concentration, ppm .(v/v)
Sample/Time
0 hour 75 58 27 27 27
iday 46 40 25
1.5 days 36
3 days 61 144 27 27 32
7 days 46
14 days 70 51 25 23 32
63 48 26 26 35
13 8 1 2 7
CV 20.2% 16.4% 4.4% 8.9% 20.5%
Sample/Time lÀ’ 2A’ ‘ ‘
0 hour 50
1 day 120 99 96 84 59
8 days 110 98 91 79 42
14 days 121 125 106 91 52
117 107 98 85 51
6 15 8 6 7
CV = 5.1% 14.2% 7.8% 7.1%
-------
was 79 ppm. This does not appear to be a significant loss as
compared to an original 100 ppm spike. Benzene appears to be
stable in the calibrated instrument’s 5—layered bags.
B. Tedlar Bags
Data were obtained for the Tedlar bag samples spiked with the
test gas mixture, at intervals up to 14 days of storage. The
data are presented in Table 6. The mean (X), standard deviation
(c) and coefficient of variation (CV) were calculated for each
set of 5 replicates.
Four of the six model compounds are apparently stable up to
14 days of storage. Ethane, propylene, methyl chloride and
benzene show no apparent trend towards loss or gain within the
calculated CV of the replicate values.
Benzene was quantified in 3 replicate Tedlar bags, after 51 days
of storage. The calculated mean concentration at 51 days was
126 ppm. This does not appear to be significantly higher than
the original 100 ppm spike concentration, within the precision
of the measurement. Benzene, therefore, appears to beaable in
the Dupont Tedlar bags.
The methanol data show apparent losses at 1 day, 3 days, 7 days
and 14 days. The Tedlar bag samples show a mean methanol concen-
tration of 72 ppm at 0 hour, 51 ppm at 1 day, 59 ppm at 3 days,
38 ppm at 7 days and 0 ppm at 14 days. The percentage loss with
time is 29% at 1 day, 28% at 3 days, 48% at 7 days and 100% at
14 days. The coefficients of variation for the replicate Tedlar
bag samples at any given time are smaller than those calculated
for the 5—layered bags. The methanol concentration decreases
may be due to permeation through, or adsorption on, the walls of
the Tedlar bags.
23
-------
TABLE 6
Gas Saii p1 ing Study Analytical Data
Sample Container: Tedlar Bags (B )
COMPOUND CONCENTRATION, ppm* (v/v)
Sample
Time
Ethane
Propylene
Methyl
Chloride
Methanol
Ethylene
Oxide
1B/O
hrs
93
96
100
65
109
2B
93
96
104
65
98
3B
104
96
95
74
80
4B
101
95
95
89
46
5B
95
l( 1
87
65
84
Y=
=
CV =
97
5
5.2%
=97
0= 2
CV = 2.4%
=96
c= 6
CV = 6.6%
(=72
0=11
CV = 14.6%
=6l
0=29
CV = 46.9%
lB/i
day
87
101
87
50
46
2B
89
101
87
50
67
3B
95
96
95
50
59
4B
99
99
95
59
46
5B
99
97
91
47
59
X=
0=
CV =
94
6
5.9%
5=99
0=2
CV = 2.3%
=9l
0=4
CV = 4.4%
=51
0=5
CV = 8.8%
=55
a=9
CV = 16.5%
1B/3
days
107
97
101
60
35
2B
96
104
108
57
48
3B
105
102
105
60
46
4B
107
102
105
57
27
SB
105
99
97
60
50
=
cY
CV=
104
5
4.4%
= 101
0= 3
CV= 2.7%
x= 103
0 4
CV= 4.1%
= 59
0= 4
CV= 2.8%
Y= 41
o= 10
CV 23.9%
*Ppm = Calculated from a standard curve. (Concentration vs. Peak Area)
Mean = X
Standard Deviation = 0 -
Coefficient of Variation = a ix = CV
-------
TABLE 6 (cont’d)
*ppm = Calculated from standard curve.
Mean = X
Standard Deviation = —
Coefficient of Variation = cY/X = CV
S amp le C on ta me r: I ’ ed lar B s (B
COMPOUND CONCENTRATION, ppm* (v/v)
LI
1B/7 days
Ethane
Propylene
Methyl
Chloride
Methanol
Ethylene
Oxide
105
101
103
38
39
2B
103
101
99
40
56
3B
103
100
99
37
52
4B
103
99
107
42
96
5B
105
97
110
35
65
X= 104
°= 1
CV = 1.1%
X= 100
= 2
CV = 2%
X= 104
= 5
CV = 4 .7%
X= 38
a=
CV = 7%
X= 62
= 21
CV = 34.7%
1B/14 days
96
118
94
0
36
2B
98
118
98
0
48
3B
123
121
94
0
48
4B
123
118
94
0
36
5B
123
113
86
0
51
= 112
a= 14
CV = 12.6%
X= 118
0= 3
cv = 2.4%
X= 93
0= 4
CV = 4.7%
= 0
0=0
CV = 0
= 44
0= 7
CV = 16.5%
(Concentration vs. Peak Area)
-------
The ethylene oxide data in the Tedlar bags show high coefficients
of variation and indicate a slight loss in the concentration
level. The data indicate that the high coefficients of variation
for the Tedlar bags, unlike those for the 5—layered bags, cannot
simply be attributed to a high coefficient of variation in the
initial concentrations. The Tedlar bags have a mean ethylene
oxide concentration of 51 ppm at 0 hour, 55 ppm at 1 day, 41 ppm
at 3 days, 62 ppm at 7 days and 44 ppm at 14 days. It should be
noted however, that the percentage loss with time is not significantly
larger than the coefficient of variation for this analysis.
C. Glass Bulbs
Data were obtained for the glass bulb samples spiked with test
gas mixture up to 14 days of storage and are presented in
Table 7. The mean, standard deviation and coefficient of
variation have been calculated for each set of 4 replicates.
All six model compounds are apparently stable for up to 14
days of storage.
Ethane, ethylene oxide, propylene, and methyl chloride show
no changes in concentration that are significant when compared
to the coefficient of variation calculated for replicate
samples.
Methanol shows no significant trend towards loss or gain up
to 14 days, when compared to the calculated coefficients of
variation for the measurements. There is an apparent increase
in the methanol concentration, 39% in fourteen days, and the
coefficients of variation are large, 14.5% to 39.2%. These
observations may be due to the effects of the gas sample
mixture equilibrating and may reflect mixing difficulties
in the glass bulbs.
26
-------
TABLE 7
Gas Samn 1 ing St 4y_Ana ytica1 Data
Sample Container: Ace Thred Glass Bulbs (C)
COMPOUND CONCENTRATION, ppm (v/v)
.
1C/O hr
Ethane
Propylene
Methyl
Cloride
Methanol
Ethylene
Oxide
96
89
83
76
106
2C
101
99
92
67
95
3C
108
108
108
84
98
4C
105
113
120
60
103
x = 101
6
CV = 6.1%
= 102
0= 11
CV = 10.3%
x = 101
ci= 17
CV = 16.3%
= 72
a=1O
CV = 14.5%
= 99
0= 3
CV = 3.4%
ic/i
day
100
96
103
48
107
2C
95
96
98
76
93
3C
111
116
130
56
103
4C
111
121
135
100
102
x = 104
a 8
C v = 7.7%
= 107
13
CV = 12.2%
= 116
19
cv = 16.1%
= 76
=23
cv = 33%
101
= 6
Cv = 5.8%
1C/4
days
96
106
92
30
2C
96
107
99
69
3C
108
124
iii
85
99
3C
105
124
115
88
99
= 101
a= 6
CV = 6.1%
= 115
a=jj
CV = 8.7%
X = 104
0= 11
CV = 10.2%
= 68
0=27
CV = 39.2%
X = 101
0= 9
CV = 8.8%
*ppm — Calculated from a standard curve.
Mean =
Standard Deviation = 0
Coefficient of Variation = = CV
(Concentration vs. Peak Area.)
-------
TABLE 7 (Cont’d)
Gas Sampling and Storage Study Analytical Data
Sample Container: Ace Thred Glass Bulbs (C )
COMPOUND CONCENTRATION, ppm* (vlv)
Sample
Time
Ethane
Propylene
Methyl
Chloride
Methanol
Ethyle
Oxide
ne
1C/7 days
83
96
94
72
115
2C
100
92
94
88
93
3C
109
109
121
108
105
4C
109
112
121
112
96
x= 100
a= 12
CV = 12.2%
x= 102
0= 16
CV = 9.5%
x= 108
a= 16
CV = 14.5%
x= 95
0=19
CV = 19.5%
x=
0=
CV =
102
10
9.6%
lC/l4 days
95
105
100
116
87
2C
105
109
105
106
94
3C
100
136
120
61
81
4C
138
134
125
116
93
= 110
o = 19
CV = 17.3%
= 121
o = 16.3
CV = 13.4%
= 113
a = 11.9
CV = 10.6%
= 100
a = 26.2
CV = 26.3%
=
a =
CV =
89
6.0
6.8%
*
ppm = Calculated from standard curve. (Concentration vs. Peak Area)
Mean =
Standard Deviation = a
Coefficient of Variation = a/ CV
-------
Benzene was quantified in duplicate Ace Thred glass sampling
bulbs. The calculated mean concentration at 15 days was 143
ppm. This is probably not significantly higher than the
original 100 ppm spike since the analytical results for the
other five model compounds were also greater than 100 ppm.
Benzene therefore appears to be stable in the glass bulbs,
within the precision of the measurement.
D. Comparison of Analytical Data for the Three Types of Containers
The data collected for the three containers is summarized in
Table 8. Figures 2 to 6 compare the individual compound concen-
tration values for the three types of containers.
Ethane (Figure 2), propylene (Figure 3), and methyl chloride
(Figure 4) were apparently stable in all three containers.
Benzene was also apparently stable, however the data has not
been plotted, because there were too few data points. Substan-
tial methanol (Figure 5) losses were apparent in the two types
of polymeric bags. The glass bulb samples did not show similar
losses. In the glass containers the methanol concentration
showed apparent increases which may not be significant compared
to the precision of the measurement. If real, the increases may
be due to imperfect mixing in the glass bulb sampled. Ethylene
oxide (Figure 6) was apparently stable in the glass bulbs.
Ethylene oxide in the polymeric bags did not appear to be as
stable, and high coefficients of variation were observed.
E. Nitrogen Blanks in the 5—Layered and Tedlar Bags
The contents of 5—layered and Tedlar bags filled with nitrogen
only were analyzed over a period of 14 days. Both containers
had interfering gas chromatography peaks in the —161°C to 72°C
boiling point range, on the Porapak Q CC column. The heights
29
-------
Table
Summary of Analy±ical Data
Mean Concerrtration Data, pp
(b) ( ) Cd) Ethylen
Ethane Propylene Methyl UrilorideC Methanol Cxide(e
i -.i i_ __ JA iII II I
0 109 11)4 97 101 1lLi 112 97102 103 103 96 101 98 91 72 72 61 99
1 107 108 94 104 108 103 99 107 102 97 91 116 52 51 76 55 101
3 107 104 108 101 101 103 14 59
101 115 104 68 101
7 104 100 100 100 102 104 108 38 95 62 102
8 98 99 91 40
14 114 96 112 110 108 98 118 121 109 95 93 113 51 0 100 89
A- 5-Layered Bags
A’- Duplicate - 5-Layered Bags
B - Tedlar Bags
C - Glass Bulbs
(a) Coefficient of variation averaged 6.17
(b) Coefficient of variation averaged 5.1%
(c) Coefficient of variation averaged 8.5%
(d) Coefficient of variation averaged 20%
Ce) Coefficient of variation averaged 17%
-------
A
B
C
120
100
E 80
C
0
Co
J60
40
20
0
A
0 2 4 6 8 10 12
Time, Days
14
FIGURE 2 ETHANE SAMPLE STABILITY
31
-------
120
C
B
A
A’
100
180
C
0
4-,
1
so
4°
20
0
0 2 4 6 8 10 12
Time, Days
14
FIGURE 3 PROPYLENE SAMPLE STABILITY
32
-------
120
Time, Days
FIGURE 4 METHYL CHLORIDE SAMPLE STABILITY
C
A
A’
B
100
80
E
0.
a.
C
0
60
8
C
(3
40
20
0
0 2 4 6 8 10 12 14
33
-------
0 2 4 6 8 10 12 14
Time, Days
FIGURE 5 METHANOL SAMPLE STABILITY
C
A’
A
B
E
a
a
C
0
4-,
4-
C
U
C
(3
120
100
80
60
40
20
0
34
-------
120
100
80
E
a
a
C
::
20
C
B
0
0 2 4 6 8 10 12
Time, Days
14
FIGURE 6 ETHYLENE OXIDE SAMPLE STABILITY
35
-------
of the gas chromatography peaks were compared at 1 day and 7
days. The 5—layered bags show no apparent trend towards loss
or gain of those peaks within that time period. The Tedlar
bags did show a substantial increase at 1 day and 7 days.
These interferences were not a problem in the study presented
in this report, which was conducted at the ppm level. The
interferences would be of concern at the ppb level.
F. Methanol Preconditioning
The 5-layered and Tedlar bags were checked to see if precondi—
tioning with methanol would decrease the losses in methanol
concentration (Table 9). No benefits were observed, and losses
were still substantial after storing for 3 days and 7 days.
The overall loss was 50%. This finding offers some preliminary
evidence that methanol losses may not be due entirely to
absorption on the walls of the polymeric bags.
36
-------
TABLE .9
Gas Sampling Study Analytical Data
Methanol Concentration, ppm (v/v)
Sample
Time
5—Layered Bags
lA 2A
Tedlar Bags
lB 2B
0
hr.
147
128
116
105
1.5
hr.
124
4
hr.
101
101
3
days
70
74
63
59
7
days
59
32
-------
VI. CONCLIJS IONS AND RECONMENDAT IONS
Overall, glass sampling bulbs were found to be the most satisfactory
containers for storing the test gas mixture, as compared to the Tedlar
and 5—layered bags.
Calibrated Instruments ’ 5—layered bags were satisfactory for
ethane, propylene, methyl chloride, ethylene oxide and benzene. Methanol,
however, did not show adequate storage stability. The experiments run to
check whether preconditioning the 5—layered bags with methanol would
decrease methanol losses, did not indicate that this technique would be
successful. Some interfering gas chromatography peaks were observed
from the 5—layered bags, which would interfere at ppb analysis in the
—161°C to 72°C boiling point range.
Single—layer Tedlar bags were also satisfactory for ethane, propylene,
methyl chloride and benzene in terms of storage stability. However, sub-
stantial interfering gas chromatography peaks which increased with time
were observed in samples taken from the Tedlar bags. The interferences
have been observed and studied by other Investigators 9 ’ 10 , who have con-
cluded that these GC peaks are contaminants in the polymeric bags, pre-
sumably the result of outgassing from the materials used to construct the
bag. In this study, the interferences were observed in the —160°C to 72°C
boiling point range, and as observed for the 5—layered bags, would be of
concern at ppb concentration levels. Ethylene oxide results were erratic
and methanol was not stable. As in the 5—layered bags, methanol precon—
ditioning of Tedlar bags did not decrease methanol losses. Furthermore,
the Tedlar bags were fragile and prone to develop leaks.
Glass sampling bulbs showed no significant deterioration in stored
samples. Coefficients of variation for analyses of replicate samples
were somewhat higher than expected, perhaps reflecting difficulties in
mixing the samples within the bulbs. Mixing difficulties could
possibly be minimized by using glass beads. The glass sampling bulbs
can be cleaned and silanized, decreasing the potential memory problems
from reusing containers. There are doubts as to whether or not the
polymeric bags could be cleaned sufficiently to eliminate “memory
effects” from previous gas sampling. The polymeric bags would appear
38
-------
to be more convenient for field sampling and shipping. However, the
cylindrical glass sampling bulbs used in this study do seem to be quite
rugged. High vacuum stopcocks can be used with the glass bulbs to
prevent leakage around the sampling valves during sample shipment.
The study indicates that glass sampling bulbs are the best overall
choice, in terms of gas storage stability, and are recommended for
collecting relatively small volumes of gas samples. Most compounds
in the low boiling point range, with the exception of certain reactive
species, may be readily and reliably identified and quantitatif led by
gas chromatography/mass spectroscopy (GC/MS) analysis methods. In
terms of the analysis method detection limits, it will be sufficient
for Level 1 on—site gas analysis and GC/MS analysis purposes, in most
cases, to collect the equivalent of 0.1 to 3.0 liters of gas sample as
grab or time integrated samples. These quantities are conveniently col-
lected in glass sample bulbs.
However, if much larger volumes are required, e.g., 500 liters for
biotesting, the polymeric bags are almost certainly more appropriate.
This study indicates that the 5—layered bag is superior to the Tedlar
bag, and would be appropriate for sampling non—polar gases at ppm levels.
When collecting very large samples, the increased container size has been
suggested 6 ’ 10 as a means of decreasing sample losses by minimizing the
container walls’ absorptive affects, minimizing water condensation and
slowing down reactions among sample components.
39
-------
VII. REFERENCES
1. Baker, R.A. and Doerr, R.C. “Methods of Sampling and
Storage of Air Containing Vapors and Gases,” [ nt. J. Air
Poll. Pergamon Press 2: 142—158 (1959).
2. Simnil, V.L. and Hershberger, A., “Permeability of Polymeric
Films to Organic Vapors,” Mod. Plast. 27. 97 (1950).
3. Cleminens, C.A. and Altshuller, A.P. “Plastic Containers for
Sampling and Storage of Atmospheric Hydrocarbons prior to
Gas Chromatographic Analysis,” J. Air Poll. Control Assoc.
14: 407 (October 1964).
4. Conner, W.D. and Nader, J.S. “Air Sampling with Plastic Bags,”
md. Hygiene Journ. 291—297 (May/June 1964).
5. Wright, B.J. “Stability Studies of Several Malodorous Sulfur
Compounds in Teflon, Tedlar and Glass Sample Containers,”
AII Report No. 73, Air and md. Hygiene Lab. (January 1970).
6. Schuetzle, D., Prater, T.J. and Ruddell, S.R. “Sampling and
Analysis of Emissions from Stationary Sources 1. Odor and
Total Hydrocarbons,” J. of the Air Poll. Control Assoc.
25, 9:925 (September 1975).
7. Levine, S.P., Hebel, K.G., Bolton, J. Jr. and Kugel, R.E.
“Industrial Analytical Chemists and OSHA Regulations for
Vinyl Chloride,” Anal. Chem. 47:1O75A (October 1975).
8. Polasek, J.C., and Bullin, J.A., “Evaluation of Bag Sequential
Sampling Technique for Ambient Air Analysis,” Env. Sci. and
Technology, 12, 6:708 (June 1978).
9. VanHaam, Joop, “Objections to the use of Polyvinyl Fluoride in
Smog chamber Experiments,” Chemosphere, 4:315—318 (1978).
40
-------
10. Dimitriades, Band Seizinger, D.E., “A Procedure for Routine Use
in Chromatographic Analysis of Automative Hydrocarbon Emissions,”
Env. ci. and Technology, 5, 3:223—229 (1971).
11. Calibrated Instruments, Inc., “Calibrated Instruments Summation
of Tests Conducted for Various Properties of Gas Sampling Bags,
Snout Type ” (November 1974).
12. Cleland, J.G. and Kingsbury, G.L. “Multimedia Environmental Goals
for Environmental Assessment, Volume II. NEC Charts and
Background Information,” EPA—600/7—77—136 (NTIS PB 276920)
(November 1977).
41
-------
TECHNICAL REPORT DATA
(Please read 1, smjczions on the reverse before completing)
1. REPORT NO. 2.
EPA-600/7-79-095
3. RECIPIENT’S ACCESSIOI*NO.
4. TITLE AND SUBTITLE
Gas Sample Storage
5. REPORT DATE
April 1979
6. PERFORMING ORGANIZATION CODE
7. AUThOR(S)
K.E. Thrun, J. C. Harris, and K. Beltis
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Arthur P. Little, Inc.
Acorn Park
Cambridge, Massachusetts 02140
10. PROGRAM ELEMENT NO.
INE 624
11.CONTRACT/GRANTNO.
68-02-2150, T.D. 10702
12. SPONSORING AGENCY NAME AND ADDRESS
EPA Office of Research and Development
.
Industrial Environmental Research Laboratory
Research Triangle Park, NC 27711
13. TYPE OF REPORT AND PERIQD COVERED
Task Final; 1/78 - 2/79
14. SPONSORING AGENCY CODE
EPA/600/l3
15.SUPPLEMENTARY NOTES IERL-RTP project officer is Larry D. Johnson, MD-62, 919/541-
2 557.
16. ABSTRACT
The report gives results of a laboratory evaluation to compare the storage stability
of selected gases covering a range of compound categories, in three types of con-
tainers: glass bulbs and two different polymeric sample bags. The studies indicate
that glass bulbs are the best overall choice, with no significant gas losses.
Reactive and highly polar gases were preferentially lost from samples stored in the
polymeric bags. Some contaminants were detected in the samples taken from the
polymeric sample bags, presumably the result of outgassing from the materials
used to construct the bags.
17. KEY WORDS AND DOCUMENT ANALYSIS
a. DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS
C. COSATI Field/Group
Pollution Polymeric Films
Gas Sampling
Gas Storage
Gases
Stability
Glassware
Pollution Control
Stationary Sources
l3B 111
14B
1SE
07D
11B
18. DISTRIBUTION STATEMENT
Unlimited
19. SECURITY CLASS (ThisReport )
Unclassified
21. NO. OF PAGES
46
20. SECURITY CLASS (This page)
Unclassified
22. PRICE
EPA Form 2220-1 (9:73)
42
------- |