EPA560/7-77-002
ACCUMULATION OF ORGANIC POLLUTANTS
BY
SOLID ADSORBENTS
FINAL REPORT
JUNE 1976
OFFICE OF TOXIC SUBSTANCES
U.S. ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON, D. C. 20460
-------
EPA 560/7-77-002
ACCUMULATION OF ORGANIC POLLUTANTS
BY SOLID ADSORBENTS
FEBRUARY 1977
FINAL REPORT
U.S. ENVIRONMENTAL PROTECTION AGENCY
OFFICE OF TOXIC SUBSTANCES
WASHINGTON, D. C.
-------
THIS REPORT WAS PREPARED BY ENERGY RESOURCES COMPANY INC.,
FOR THE U. S. ENVIRONMENTAL PROTECTION AGENCY, UNDER
CONTRACT NO. 68-01-2925. THIS REPORT HAS BEEN REVIEWED BY EPA,
AND APPROVED FOR PUBLICATION. APPROVAL DOES NOT SIGNIFY
THAT THE CONTENTS NECESSARILY REFLECT THE VIEWS AND
POLICIES OF THE ENVIRONMENTAL PROTECTION AGENCY, NOR
DOES MENTION OF TRADE NAMES OR COMMERCIAL PRODUCTS
CONSTITUTE ENDORSEMENT OR RECOMMENDATION FOR USE.
-------
TABLE OF CONTENTS
Pac
SECTION
1.1
1.2
1.3
1.4
1.5
SECTION
2.1
2.2
2.3
2.4
2.5
INTRODUCTION
ONE COMPARISON OF SOLID ADSORBENTS USING
STANDARD MIXTURES
Summary
Introduction
Experimental
1.3.1 Materials
1.3.2 Organic Mixture No. 1 - Volatile Organics
1.3.3 Organic Mixture No. 2 - Higher Molecular
Weight Organics
1.3.4 Polychlorinated Biphyenyls (PCB's)
1.3.5 Ethylene Glycol
Results and Discussion
Conclusions
TWO ANALYSIS OF VOLATILE ORGANICS BY DIRECT
ADSORPTION ONTO TENAX RESIN
Summary
Introduction
Experimental
Results and Discussion
2.4.1 Preliminary Studies
2.4.2 Comparison with Volatile Organics Analysis
2.4.3 Sampling and Analysis of Large Volumes of
Drinking Water
2.4.4 Reproducibility of Direct Adsorption
Conclusions
1
5
5
5
7
7
7
9
9
9
10
14
15
15
15
16
17
17
19
23
24
27
-------
TABLE OF CONTENTS (CONT.)
Paqe
SECTION THREE ACCUMULATION OF ORGANICS FROM NATURAL 29
WATERS ONTO TENAX RESIN
3.1 Summary - 29
3.2 Introduction 29.
3.3 Experimental 29
3.4 Results and Discussion 30
SECTION FOUR RESULTS AND FURTHER LINES OF INQUIRY 35
APPENDIX A DESCRIPTION OF METHOD FOR SAMPLING AND A-l
ANALYSIS OF ORGANIC WATER POLLUTANTS
-11-
-------
LIST OF TABLES
SECTION ONE COMPARISON OF SOLID ABSORBENTS USING
STANDARD MIXTURES
1-1 Characteristics of Solid Adsorbents 6
1-2 Characteristics of Test Compounds 8
1-3 Retention of Organic Compounds 12
SECTION TWO ANALYSIS OF VOLATILE ORGANICS BY DIRECT
ADSORPTION ONTO TENAX RESIN
2-1 Total Recovery Compared to Calculated Concentration 21
2-2 Concentrations of Chlorinated Hydrocarbons in 24
Cambridge Tap Water (yg/1)
2-3 Volatile Organics from Tap Water 27
-iii-
-------
LIST OF FIGURES
SECTION TWO ANALYSIS OF VOLATILE ORGANICS BY DIRECT
ADSORPTION ONTO TENAX RESIN
2-1 Heat-desorption apparatus, direct to GC 18
2-2 Heat-desorption Apparatus 20
2-3 GC on Chromosorb 101 '22
2-4 Sampling Arrangement 24
2-5 Volatile Components of Drinking Water (30 liters) 26
SECTION THREE ACCUMULATION OF ORGANICS FROM NATURAL
WATERS ONTO TENAX RESIN
3-1 Extracted Components'from Acushnet River 31
APPENDIX A DESCRIPTION OF METHOD FOR SAMPING AND
ANALYSIS OF ORGANIC WATER POLLUTANTS
A-l Procedure for Analysis of Organic Water Pollutants A^-2
A-2 Sampling System A-3
A-3 Heat-desorption Apparatus A-4
-iv-
-------
INTRODUCTION
With the passage of the Federal water and air pollution
control acts and the increasing concern over degradation of
the environment, there has been a growing demand for the
development of techniques for measuring trace amounts of
chemical contaminants in the environment. Efforts to improve
the ability to detect trace substances have proceeded on two
fronts; new instrumentation has been developed, and older
methods have been improved to detect ever-smaller quantities
of material. In conjunction with these efforts, methods
have been developed to quantitatively accumulate trace
chemicals from environmental media, thus concentrating the
sample prior to analysis.
A review of the literature on the available preconcen-
tration techniques for all types of environmental contaminants
was prepared as the first phase of this project and has been
published as "A Review of Concentration Techniques for Trace
Chemicals in the Environment" (EPA-560/2-75-003). An evalu-
ation of this material indicated that the area in which
there is the most need for an improvement in preconcentration
techniques is that of the analysis of trace organics.
Recently there has been an increasing concern with the
contamination of the nation's water by organic chemicals.
It has, therefore, become important to develop techniques
for the analysis of organic compounds in the very low concen-
trations (parts per trillion to parts per billion) in which.
they occur in the environment. While current techniques for
the analysis of volatile organics (gas chromatography and
gas chromatography/mass spectrometry) (GC/MS) can detect
minute quantities of a given compound (from 10~H to 10~13
grams), the severe limitation in the sample volume which can
be injected into a gas chromatograph limits the over-all
sensitivity of these analyses. The most promising approach
to increasing the sensitivity of the analysis of trace
organics is the design of a system to accumulate the organic
material from very large sample volumes. The standard
preconcentration technique is solvent extraction, but this
can become very cumbersome when applied to the large volumes
of water which are required for the analysis of very low
levels of contamination.
A more convenient technique for sampling large volumes
of water is the use of a solid to adsorb the organic contami-
nants. Activated carbon, which is well known for its ability
to bind organic molecules, has often been used for this
purpose. This is the basis for the Environmental Protection
-1-
-------
Agency (EPA) standard method for the determination of gross
organic contamination of water.1 Carbon, however, has
disadvantages as an adsorbent. Its efficiency is highly
variable depending upon its source, the activation proce-
dures, and the moisture content of the carbon during extrac-
tion. Its high activity can also lead to irreversible
adsorption of some species and decomposition of others.
Grob has designed a sampling system which uses carbon
as the adsorbent medium.2 This system shows a very high
degree of retention of organic compounds, but great care
must be taken to avoid contamination once it has been acti-
vated.
Recently, a number of polymeric resins have been studied
as replacements for carbon in adsorption columns. Junk and
coworkers at Iowa have examined XAD resins, and in particular
XAD-2, for their retention of organics from standard solutions
and from natural waters.3 Dr. Ronald Webb of the EPA South-
eastern Environmental Research Laboratory has investigated
these same resins for their ability to adsorb organics from
sewage treatment and industrial discharges.4 Laboratory
experiments indicate that XAD resins will adsorb a wide
variety of lipophilic organic compounds from dilute solution.
They are much less effective for the accumulation of alkanes
than for aromatic compounds, however. Among other polymers
R.W. Buelow, J.K. Car-swell, and J.M. Symons, "An
Improved Method for Determining Organics by Activated Carbon
Adsorption and Solvent Extraction - Part i;" Journal of the
American Water Works Association 65 (1973): 57; and R.W.
Buelow, J.K. Carswell and J.M. Symons, "An Improved Method
for Determining Organics by Activated Adsorption and Solvent
Extraction - Part II" (Test Method), JAWWA 65 (1973): 195.
2
Kurt Grob, "Organic Substances in Potable Water and in
Its Precursor, Part 1. Methods for Their Determination by
Gas-Liquid Chromatography," Journal of Chromatography 84
(1973): 255-273.
G.A. Junk et al., "Use of Macroeticular Resins in the
Analysis of Water for Trace Organic Contaminants," Journal of
Chromatography 9J9_ (1974): 745-762; and A.K. Burnham et al.,
"Identification and Estimation of Neutral Organic Contaminants
in Potable Water," Analytical Chemistry 4_4_ (1972): 139-142.
4
Ronald G. Webb, Isolating Organic Water Pollutants;
XAD Resins, Urethane Foams, Solvent Extraction, for the U.S.
Environmental Protection Agency, EPA-660/4-75-003, June 1975.
-2-
-------
which have been evaluated for this purpose, polyurethane and 5
coated polyurethanes have been used in sampling for pesticides
and Tenax GC will efficiently adsorb many pesticides and
high molecular-weight aromatics from dilute solution.6
This resin is also used as a trapping material for volatile
organic components which are removed from aqueous solution
using a gas sparging technique7 such as the EPA Volatile
Organics Analysis.8
The relative adsorption efficiencies of a number of
commercially available solids were obtained in a laboratory
study using standard mixtures containing a range of environ-
mentally interesting compounds. Section One describes these
experiments. The major differences which were observed in
the retention efficiencies of polymeric resins occurred for
moderately polar volatile organics such as benzene and
chloroform.
The few materials (Tenax GC, Chromosorb 101, and activated
carbon) which showed the most promise as accumulators of
volatile organics from solution were then examined for their
utility when heat-desorption was used to release adsorbed
compounds directly onto a gas chromatographic column for
analysis. Section Two describes the experiments which led
to the choice of Tenax GC as a good general adsorbent, and
to the design of a technique for the analysis of purgeable
organics using direct adsorption from water.
J.F. Uthe, J. Reinke and H. Gesser, "Extraction of
Organochlorine Pesticides from Water by Porous Polyurethane
Coated with Selective Adsorbent," Environmental Letters 3_
(1972) : 117.
V. Leoni, G. Puccetti and A. Grella, "Preliminary
Results on the Use of Tenax for the Extraction of Pesticides
and Polynuclear Aromatic Hydrocarbons from Surface and
Drinking Waters for Analytical Purposes," Journal of Chroma-
tography 106 (1975): 119.
Thomas G. Bellar and J.J. Lichtenberg, "Determining
Volatile Organics at Microgram-per-Litre Levels by Gas
Chromatography," for the U.S. Environmental Protection
Agency, EPA-670/4-74-009, November 1974, in Journal of
the American Water Works Association 66 (1974): 739-744.
g
A. Zlatkis, H.A. Lichtenstein and A. Tishbee, "Concentra-
tion and Analysis of Trace Volatile Organics in Gases and
Biological Fluids with a New Solid Adsorbent," Chromatographia
6 (1973): 67.
-3-
-------
Once it had been shown that Tenax could effectively be
used to concentrate volatile organics from aqueous solution,
its utility as an adsorbent for the less volatile materials
which are found in the environment was investigated, as
described in Section Three. Drinking water and surface
water samples were passed through Tenax columns; the volatile
components were heat-desorbed and the remaining organic
compounds were extracted with ethyl ether. The GC/MS iden-
tification of the compounds which were obtained by this
method indicate the range of materials which this resin will
accumulate. Compounds which have been identified include
both aliphatic and aromatic materials.
Using Tenax resin as the adsorbent, a complete sampling
and analytical procedure for both volatile and less-volatile
gas chromatographic organic analysis has been designed.
This procedure is described in Appendix A. A simple field
sampling system for use with adsorbent columns has also been
constructed and tested.
This standard sampling procedure can now be used to
compare a number of resins in terms of their utility as
adsorbents in sampling a variety of environmental water
samples. Such a comparison would be of value in finding
methods to extend the range of compounds which can be accu-
mulated by an adsorbent column. A thorough evaluation of
the ability of Tenax and other resins to retain alkanes
would be particularly useful. An efficient accumulation
system for these compounds is very desirable at the moment
for use in baseline oil contamination studies. Further
efforts which could be made in this area are described in
Section Four.
-4-
-------
SECTION ONE
COMPARISON OF SOLID ADSORBENTS USING STANDARD MIXTURES
1.1 Summary
A laboratory study has been conducted to determine the
retention of a number of selected organic compounds on poly-
meric resins, activated carbon, polyurethane foam, and
uncoated glass beads when eluted with water. Of the adsor-
bents tested, carbon was the only one which gave close to
quantitative retention of volatile compounds (CHC13, benzene,
CH3CC13) while Chromosorb 101 and Porapak Q gave moderate
recoveries (65 to 70 percent) of benzene and methylchloroform.
Higher molecular weight compounds were retained well by all
of the solids, including significant retention by the glass
beads. These compounds could not be eluted efficiently from
the activated carbon, however, and this appears to be a
result of irreversible adsorption under the experimental
conditions.
1.2 Introduction
In order to identify and determine the concentration of
organic pollutants in the environment it is necessary to
isolate them from the environmental matrix (air, water, or
soil) and to concentrate the sample by several orders of
magnitude. Liquid/liquid extraction of aqueous samples
becomes very cumbersome for large volumes and the need for
an effective adsorbent to replace carbon for the accumulation
of organic compounds has been widely recognized. Therefore
the retention efficiencies of a variety of lipophilic solids
for selected groups of organic compounds have been examined.
Most of these solids are porous polymers which have been
produced for liquid or gas chromatography. The physical and
chemical characteristics of these solids are listed in Table
1-1.
Although the basic structures of most of these polymers
are very similar, and therefore their retention of organics
should be similar, a comparison is useful in order to ascer-
tain the effect of differences in polarity, surface area,
and functional groups.
-5-
-------
TABLE 1-1
CHARACTERISTICS OF SOLID ADSORBENTS
ABSORBENT
CARBON
POLYURETHANE
XAD-2
XAD-7
TENAX GC
CHROMOSORB 101
CHROMOSORB 103
CHROMOSORB 104
PORAPAK Q
ABBREVIATIONS :
MATERIAL
Coconut Charcoal
Polystyrene
Cross-linked
Acrylic Ester
Poly 2,6-diphenyl
p-phenyleneoxide
STY-DVB
Polystyrene
ACN-DVB
EVB-DVB
STY = Styrene
DVB = Di-vinyl Benzene
SURFACE PORE
MESH AREA SIZE
SIZE (M2/g) (y)
60/80
20/50 330 .009
20/50 450 .008
60/80
60/80 <50 .3-. 4
60/80 15-25 .3-. 4
60/80 100-200 .06-. 08
50/80 50-600
ACN = Acrylonitrile
EVB = Ethylvinylbenzene
TEMP.
MAX.
<°0
>700
-87
250
150
320
275
275
250
250
POLARITY
Nonpolar
Moderately Polar
Nonpolar
Dipole = . 3
Moderately Polar
Dipole =1.8
Nonpolar
Strongly Polar
Absorbs Acids
Highly Polar
Nonpolar
-------
The compounds were selected for inclusion in the standard
mixtures on the basis of the variety of chemical structures
which they represented and because of current interest in
them as environmental pollutants. They are listed in Table
1-2.
1.3 Experimental
1.3.1 Materials
The solid adsorbents and the sodium sulfate for drying
were cleaned by exhaustive Soxhlet extraction with ethyl
ether and with methanol. An ether eluate of the materials
after this procedure showed no significant organic contami-
nation.
Identical glass columns (20 ml burets) were packed with
approximately 1 gram (g) or 5 cubic centimeters (cm-*) of an
adsorbent, with plugs of silanized glass wool above and
below the packing material. The packed columns were rinsed
with ether, methanol, and organic-free water.
All chemicals were reagent grade, used without further
purification.
1.3.2 Organic Mixture No. 1 - Volatile Organics
This mixture consisted of benzene (2.2 g), chloroform
(22.48 g) , methylene chloride (13.36 g) acrylonitrile (3.96
g), and methyl chloroform (23.19 g) .
In these experiments, a 50 microliter (yl) quantity of
the standard solution was placed directly on each column by
means of an Eppendorf pipet. One liter of tap water was
then run through the column.
After elution with water, the columns were extracted
with methanol (first run) or N,N-dimethyl formamide (DMF)
(second and third runs). The first 5.0 milliliters (ml)
of extract were collected. Concentration by evaporation was
omitted due to the volatility of these organic accumulants.
Gas chromatographic separation was accomplished on a
Chromosorb 101 column, 6 feet in length, using a flame-
ionization detector. The temperature program was an initial
140° C increasing at the rate of 8"/minute to a final
temperature of 220° C.
-7-
-------
oo
TABLE 1-2
CHARACTERISTICS OF TEST COMPOUNDS
TEST
COMPOUNDS
CH2C12
CHC13
CH3CC13
CC13CC13
ACRYLONITRILE
BENZENE
ETHYLENE GLYCOL
1-NAPHYTHYL AMINE
PYRENE
TRI-p CRESYL
PHOSPHATE
3,3' DICHLORO-
BENZIDINE
M.W.
(a.u.)
85
119
133
237
53
78
62
193
202
368
253
m.p.
CO
-97
-63
-33
187
, -82
6
-13
50
156
-28
132
b.p.
CO
41
61
74
sub.
78
80
198
301
sub.
404
410
v.p.
(25°) H20
(nun Hg) solubility
400 1:50
250 1:200
150 i
< 1 i
125 1:14
100 1:1,430
< 1 °°
«1 1:590
<1:500
V . S . S .
dipole
1.57
1.02
1.79
3.83
0
2.20
ABBREVIATIONS: i = insoluble
v.s.s. = very slightly soluble
-------
1.3.3 Organic Mixture No. 2 - Higher Molecular
Weight Organics
This mixture consisted of a solution of: pyrene (0.64 g) ,
hexachloroethane (5.00 g) , 1-naphthylamine (0.64 g), 3,3' di-
chlorobenzidine (2.00 g) , and tri-p-tolyl phosphate (1.60 g)
in 100 ml of methylene chloride.
A 0.1 ml aliquot of this mixture was pipetted directly
onto each column. The passage of 1.0 liter of tap water
followed immediately.
The columns were then eluted with ethyl ether, a total
of 5 bed-volumes, about 25 ml. The ether extract was dried
with anhydrous, previously cleaned, sodium sulfate and
concentrated in a Kuderna-Danish evaporator to a standard
sample volume of 0.5 ml.
A 3 percent OV-1 column was used to separate these com-
pounds with a program of 70° C for 4 minutes rising to
250° C for 4 minutes at a rate of 32°/minute.
1.3.4 Polychlorinated Biphenyls (PCS's)
An aliquot of 0.1 ml of an Aroclor 1242 solution (1
in iso-octane) was placed on each wet column. One liter of
tap water was then passed through the column and, following
elution with water, the columns were extracted with 25 ml of
ethyl ether,.! bed-volume followed by 4 bed-volumes. The
ether extract was filtered through anhydrous sodium sulfate
and concentrated in a Kuderna-Danish evaporator to 0.5 ml.
The analysis was performed with a Ni electron capture
detector using a 5-foot, 1/8 inch stainless steel column
packed with 3 percent OV-1. A temperature program of 4
minutes at 172° C increasing by 4°/minute to 192° C and
remaining there for 8 minutes was used.
1.3.5 Ethylene Glycol
A 1 cm sample of a solution of ethylene glycol in
water (1 cm^/l) was added to the top of each column. This
was followed by elution with 1 liter of tap water.
The columns were extracted with 4 bed volumes of
methanol and the solution was concentrated to a volume of
0.5 cm3.
— 9 —
-------
Analysis was performed on a Carbowax 20M column with a
flame ionization detector and a constant temperature of 155'
C.
1.4 Results and Discussion
The method used in determining retention efficiencies
for the solutions described above was that of direct appli-
cation of the material of interest to the top of a column
followed by elution with water. This does not mimic the
situation which occurs in environmental monitoring, in which
the solute is adsorbed from a very dilute solution in water.
An adsorbent column used in this manner, however, is also a
chromatographic column and the same theoretical considera-
tions apply. The volume of solvent which is required to
elute a given solute from a column of a given size is
constant regardless of the amount of solute which is ad-
sorbed as long as this amount does not load a significant
portion of the column. Therefore, if a material is eluted
from a column with a liter of water, the same material,
dissolved in the water, would also have been eluted from the
column. If elution does not occur, the adsorbent has a much
higher affinity for the solute than water does, and retention
from a dilute solution would be quantitative as long as
there is sufficient contact with the adsorbent for proper
adsorption.
This technique was used to avoid low yields due to
adsorption of standard solutions onto container and glass
column walls. Methylene chloride extractions of duplicate
standard solutions in water gave recoveries which varied
with the time elapsed prior to extraction, leading to
uncertainty about the quantity which was actually applied to
the adsorbent. In similar experiments standard solutions
have been made directly in the chromatographic column in
order to avoid this problem.9 The subsequent extraction of
the column, however, would also elute any material which had
adsorbed to the glass walls of the apparatus, a factor that
was desirable to avoid. A column packed with unsilanized
glass beads was included in the experiment to indicate the
amount of adsorption which could be attributed to glassware.
9
A.K. Burnham et al., Analytical Chemistry 44 (1972):
193-142; G.A. Junk, Journal of Chromatography 99 (1974):
745-762.
-10-
-------
The retention efficiencies which were obtained in these
studies are listed in Table 1-3. A precision of + 6 percent
was obtained for two replicate experiments unless otherwise
indicated.
Since all of the porous polymers are lipophilic in
character, it was expected that there would be no major dif-
ferences in their accumulation of lipophilic organic compounds
Differences in polarity and surface area, however, were
expected to lead to differences in the accumulation of the
more volatile or water-soluble test chemicals.
The results are consistent with these expectations.
None of the adsorbents accumulated the very water-soluble
ethylene glycol, or the polar acrylonitrile, while the other
low molecular weight compounds (CHC13, CHCC13, and benzene)
were only partially retained by the polymers.
Activated carbon gave by far the highest retention of
volatile compounds. Recovery of the higher molecular-weight
compounds from activated carbon, however, was dependent on
the past history of the material. Recoveries were very low
when fresh carbon was used, but they improved substantially
when the same column was reused without activation, other
than thorough elution with an organic solvent. This is
indicated by the second run recoveries of pyrene, tricresyl
phosphate, and dichlorobenzidine shown in Table 1-3. These
results are apparently due to the irreversible adsorption of
these materials by carbon under the experimental conditions.
The most active sites were already occupied (deactivated)
when the second standard solution was added, permitting a
larger percentage of this standard to be eluted. A simple
change of solvent from ether to methylene chloride or
chloroform did not improve the recoveries. Carbon, therefore,
appears to be a more efficient adsorbent of a wide variety
of organic compounds, but it requires much more stringent
conditions for activation and desorption than the polymeric
materials.
The polymers, Tenax GC, Porapak Q, and Chromosorb 103
are excellent accumulators of benzene and chloroform in the
vapor phase. " Porapak Q and Chromosorbs 101 and 103 also
retain these compounds efficiently from an aqueous system
while XAD-2 shows moderate efficiency; the low recoveries
from Tenax are surprising, however. Further extraction of
10T.G.'Bellar and J.J. Lichtenberg, JAWWA 66 (1974):
739-744.
-11-
-------
TABLE 1-3
RETENTION OF ORGANIC COMPOUNDS
CARBONa FIRST
SECOND
1
-^ . PORAPAK Q
^J
1
TENAX GC
CHROMOSORB 101
CHROMOSORB 103
CHROMOSORB 104
XAD-2
XAD-7
POLYURETHANE
GLASS BEADS
01
i-l
B
50
1
<2
<1
3
—
5
—
—
0
H
y
5
99
15
7
27
13
5
25
<10
5
<2
n
?,
5
94
69b
29
70
63
50
53
29
<2
<3
W
IS]
&
w
m
(P E
83
67b
31
68
59
44
48
38
0
<2
vo
rH
u1
R C E
38
48
49
b
48
46
50
52
59
52b
49
43
§
S
E U
u &
H H
Q Q
1 H
- tSJ
n *z
ro pq
N T)
10
30
76
78
72
b
74
60
81
78
76
70
&
g-i
ffi
Sig
? S
rH i^
4
4
4
b
45°
32b
b
38
31
51b
44
4
<2
H
2
§
SH
P4
17
61
76
72
71
76
71
80
84
88
93b
^
>i
w
§g
t> rt;
1 X
Q. d<
1 Ul
H O
gg
37
85
80
84
79
80
76
83b
86
92
98b
«
s
U
-------
the Tenax column did not result in the elution of significant
amounts of these materials, indicating that complete elution
had been achieved. While these low recoveries were consis-
tently reproducible under the experimental conditions used
in this study, further work using heat-desorption of volatile
organics (see Section Two) has shown that Tenax is highly
effective in concentrating volatile organics from dilute
aqueous solution. This discrepancy has not been explained.
Most of the larger aromatic compounds were well retained
by all of the materials. A variety of aromatic compounds,
including phenols and cresols, have previously been effi-
ciently accumulated using XAD-2,H Chromosorb 102,12 and
activated charcoal,^ while pyrene has been collected using
Tenax GC^4 as the adsorbent, and polyurethane15 has been used
for PCS recovery. The low recovery of 1-naphthylamine (40
to 60 percent) in all cases was due, at least in part, to
partial decomposition of this compound during the experi-
mental procedure.
The very high retention of pyrene, tricresyl phosphate,
PCB's and dichlorobenzidine by the glass beads was unexpected.
Unsilanized glass is not free of adsorptive properties, as
was clearly indicated by the decrease in concentration of
standards made up in glass containers. Low surface-area
A.K. Burnham et al., Analytical Chemistry 4_4_ (1972):
139-142; G.A. Junk, Journal of Chromatography 99 (1974):
745-762.
J.P. Mieure and M.W. Dietrich, "Determination of Trace
Organics in Air and Water," Journal of Chromatographic
Science 11 (1973): 559-570.
James W. Eichelberger, Ronald C. Dressman, and James
E. Longbottom, "Separation of Phenolic Compounds from Carbon
Chloroform Extract for Individual Chromatographic Identifi-
cation and Measurement," Environmental Science and Technology
4_ (1974) : 576-578.
14
V. Leoni, G. Puccetti and A. Grella, Journal of Chroma-
tography 106 (1975): 119.
James W. Bedford, "The Use of Polyurethane Foam Plugs
for Extraction of Polychlorinated Biphenyls (PCB's) from
Natural Waters," Bulletin of Environmental Contamination and
Toxicology 12 (1974): 5.
-13-
-------
glass beads, however, were not expected to show the observed
high retention of these aromatic hydrocarbons. Passage of a
dilute solution of tricresyl phosphate through a bed of glass
beads resulted in greater than 90 percent retention by the
beads, a result comparable to that obtained by direct addition
of a concentrated solution to the column. The values obtained
for pyrene on all of the adsorbents, however, are more
closely related to the ease of extraction through the material
than to adsorption effects since, with its very low water
solubility, it precipitated immediately upon contact with
the wet column and did not totally dissolve in the volume of
elution water.
1.5 Conclusions
A comparison of the abilities of a number of polymeric
solids to retain organic chemicals when eluted with water
has shown that there is little variation in their retention
of highly water-insoluble compounds. The nonpolar polystyrene
resins are the most effective polymers for the accumulation
of the volatile organics from water, complementing their
adsorption properties for vapor-phase organics. While
activated carbon is a more powerful adsorbent, particularly
for the volatile organic substances, irreversible adsorption
may lead to a loss of some components unless stringent
extraction procedures are used.
The information gained from this study has been used to
choose the solid adsorbents for investigation in the develop-
ment of a technique for sampling and analyzing the organic
contaminants of drinking and surface waters using an adsorbent-
packed column.
-14-
-------
SECTION TWO
ANALYSIS OF VOLATILE ORGANICS BY
DIRECT ADSORPTION ONTO TENAX RESIN
2.1 Summary
After a preliminary study of a number of adsorbents,
Tenax GC was chosen for the development of a technique for
analyzing the most volatile components of drinking water by
direct adsorption from water followed by heat-desorption
onto a GC analytical column. This method gives results
which are comparable to the EPA Volatile Organics Analysis
(VOA)16 gas-sparging technique for the most common volatile
organics.
2.2 Introduction
The VOA gas-sparging method is a very sensitive technique
for measuring the concentrations of low molecular weight
organic contaminants in drinking waters. Because the entire
organic content of the sample is analyzed at one time, rela-
tively small sample volumes are required in order to obtain
measurements at the part per billion (ppb) level. For less
volatile compounds which are generally isolated by extraction,
however, only a portion of the extract can be analyzed at
one time. Very large sample volumes are then necessary to
isolate the quantities of material which are needed for
qualitative and quantitative analyses. The use of a solid
adsorbent to accumulate the trace organic constituents of a
water sample is generally regarded as the most efficient
sampling method for these large volumes. It would be advan-
tageous, therefore, to collect the most volatile consti-
tuents of a sample in the same manner, thus eliminating the
need to ship bottles of water from the sampling site to the
analytical laboratory with the consequent danger of loss of
components. A sampling cartridge which collects the entire
volatile range of lipophilic organics (the range which can
be analyzed by gas chromatography) would be ideal.
16T.A. Bellar and J.J. Lichtenberg, JAWWA 66 (1974): 739.
-15-
-------
An adsorbent should meet four requirements if it is to
be used as an accumulator of the total range of volatile
organic contaminants from drinking and surface waters: (1)
it must retain volatile organics strongly, (2) it must be
heat-stable to allow for heat-desorption of these volatile
components, (3) it should reversibly adsorb less volatile
organics, and (4) it should be easily purified.
Previous experiments (Section One) have indicated that
activated carbon is an excellent adsorbent for low molecular
weight organics, and that a number of polystyrene resins are
also moderately efficient in retaining such compounds when
eluted with water. Initially, consideration was given to
activated carbon, Chromosorb 101, and Tenax GC as possible
"universal" adsorbents. XAD-2, which shows some promise for
retaining the volatile compounds, was not investigated
because of its low heat stability. Of these, Tenax GC was
chosen as the material to be used in designing and testing a
general procedure for sampling and analysis of trace organics,
2.3 Experimental
The following procedure was used for most of the experi-
ments discussed in this chapter.
1. Each sampling column consisted of a 1/2-inch by
6-inch stainless steel tube packed with approxi-
mately 1 gram of pre-extracted adsorbent, held in
place with silanized glass wool.
2. Prior to its use for sampling, each column was
thoroughly washed with methanol and ether, and
heated to above 150° C in a stream of nitrogen.
Heat desorption into the GC was then used to
verify that no volatile impurities remained on the
column.
3. Standard solutions were passed through the adsor-
bent columns under gravity or drinking water
samples were collected by attaching the column to
a tap with pipe fittings and running the water
through it as a rate of 30 cmvmin. Standard
solution A contained 250 y/1 each of methylene
chloride, chloroform, trichloroethylene and ethyl-
enedibromide in water. Standard solution B
contained 50 y/1 each of methylene chloride,
chloroform, trichloroethylene, dibromoethane, and
a-chlorobenzene in water.
-16-
-------
4. The column was then attached to a stream of pre-
purified nitrogen or argon and the free water was
blown from the column with a rapid stream of gas.
The gas flow was then reduced to 30 cm-^/min and
continued for 2 hours to further dry the packing
material.
5. The column was connected into the carrier-gas
line of a Perkin-Elmer 3920 gas chromatograph as
indicated in Figure 2-1. The gas passed through a
trap packed with precleaned sodium sulfate to
remove most of the water vapor which resulted from
heating the column. Heat-desorption was carried
out at 100° C or lower onto a Chromosorb 101
analytical column which was held at 50° C during
this period. The carrier gas was then set to
bypass the sampling column while the analysis was
performed.
6. For later samples on which gas chromatography/
mass spectrometric analysis was performed, the
columns were heat-desorbed in a gas stream which
was then run through a condenser to remove the
water vapor and through two smaller Tenax traps.
These traps were placed directly in the injection
port of the gas chromatograph, heat-desorbed at
250° C, and analyzed (see Appendix A). A Hewlett-
Packard 5981A GC/MS was used for the identification
of drinking water contaminants.
2.4 Results and Discussion
2.4.1 Preliminary Studies
A standard solution of methylene chloride, chloroform,
trichloroethylene, and dibromoethane (250 yg/1) was used to
examine the feasibility of direct adsorption from water
followed by heat-desorption as a method for analyzing volatile
water contaminants. All of these compounds could be effec-
tively desorbed from the three solids which were tested —
carbon, Chromosorb 101, and Tenax GC — but the water which
remained in the column after adsorption interfered with the
subsequent gas chromatographic analysis. This problem could
be eliminated by using a stream of prepurified nitrogen to
elute most of the water from the column prior to the desorp-
tion step. When nitrogen was passed through the columns at
a low flow rate of 30 cm3/min for 2 hours, only the methylene
chloride and a fraction of the chloroform were eluted from
the resins, and no material was removed from the carbon.
-17-
-------
BYPASS
oo
THREE-WAY-VALVES
INJECTION PORT
DRYING COLUMN
SAMPLING COLUMN
VARIAC
Figure 2-1. Heat-desorption apparatus, direct to GC.
-------
A drying tube packed with prewashed sodium sulfate, placed
in-line to the GC after the sampling column, was used to
remove the remaining water from the carrier gas stream.
Using these procedures, any water which did get onto the
analytical column was eluted at least 5 minutes before the
chloroform peak, and therefore did not interfere with the
analysis in any way.
Only one adsorbent was needed for the further development
of an analytical technique for organic water contaminants.
After initial consideration of Chromosorb 101, Tenax GC, and
activated carbon, the carbon was rejected because of extreme
difficulty in purifying it and in avoiding contamination prior
to analysis. Chromosorb and Tenax both retain volatile
organics quite well, even after a nitrogen drying step.
These two materials were compared for their retention and
desorption properties using 100 cm^ of standard solution A.
Each column was heat-desorbed at 80° C onto a Chromosorb 101
analytical column and analyzed. After a total of 48 minutes
of heat-desorption over 90 percent of all the materials had
been removed from the Tenax column, while the Chromosorb
column still retained from 20 to 40 percent of the standards
and carbon retained more than 20 percent of the dibromoethane.
Tenax GC was therefore chosen as the material to be
used in the development of a standard procedure of organic
sampling and analysis. Tenax GC is stable to high temperatures
(375° C), is easily purified, requires mild conditions for
the desorption of volatile organics, and shows a high reten-
tion of these materials.
2.4.2 Comparison With Volatile Organics Analysis
Most of the analyses which have been performed for the
.identification of volatile organic contamination in water
systems have used the VOA gas-sparging technique. This is,
therefore, the technique against which any new method for
the analysis of volatile organics must be compared. This
comparison has been made using both a spiked water solution
and using duplicate drinking water samples.
One hundred milliliters of standard solution B were
analyzed by the VOA method and by direct adsorption onto a
Tenax column. In this experiment the column was heat-
desorbed into the heat-desorption apparatus described in
Appendix A and shown in Figure 2-2. The same small Tenax
traps were used for both types of analysis. Table 2-1 gives
the total recovery from each method as compared with the
calculated concentration of the standard. The low values
-19-
-------
LARGE
TENAX
COLUMN
FLOW METER
t
TENAX TRAP
CONDENSER
Figure 2-2. Heat-desorption apparatus.
-20-
-------
for the recovery of some of these "components may be due
either to adsorption onto glassware or evaporation of the
standard during its preparation and use.
TABLE 2-1
TOTAL RECOVERY COMPARED TO CALCULATED CONCENTRATION
VOA ' TENAX GC
chloroform 92 87
trichloroethylene 66 63
dibromoethane 75 64
a-dichlorobenzene 96 74
Gas chromatographic analysis using an electron-capture
detector (BCD) gave a recovery of greater than 100 percent
for methylene chloride for both methods because of the inter-
ference of a small amount of water vapor in the BCD scan.
These data show that the Tenax recoveries, while slightly
lower than those from the VOA analysis are comparable for
highly volatile organics. The much lower recovery of dichloro-
benzene from the Tenax column is probably due to condensation
in the unheated heat-desorption apparatus.
The concentrations (50 ppb) of the standard solution
were rather high in order to avoid interferences from true
organic contaminants in the water which was used to make
the standards. The same method was therefore used to analyze
Cambridge tap water samples (.100 cm3) taken on the same day.
Four major BCD detectable substances were observed by each
method as shown in Figure 2-3. The quantities observed were
very similar, although not identical. The three largest
peaks were identified by GC/MS analysis as chloroform (peak
a), trichloroethylene (peak c) and bromodichloromethane
(peak d) .
-21-
-------
a c
Ni
I
Tenax adsorption analysis
100 cm3 tap water
VGA analysis -
100 cm3 tap water
Figure 2-3. GC on Chromosorb 101 (50° C to 250° C at
8"/minute BCD Detector.
-------
2.4.3 Sampling and Analysis of Large Volumes of
Drinking Water
The potential advantage of using a direct adsorption
technique instead of gas sparging for the analysis of
volatile water components is the ease with which large
volumes (i.e., gallons) can be sampled and analyzed at one
time. In order to obtain quantitative recoveries of compounds,
breakthrough of components must be prevented. In other words,
the volume of water required to elute an adsorbate must be
larger than the volume that is to be sampled. Because the
low-molecular weight organics are more easily eluted than
larger lipophilic molecules, the retention of these materials
with a Tenax column was investigated.
Progressively larger samples of Cambridge tap water
were analyzed and the retention of the three largest components,
chloroform, trichloroethylene, and bromodichloromethane were
computed. Table 2-2 lists the concentrations of these three
contaminants which were observed when 4-, 8-, and 30-liter
samples of tap water were analyzed. Since the Tenax columns
were attached directly to the taps for sampling, and the
samples were taken sequentially, there is no basis for
estimating the variability to be expected between samples.
The values for the two 4-liter samples may, however, give
some indication of the variability which may be expected.
Only the chloroform shows a decrease in apparent concentration
which is large enough and consistent enough to be attributed
to losses due to elution during the .sampling period. The
sampling column used in this work contained approximately 1
gram of Tenax GC. The use of a larger column could improve
the recovery of chloroform, but this technique cannot be
considered quantitative for compounds with chloroform's high
volatility.
-23-
-------
TABLE 2-2
CONCENTRATIONS OF CHLORINATED HYDROCARBONS
IN CAMBRIDGE TAP WATER (yg/1)
COMPOUND
CHC13
C-HC1,,
2 3
CHCl2Br
VOLUME OF WATER SAMPLES (1)
44 8 30
6.2 4.8 2.3 1.1
0.2 0.3 0.4 0.4
6.4 5.3 6.5 3.8
2.4.4 Reproducibility of Direct Adsorption
The reproducibility of Tenax adsorption of trace organics
from drinking water was investigated by splitting a stream
of tap water through two sets of sampling columns (Figure
2-4). After 30 liters of water had passed through each
column, heat-desorption and GC/MS analysis led to the iden-
tification of the nine major organic contaminants listed
below.
Tap
J
B
Figure 2-4. Sampling arrangement.
-24-
-------
a. chloroform CHC13
b. trichloroethylene C2HC13
c. dichlorobromomethane CHCl2Br
d. mesityl oxide (CH3)C=CHCOCH3
e. toluene CgH,-CH3
e'. tetrachloroethylene <~2C^4
f. chlorodibromomethane CHClBr-
g. ethyl benzene CgHj^Hc
h. tetrachloroethane C2H2C14
Figure 2-5 shows the chromatogram obtained from Column
A with the identified peaks labelled. Table 2-3 gives a
comparison of the GC peak heights and the calculated concen-
tration of components which were obtained from the two
duplicate columns. The values obtained from Column A are
consistently higher than for Column B, probably an indication
of unequal splitting of the stream due to inaccurate flow
control. Except for the most volatile species, however, the
results show that the technique gives essentially reproducible
results.
-25-
-------
Figure 2-5. Volatile components of drinking water (30 liters)
-26-
-------
TABLE 2-3
VOLATILE ORGANICS FROM TAP WATER
COLUMN A
COMPOUND PEAK HEIGHT
a.
b.
c.
d.
e.
f.
f '
g-
h.
h1
CHC13
C2HC13
CHCl2Br
Mesityl
oxide
C2C14 +
toluene
CHClBr2
_ p
Ethyl
benzene
C2H2C14
- ?
46
62
297
25
138
47
10
45
36
15
CONG.
0
0
13
0
yg/i
.8
.4
.0
.1
0.3
(toluene)
2
0
0
.1
-
.07
.2
-
COLUMN
PEAK HEIGHT
23
29
182
22
100
38
9
44
29
13
B
CONC.
0
0
8
0
yg/i
.4
.2
.0
.1
0.2
(toluene)
1
0
0
.7
-
.07
.2
-
Columns C and D were also analyzed to give an indication
of the degree to which components were eluted from the
first sampling columns into the second ones. The gas chroma-
tograms of the heat-desorbed organic components from these
columns show less than 10 percent of the material found on
the first column for most of the identified compounds.
Toluene is an exception, with a recovery of 20 percent on
the second column.
2.5 Conclusions
These studies show that a direct adsorption of volatile
organic contaminants of drinking waters can be accomplished
using Tenax GC as an absorbent. This method is comparable
-27-
-------
to the VOA analysis for compounds with volatilities above
that of chloroform, and it simplifies the sampling of very
large volumes of water (and therefore the detection of very
minor contaminants). Heat-desorption may also be followed
by extraction of the column to obtain less volatile components
as described in Section Three.
-28-
-------
SECTION THREE
ACCUMULATION OF ORGANICS FROM NATURAL
WATERS ONTO TENAX RESIN
3.1 Summary
The general utility of Tenax GC as an adsorbent for
trace water contaminants was examined. Surface and drinking
water samples were passed through a trapping column which
was subsequently heat-desorbed to remove volatile components.
Extraction with ether was used to remove less volatile
contaminants which were then analyzed by gas chromatography/
mass spectrometry. Among the compounds which were accumulated
were alkanes, PCB's, and phthalates.
3.2 Introduction
An adsorbent column packed with Tenax GC has been used
to accumulate trace volatile organic contaminants of drinking
water by direct adsorption (see Section Two). Since there is
a need for a system which will collect a wide range of
organic contaminants, the ability of this material to retain
less volatile materials which can be removed subsequent to
heat-desorption of the volatiles was examined. Laboratory
studies^ have already shown high retentions of non-polar
materials by Tenax. To derive more information on its
utility for field sampling, the compounds which were accumulated
by Tenax from drinking and surface water samples were examined.
3.3 Experimental
Surface water samples were pumped through a stainless
steel line and Tenax-packed glass sampling columns (described
in Appendix A) at a rate of 5 liters per hour. Drinking
water samples were collected by attaching stainless steel
sampling columns directly to the tap and running the water
at a flow rate of approximately 50 cm^/minute until 20
liters had been sampled.
V. Leoni, G. Puccetti and A. Grella, Journal of
Chromatography 106 (1975): 119.
-29-
-------
Both sets of columns were heat-desorbed (see Section
Two), cooled, and extracted with anhydrous ethyl ether. The
larger glass columns were extracted with 250 cm^, while 50
cm^ were used for the smaller stainless steel columns. The
ether was then evaporated to less than 0.5 cm^ on a Kudern-
Danish evaporator, transferred to a graduated vial and
further reduced under a stream of nitrogen to 0.1 cm .
Preliminary GC analyses were run on a Perkin-Elmer 3920
GC using a 12-foot SP2100 (Supelco) column and both FID and
ECD detectors on a split effluent line.
GC/MS analyses were carried out on a Hewlett-Packard
5981A system using either an SP2100 or an OV-1 GC column.
3.4 Results and Discussion
The two surface water samples were taken from the
Acushnet River, at New Bedford, Massachusetts, and the south
branch of the Pawtuxet River, in southeastern Rhode Island.
The Acushnet at the sampling point is tidal, and quite
polluted from the manufacturing plants which line its banks.
Near the sampling point (which was just offshore), were a
machine shop, capacitor manufacturer, and leather goods
factory. Sewage pollution was also obvious. The Pawtuxet
River is a small, swift-flowing stream which passes by a
number of small industrial plants. The sampling location
was about one mile downstream from a large chemical plant.
The major classes of contaminants which were accumulated
from a 5-gallon sample from the Acushnet River by the Tenax
were alkanes, PCB's and phthalates. The mass chromatogram
of the extracted components is shown in Figure 3-1. Those
compounds which could be identified are listed below next to
their assigned peak number.
2 C,. alkane
3 C15 alkane
5 dichlorobiphenyl
6 C,g alkane
7 a phthalate
8 C,_ alkane
9 C,_ alkane
10-14 trichlorobiphenyls
15-16 tetrachlorobiphenyls
-30-
-------
I
U)
28
Figure 3-1. Extracted components from Acushnet River, mass chromotogram
3 percent OV-1.
-------
17-18 pentachlorobiphenyls + C,g alkane
19 C2Q alkane
20 diphenyl phenol or biphenyl phenyl ether
21 C21 alkane
22 trihydroxyxanthone
23 dibutyl phthalate
26 C25 alkane
27-28 dioctyl phthalate
The concentration of PCB's in this water was 0.07 yg/1
based upon a comparison with a standard Arochlor mixture.
The alkanes were recovered in the following amounts:
C14 0.3 yg C2Q 1.1 yg
C15 1.0 yg C21 1.9 yg
C17 3.2 yg C25 0.5 yg
Other alkanes were undoubtedly present, since alkane peaks
were present in the background of many of the mass spectra
in which other compounds were also identified. The GC con-
ditions could not be improved to the point where a separation
of these components could be obtained.
A similar analysis of a 6-liter sample of water col-
lected at the same time and place led to the recovery of the
same range of hydrocarbons and PCB's/ and in addition,
diphenyl amine and diphenyl quinone were also observed in
this sample. Quantitative comparisons of the recoveries
were beyond the scope of this work which was carried out
using packed columns which could not give the necessary
resolution of the hydrocarbon background and other organics.
The Pawtuxet River sample (5 gal), also contained a large
hydrocarbon background in the Cig to C25 region. Above this
background the series of compounds listed below were observed.
bromochloroaniline (2 isomers)
tribromoaniline
chlorodibromoaniline (.3 isomers)
N-ethyl carbazole
dichlorobiphenyl
dibromodichloroaniline
dioctyl phthalate
-32-
-------
The Tenax traps used for both of these surface water
samples were heat-desorbed for volatiles prior to extraction
with ether. Only trace quantities of volatile compounds
were observed.
The less volatile components of a 20-liter drinking
water sample have been obtained and analyzed in the same
manner as the surface waters. Below are listed the compounds
which have been identified from Cambridge tap water. In this
case the major classes of substances which were observed using
this analytical procedure were halogenated compounds or
simple aromatic molecules.
mesityl oxide
toluene
chlorodibromomethane
bromoform
tetrachloroethylene
ethyl benzene
xylenes
benzaldehyde
ethyl toluene
C. - benzenes
naphthalene
methyl naphthalenes
coumarin
Two adsorbent columns were used in series for the
analysis of the drinking water and Pawtuxet samples. The
extracts from the second columns had much lower levels of
organic material, although some hydrocarbons and phthalates
did pass through from the first columns. Since XAD resins,
which are often used for organic accumulation, do not retain
alkanes well, the efficiency of Tenax for the recovery of
these materials should be investigated in more detail.
These analyses have shown that Tenax GC is an effective
adsorbent of organic contaminants from clean, polluted, fresh,
and brackish waters. In general the types of compounds
which it will accumulate are consistent with its lipophilic
chemical structure and with the results of previous laboratory
tests with standard solutions. Water-insoluble lipophilic
compounds such as alkanes, aromatics and PCB's are retained
by the resin. In addition to these classes of compounds the
-33-
-------
more polar substituted anilines and halogenated methanes are
adsorbed by Tenax. As expected, however, no water-soluble
compounds were observed in the GC/MS analyses.
Specific recovery studies must be carried out if Tenax
GC is to be used in quantitative studies of individual
pollutants. It seems to meet the need for a general adsorbent
of organics, however, because of its high retention of a
wide range of common organic pollutants. This property makes
it particularly appropriate for use in drinking water
analyses.
-34-
-------
SECTION FOUR
RESULTS AND FURTHER LINES OF INQUIRY
The results of this work have shown that a system can
be devised, using a polymeric adsorbent/ for the analysis of
a wide range of organic water pollutants from the volatile
halogenated methanes and ethanes to higher molecular weight
substituted aromatics and alkanes. A number of additional
experiments would be useful first in determining the full
range of utility of the present technique, and secondly in
developing possible modifications to broaden its application.
One possible method of extending this technique would be
to use a different (or additional) adsorbent from Tenax GC.
Tenax was used as the adsorbent in the development of an
analytical technique because of its high retention of the
small halogenated drinking water contaminants and of the
ease of heat-desorbing materials from this resin. Other
resins, notably the Chromosorb or Porapak series, are also
efficient accumulators of volatile organics. Desorption
from these materials is slower, and they retain a larger
amount of water, but with the recent improvements which have
been made in the heat-desorption procedure they may prove to
be at least as effective as Tenax. There are two major
advantages to finding an alternative to Tenax as an adsorbent:
a reduction in the cost of the resin, and the ability to use~
chlorinated extraction solvents in which Tenax dissolves.
A number of resins are unsuitable for heat-desorption
because of their low heat-stability. The use of other
materials in addition to Tenax, however, could extend
the range of accumulated materials for extraction. Polar
resins, support-bonded liquids or ion-exchange resins may
be effective in adsorbing those polar compounds which pass
through Tenax and other lipophilic resins. A support-
bonded long-chain alkane may prove to be the most effective
adsorbent for the analysis of trace hydrocarbon contamination.
New materials can be compared to Tenax as general adsor-
bents of organic water contaminants by using the standard
analytical procedure which has been developed with Tenax.
A direct comparison can be made of the compounds which are
obtained from drinking and surface waters using a number of
different adsorbents. This should highlight differences in
organic retention under field-sampling conditions. Materials
-35-
-------
which look promising can then be-studied for their retention
of particular compounds of interest.
•
The Tenax system itself should be further tested for
its ability to retain hydrocarbons and, in particular,
alkanes. Analyses on river water samples have shown that
Tenax will accumulate these compounds, but the retention
efficiency and recovery for a range of alkanes should be
examined further using dilute standard solutions. A good
adsorbent for aliphatic and aromatic hydrocarbons would be
an immense aid in analyses for trace oil contamination or
baseline hydrocarbon levels.
The sampling and analysis technique using a resin
adsorbent can be further developed by the design of an
automated sampler which incorporates all of the necessary
control apparatus in a neat package. Part of this develop-
ment would involve a more thorough investigation of the
optimum sampling rates and column sizes as related to the
sample size. In the experiments run to date the amount of
packing and flow-rates used have been very conservative to
assure that contact time with the resin is sufficient for
complete adsorption.
In addition to the studies suggested above, more analyses
conducted using the present system on a wide range of samples
will lead to a greater understanding of its utility and its
limitations.
-36-
-------
APPENDIX A
DESCRIPTION OF METHOD FOR
SAMPLING AND ANALYSIS OF ORGANIC WATER POLLUTANTS
-------
APPENDIX A
DESCRIPTION OF METHOD FOR
SAMPLING AND ANALYSIS OF ORGANIC WATER POLLUTANTS
A procedure has been developed whereby the organic com-
ponents of a drinking or surface water are adsorbed onto a
Tenax-packed column. The very volatile organics and the
less volatile compounds are then removed sequentially by
heat desorption and extraction. An outline of the procedure
is given in Figure A-l.
A.I Experimental Details
A.1.1 Preparation
Commercial Tenax GC is exhaustively extracted in a
Soxhlet for 8 hours with methanol followed by a similar
extraction with ether. Five grams of resin are packed
into a 1-inch by 4-inch glass tube and held in place with
glass beads and silanized glass wool. This sampling column
is then washed with methanol and ether to remove any extrac-
table impurities.
Volatile contaminants are removed by passing a stream
of prepurified nitrogen through the column while it is
heated to 200° C for one hour. The results of this purifica-
tion step may be checked by heat-desorption into the GC and
analysis of the vapor. The column is then capped until it
is used.
A.1.2 Sample Collection
Water is sampled by passing it (20 liters or more for a
part per billion analysis), through the column at a flow rate
of from 50 to 100 cm^ per minute. The water touches only
stainless steel and glass before entering the column to
avoid any organic contamination. A small, battery-operated
centrifugal pump (Teel) was used to pull surface water
samples through the resin bed. A diagram of the sampling
system is shown in Figure A-2.
A-l
-------
CLEAN ADSORBENT
(ETHER AND METHANOL)
SAMPLE BY ADSORPTION
(40 TO 50 cm /min
HEAT TO DESORB
VOLATILES
(140°C, ONE HOUR)
EXTRACT WITH ETHER
(50 cm3)
CONCENTRATE
(0.1 TO 0.3 cm3)
SMALL
TENAX
TRAP
ANALYSIS ON
CHROMOSORB 101
ANALYSIS ON
SP-2100
Figure A-l. Procedure for analysis of organic water
pollutants.
A-2
-------
SAMPLING COLUMN
1/2" ss
1/4" ss
TUBING -
OUTLET
FLOW METER
-CENTRIFUGAL
PUMP
WATER LEVEL
METAL
IN-LINE FILTER
Figure A-2. Sampling System.
A.1.3 Heat-Desorption
The sample column is then attached to a prepurified
nitrogen or argon line and the heat-desorption apparatus
shown in Figure A-3.
The heat-desorption apparatus consists of a heating
unit which fits tightly over the sampling column, a flask
for collecting condensed water vapor, a condenser to remove
water vapor from the gas stream, and two small Tenax traps
which readsorb the organics from the gas stream. Flow
meters are used to control the flow rate through each of
these traps. The gas stream is directed through the water
to prevent any loss of purgeable organics by absorption.
The smaller Tenax traps have been designed to fit directly
A-3
-------
•« FLOW METER
LARGE
TENAX
COLUMN
TENAX TRAP
CONDENSER
Figure A-3. Heat-desorption apparatus,
A-4
-------
into the injection port of a Perkin-Elmer 3920 gas chroma-
tograph. The use of two or more of these columns allows for
duplicate analyses of the same sample.
Ten minutes after the gas flow is started, when the air
and major portion of free water has been removed from the
sampling column, the heating coil is turned on. Heat-
desorption is continued for one hour at 140° C.
A.1.4 Analysis of Volatile Organics
The small Tenax traps are inserted directly into a GC
injection port and heat-desorbed for 4 minutes at 250° C
onto a 6-foot Chromosorb 101 analytical column. The temp-
erature is then'programmed from 50° C to 250° C at 8°/minute
for the analysis of a water sample.
A.1.5 Analysis of Extractable Organics
After cooling to room temperature, the resin column is
removed from the heat-desorption apparatus and extracted
with 250 cm^ of ethyl ether. This ether solution is concen-
trated to 0.2 cm3 on a Kuderna-Danish evaporator followed by
a mini Snyder column, washed into a vial, and further
evaporated to from 0.1 to 0.3 cm^ under a stream of prepuri-
fied nitrogen.
The GC or GC/MS analysis is carried out using 2 yl to
5 yl of solution injected onto a 12-foot glass SP2100
column. A standard temperature program is from 100° C (for
4 minutes) to 280° C at 4° C cer minute.
A-5
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA 560/7-77-002
2.
3. RECIPIENT'S ACCESSION-NO.
4. TITLE AND SUBTITLE
Accumulation of Organic Pollutants by Solid
Adsorbents
5. REPORT DATE
1/75 - 6/76
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Energy Resources Co. Inc.
185 Alewife Brook Parkway
Cambridge, MA 02138
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
68-01-2925
12. SPONSORING AGENCY NAME AND ADDRESS
Office of Toxic Substances
Environmental Protection Agency
401 M Street, S.W.
Jtfashinaton, D.C. 20460
13. TYPE OF REPORT AND PERIOD COVERED
FINAL
14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
16. ABSTRACT
A number of solid adsorbents have been examined as
accumulators or organic compounds from aqueous systems.
Using Tenax GC, a method was developed to sample and analyze
both volatile and less volatile organics by direct adsorption
from water followed by heat desorption for the volatiles and
extraction for the remaining non-volatile organic. This
system has been used for the analysis of tap water and
several surface water samples.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS C. COS AT I Field/Group
18. DISTRIBUTION STATEMENT
Release Unlimited
19. SECURITY qLASS (This Report)
unlassified
21. NO. OF PAGES
48
20. SECURITY CLASS (This page)
unclassified
22. PRICE
EPA Form 2220-1 (9-73)
------- |