C^ U A U.S. Environmental Protection Agency Industrial Environmental Research
EPA-600/7-77-044
Office of Research and Development Laboratory . .. .
Research Triangle Park. North Carolina 27711 Apfll 1977
SELECTION AND EVALUATION
OF SORBENT RESINS
FOR THE COLLECTION
OF ORGANIC COMPOUNDS
Interagency
Energy-Environment
Research and Development
Program Report
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into seven series. These seven broad categories
were established to facilitate further development and application of environmental
technology. Elimination of traditional grouping was consciously planned to foster
technology transfer and a maximum interface in related fields. The seven series
are:
1. Envjronmental 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
This report has been assigned to the INTERAGENCY ENERGY-ENVIRONMENT
RESEARCH AND DEVELOPMENT series. Reports in this series result from the effort
funded under the 17-agency Federal Energy/Environment Research and Development
Program. These studies relate to EPA's mission to protect the public health and welfare
from adverse effects of pollutants associated with energy systems. The goal of the
Program is to assure the rapid development of domestic energy supplies in an environ-
mentally-compatible manner by providing the necessary environmental data and
control technology. Investigations include analyses 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 environmental issues.
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 Information
Service, Springfield, Virginia 22161.
-------
EPA-600/7-77-044
April 1977
SELECTION AND EVALUATION
OF SORBENT RESINS
FOR THE COLLECTION
OF ORGANIC COMPOUNDS
by
J. Adams, K. Menzies, and P. Levins
Arthur D. Little, Inc.
20 Acorn Park
Cambridge, Massachusetts 02140
Contract No. 68-02-1332
Task No. 24
Program Element No. EHE623
EPA Task Officer: Larry D. Johnson
Industrial Environmental Research Laboratory
Office of Energy, Minerals, and Industry
Research Triangle Park, N.C. 27711
Prepared for
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Research and Development
Washington. D.C. 20460
-------
ABSTRACT
The report gives results of an experimental program to characterize the
behavior of resins which can be used in the sorbent trap module of a sam-
pling train used for environmental assessment studies. Experimental design
considerations were based on the sorbent canister in the new source assess-
ment sampling system (SAS8) train. Both XAD-2 and Tenax-GC resins were
studied. Investigated compounds represented both a regular homologous
series and compounds of direct interest to shipboard incineration studies.
Two experimental approaches were used: a gas chromatography method using
elution analysis to determine volumetric capacity (Vg) at low pollutant con-
centrations; and a steady state apparatus for frontal analysis to determine
weight capacities of the resins. The studies showed that XAD-2 has a grea-
ter volumetric and weight capacity than Tenax-GC and is, therefore, pre-
ferred for use in the SASS train sorbent canister. A regular relationship
was observed between the capacity of the resin and the volatility of the com-
pounds studied. Under normal SASS train sampling conditions, materials
such as POMs, PCBs, and Agent Orange would be completely retained by
either the XAD-2 or Tenax-GC resin.
11
-------
Table of Contents
SUMMARY 1
I. INTRODUCTION 2
II. BACKGROUND 4
A. Previous Sorbent Studies 4
B. Available Sorbent Resins 5
C. Physical Properties 5
D. Collection Efficiency Studies 14
E. Evaluation of Other Literature
References 18
F. Conclusions and Recommendations ....... 18
III. EVALUATION PROGRAM 19
A. Approach 19
B. Experimental 20
C. Results of GC Experiments, Vg's 26
D. Results of Steady State Challenge
Experiments, Weight Capacities 40
E. Applicability to Agent Orange Ship-
board incineration tests 46
F. Recovery of TCDD 48
IV. CONCLUSIONS AND RECOMMENDATIONS 50
V. REFERENCES 51
Appendix A - Individual Specific Retention
Volume (Vg) data 53
Appendix B - Relative Specific Retention
Volumes (Vg) on Chromosorb 101
Vg Relative to Benzene 58
iii
-------
List of Tables
Table No. Page
1 Physical Properties of Sorbent Media 6
2 Sorbent Trap Breakthrough Experiments 17
3 Specific Retention Volumes (Vg) for XAD-2 29
4 Specific Retention Volumes (Vg) for Tenax-GC .. 30
5 Specific Retention Volumes (Vg) of Benzene
and Hexane on XAD-2 at Three Different
Temperatures 31
6 Relationship of SASS and Modified Method 5
TT i-iS to Specific Retention Volume Data ... 39
7 Speci :ic Retention Volumes of Other Selected
Pollutants 41
8 Steady State Challenge Capacity: XAD-2 42
9 Steady State Challenge Capacity: Tenax-GC ... 43
10 Mathias III - Agent Orange Burn 47
iv
-------
List of Figures
Figure No. Page
1 Tenax-GC 8
2 Chromosorb 102 9
3 XAD-2 10
4 XAD-4 11
5 Differential Scanning Calorimetry Data ... 12
6 Thermogravimetric Analysis Data 13
7 Sorbent Trap Breakthrough Experiment 15
8 GC Elution Profiles on XAD-2 21
9 Variation of Peak Elution Time with Sample
Size 23
10 Sorbent Trap Exposure Apparatus 24
11 Sorbent Traps 25
12 Breakthrough of Tenax-GC Challenged with
n-Decane in Dry Air Stream at 60°C .... 27
13 Temperature Dependence of Specific
Retention Volume for XAD-2 32
14 Temperature Dependence of Specific
Retention Volume for Tenax-GC 33
15 Relationship of Specific Retention Volume
with Boiling Point 35
16 Relationship of Specific Retention Volume
with Boiling Point 36
17 Relative Specific Retention Volumes for
Aromatic Compounds 37
18 Sorbent Resin Capacity Vs. BP
100 mg/cu m Challenge Concentrations ... 44
-------
SUMMARY
An experimental program has been conducted to characterize the behavior
of resins which can be used in the sorbent trap module of a sampling
train used for environmental assessment studies. Experimental design
considerations were based on the sorbent canister in the new source
assessment sampling system (SASS) train. Both XAD-2 and Tenax-GC resins
were studied. Compounds were chosen for investigation which represented
both a regular homologous series and compounds of direct interest to
shipboard incineration studies.
Two experimental approaches were used: one a gas chromatography method
using elution analysis for the determination of volumetric capacity (Vg)
at low pollutant concentrations, and the other a steady state apparatus
for frontal analysis to determine the weight capacities of the resins.
The studies showed that XAD-2 has a greater volumetric and weight capaci-
ty than Tenax-GC and is, therefore, preferred for use in the SASS train
sorbent canister. A regular relationship was observed between the capaci-
ty of the resin and the volatility of the compounds studied.
Under normal SASS train sampling conditions, materials such as polynuclear
aromatic hydrocarbons, polychlorinated biphenyls and Agent Orange would
be completely retained by either the XAD-2 or Tenax-GC resin. However,
at higher pollutant levels the use of XAD-2 with its greater weight
capacity would be necessary. Neither resin efficiently retains volatile
materials such as vinyl chloride monomer.
-------
I. INTRODUCTION
The new source assessment sampling system (SASS) train includes a sorbent
trap module designed to collect volatile materials which pass through the
high efficiency glass fiber filter.^1) The sorbent trap module consists
of a gas conditioner designed to cool the gas stream to 60°C*, followed
by a canister containing a macroreticular organic resin for the collection
of volatile organic compounds and any other species, such as volatile
metals, which may have the correct properties for collection on the resin.
The sorbent canister is followed by a receiver to collect the water and
other liquids which condense and pass through the resin.
Recent studies(2»3) have shown sorbent traps of this type to have good
characteristics for the recovery and analysis of organics from pollution
sources. However, very little systematic quantitative data have been
obtained which are directly applicable to the design conditions of the
SASS train sorbent module. The purpose of this task was to initiate
studies which would characterize the quantitative behavior of the sorbent
trap. In particular, a series of shipboard incinerator tests are to be
conducted, and it was desirable to know the behavior on the sorbent module
of the materials to be burned in these tests. Initial interests in con-
nection with the incinerator tests included the following compounds:
Selecte_ polynuclear aromatic hydrocarbons (POM's)
Selected polychlorinated biphenyls (PCB's)
Vinyl clJLoride monomer
Agent Orange (2,4-D and 2,4,5-T)
2,3,7,8 tetrachlorodibenzo-p-dioxin
At the time this task was initiated (April 1976), Tenax-GC had been
selected as the resin for the sorbent trap, based upon the behavior of
this material from POM collection.&' However, other studies which had
examined a broader range of pollutants (3 fl*' indicated that the resin
XAD-2 would be a better choice of material for the full range of compounds
to be encountered in the environmental assessment studies. In late April
1976, Arthur D. Little, Inc. had prepared a background discussion on some
of the factors affecting the selection of sorbent trap resins for the
collection of organic compounds.'5' The factors presented in that dis-
cussion are important to an understanding of the behavior of these resins
in the collection of organic compounds and in understanding the reasons
for preferring the use of XAD-2 resin. Because that discussion was not
prepared as a report for distribution, it has been incorporated in the
"Background" section of this report. It is hoped that inclusion of this
more general discussion ;rill aid in understanding the various choices
that are available in designing a system for the collection of organic
vapors and the basic mechanisms operating in the use of sorbent traps.
* The design temperature of 60°C at the time of these studies
has since been lowered to 20°C.
-2-
-------
Many factors affect the collection efficiency and capacity of sorbents
for chemical species. Some of these are:
Sorbent surface area
Sorbent pore volume
Sorbent specific adsorptivity
Pollutant vapor pressure (volatility)
Pollutant concentration
Gas flow rate
Sampling temperature
Gas stream moisture
Presence of other pollutants
A complete evaluation of all of these factors was beyond the scope or
time available for this task. Rather than attempt to examine each of
these factors in detail, data from the previous studies were used to
design a program which made it possible to obtain basic data directly
relevant to the incineration tests.
The use of a conventional sampling train for all of the studies required
for this program would have been prohibitively time-consuming and would
also have posed difficult problems in studying hazardous materials. In
addition, from analysis of the physical properties of many of the com-
pounds of interest (e.g., PCB's, Agent Orange), it was not expected to
be able to directly measure their collection efficiency and breakthrough
because of their high boiling points. The program approach was designed
to
1). test the collection behavior of compounds as a
function of some readily measurable parameter,
such as boiling point, and
2). evaluate a rapid screening approach based upon
a gas chromatograph (GC) type experiment.
Studies have been completed for a number of compounds on both the Tenax-
GC and XAD-2 resins. This report presents the research findings using
both the conditions of a conventional sampling train and the GC approach.
-3-
-------
II. BACKGROUND
A. Previous Sorbent Studies
The material in this BACKGROUND section was reported in a preliminary
form in April, 1976.(5' The purpose of this section is to present a
brief discussion of the alternatives in selecting resins for use in
sorbent traps such as those currently in use in modified Method 5
trains and the new SASS train. While Tenax-GC is currently in use in
several such trains, the use of XAD-2 resin has been favored in studies
conducted by Arthur D. Little, Inc. (ADL) except where thermal stability
is an issue. Unfortunately, complete studies on all of the characteris-
tics which should be evaluated for both resins have not yet been done
for either one. The preferences are derived from experience at ADL
with these materials over the last four years and a recent comparative
evaluation of XAD-2 and Tenax-GC for an EPA incinerator performance
evaluation program.' '
The use of such materials as charcoal, silica gel, GC column packings
(silicones on supports), etc. for the collection of trace materials
has been known and practiced for many years. These previous studies
have led to the recognition of serious deficiencies in many of these
materials. While charcoal has tremendous collection efficiency and
capacity, quantita ve recovery for analytical purposes has been poor.
Silica gel is useful in some cases but has serious limitations in humid
environments. Othe : materials show selectivity in collection and have
low capacities.
Dravnieks' ' was probably one of the first to systematically evaluate
macroreticular resins for use in collecting trace ambient pollutants.
In 1972, ADL began the routine use of Chromosorb 102 in the collection
of diesel exhaust organic pollutants for work on the diesel odor prob-
lem. t7' Chromosorb 102 was selected at that time after comparison with
Chromosorb 101, silica gel and charcoal. Chromosorb 102 has since been
used on many similar sampling problems with complete success. Battelle
Columbus Laboratories (BCL) were also beginning to evaluate these resins
in about 1972-73. (Personal Communications) Emphasis at BCL was on
techniques for direct GC interfacing and, thus, thermal desorption.
Tenax-GC was, therefore, selected as a preferred substrate for that
approach. The emphasis at ADL was on solvent desorption methods
(pentane), and the Chromosorb 102 performed well in that regard.
Zlatkis' work on the use of Tenax-GC for these purposes was first
published in 1973.<8)
Thus, there is a fortuitous situation where two different sorbent media
have been selected, each optimized and satisfactory for the particular
studies to which they were applied. In the following sections, an at-
tempt has been made to present the information currently (and conve-
niently) available on a comparative evaluation of the two basically
different types of resins.
-4-
-------
B. Available Sorbent Resins
No attempt has been made to determine all sorbents which may be available
and appropriate for these purposes. The ones considered are those which
others have found useful and, particularly, those which do not have strong
selectivity characteristics. Resins commonly considered are:
Chromosorb 101 Johns-Manville
Chromosorb 102 Johns-Manville
XAD-2 Rohm and Haas
XAD-4 Rohm and Haas
Tenax-GC Enka N.V.
Poropak series Waters Associates
C. Physical Properties
Several basic parameters are important in selecting substrates for gas
phase adsorption. Besides the chemical surface properties, the physical
parameters are:
particle size range and distribution
pore volume mean and distribution
surface area total and distribution
The particle size will affect the pressure drop across the adsorbent bed
and will also determine whether mass transfer from the gas phase to the
particle will be rate limiting and thus affect the collection efficiency.
Within a given adsorbent, the pore volume and surface area are inter-
related. A larger surface area will usually lead to greater equilibrium
adsorption capacity, but the surface must be available within the time
allowed in the bed transport. Thus, adsorbents which have lower surface
areas are sometimes more effective because they have a larger amount of
surface available in large pores, where gas phase diffusion will not be
rate limiting.
Some of the physical properties of these resins are given in Table 1.
Initial purchase costs are also given for comparison, but none of these
reflect final use cost, since each resin must be cleaned before using.
The significance of some of these differences in properties will be
further explained in the next section where performance testing is dis-
cussed.
Whereas most of the previous ADL studies had been done using Chromosorb
102, the XAD resins were evaluated because they were available in a
larger mesh size range and thus could have a lower pressure drop in a
sampling train. It is understood that Chromosorb 102 and XAD-2 are
virtually identical chemically, both being a divinylbenzene cross-linked
polystyrene. Tenax-GC is a polyphenylene oxide.
-5-
-------
Table 1
Sorbent
Chromosorb 101
Chromosorb 102
XAD-2
XAD-4
Tenax-GC
Physical Properties of Sorbent Media
Mesh Size
-
40 - 80a
20 - 50
20 - 50
35 - 60b
Bulk
Density (g/cc)
0.36
0.38
0.38
0.14
BET
Surface Area
approx.
374
350
925
25
(m2/g)
30
Pore
Volume (cc/g)
0.829
0.854
1.145
0.053
Purchase
Price ($/g)
0.24
0.0088
0.013
3.2
a. largest size range available, sold as 60 - 80
b. 60-80 mesh is also available
-------
Based upon the observed mesh size data for these materials, the pressure
drop across a packed bed should increase in the order XAD < Tenax-GC <
Chromosorb.
The density of Tenax-GC is about one-third that of the other resins.
The surface areas of Chromosorb 102 and XAD-2 are about the same, XAD-4
is three times that of XAD-2, and the Tenax-GC surface area is only one-
fourteenth that of XAD-2. Thus, XAD-4 might be expected to have the
highest equilibrium capacity and Tenax-GC the least. The pore volumes
reflect the surface area data.
The distribution, and thus availability, of the surface area is an im-
portant consideration. In Figures 1-4 the surface area distribution
is shown as a function of pore volume respectively for Tenax-GC, Chromo-
sorb 102, XAD-2 and XAD-4. Nearly all of the Tenax-GC surface area is
in very small sized pores, < 40A. Chromosorb 102 and XAD-2 have very
similar patterns with a good portion of their area in 200 - 300 A pores
and the balance in pores < 50A. XAD-4 has no area in large pores, but
all in pores < 90A.
From these data and previous filtration experience, Chromosorb 102 and
XAD-2 would be expected to be the most efficient resins for collection
at the flow rates of the SASS train.
As is shown later in this section, XAD-4 does have the greatest collec-
tion capacity, but it has been difficult to quantitatively recover
material from this resin. Presumably, the small pores greatly increase
the time required for diffusion in solvent extraction methods.
In a source assessment train sampling hot gases, thermal stability of
the resins is an important issue. Tenax-GC is known to have superior
properties in this respect. We have examined Chromosorb 102, XAD-2
and Tenax-GC by thermal analyis methods, obtaining differential scan-
ning calorimetry (DSC) traces and thermogravimetric analysis (TGA)
curves in an air atmosphere. These are shown in Figures 5a-c (DSC)
and 6a-c (TGA).
From the DSC traces the resins appear thermally stable up to
Chromosorb 102 200°C
XAD-2 210°C
Tenax-GC 400°C
The temperatures at which they begin to show a weight loss from thermal
decomposition in air are as follows:
Chromosorb 102 250°C
XAD-2 260°C
Tenax-GC .... 450°C
-7-
-------
+22.5
+27.5
+37.5
+42.5
+47.5
+52.5
+57.5
+62.5
+67.5
+72.5
+77.5
+82.5
+87.5
+92.5
+97.5
+105.0
+115.0
+125.0
+135.0
+145.0
+155.0
+170.0
+190.0
+210.0
+230.0
+250.0
+270.0
+290.0
+325.0
+375.0
+425.0
+475.0
+550.0
% of Maximum Surface Area ( +8.665 CC/G)
Versus Average Pore Diameter, Angstroms
20%
40%
60%
80%
100%
FIGURE 1 INCREMENTAL SURFACE AREA DISTRIBUTION (DESORPTION):
TENAX GC
-8-
-------
% of Maximum Surface Area ( +74.800 CC/G)
Versus Average Pore Diameter, Angstroms
+22.5
+27.5
+32.5
+37.5
+42.5
+47.5
20% 40% 60% 80% 10C
I I M 1 1 | 1 1 1 1 1 I 1 1 | I | I | i i M I I I I I I 1 I 1 | I I 1 1 I I I I 1 1 | 1 ! 1 1 1 1 1 II 1 1 1 1 1 1 1 1 1
+52.5 1
+57.5
+62.5
+67.5
+72.5
+77.5
+82.5
+87.5
+92.5
+97.5
+105.0
_
-^-»
—
—
—
_
—
—
^^^^^^^^m .
+115.0 I
+125.0 U— —
+135.0 ^—
+145.0
+155.0
+170.0
+190.0
+210.0
+230.0
+250.0
+270.0
+290.0
+325.0
+375.0
+425.0
+475.0
+550.0
— •—
__
— •
•
—
p
)%
.
FIGURE 2 INCREMENTAL SURFACE AREA DISTRIBUTION (DESORPTION):
CHROMOSORB 102
-9-
-------
+22.5
+27.5
+32.5
+37.5
+42.5
+47.5
+52.5
+57.5
+62.5
+67.5
+72.5
+77.5
+82.5
+87.5
+92.5
+97.5
+105.0
+115.0
+125.0
+135.0
+145.0
+155.0
+170.0
+190.0
+210.0
+230.0
+250.0
+270.0
+290.0
+325.0
+375.0
+425.0
+475.0
+550.0
% of Maximum Surface Area ( +68.764 CC/G)
Versus Average Pore Diameter, Angstroms
20% 40% 60%
80%
100%
I I I I I I I I I I I I I I I I I 1 | I I I I I I I I I i I I I i I I I I I I I I I I I I I I I I I I I I Nit :l I 1 I I I I I
FIGURE 3 INCREMENTAL SURFACE AREA DISTRIBUTION (DESORPTION):
XAD-2
-10-
-------
% of Maximum Surface Area ( +217.774 CC/G)
Versus Average Pore Diameter, Angstroms
+22.5
+27.5
+32.5
+37.5
+42.5
+47.5
+52.5
+57.5
+62.5
+67.5
+72.5
+77.5
+82.5
+87.5
+92.5
+97.5
+105.0
+115.0
+125.0
+135.0
+145.0
+155.0
+170.0
+190.0
+210.0
+230.0
+250.0
+270.0
+290.0
+325.0
+375.0
+425.0
+475.0
+550.0
20% 40% 60% 80% 1C
llllllllllllllllllllllllllllll^l.lllltllllfllllljllllllllllltll!
•MBIHIMBM
—
•
P
0%
FIGURE 4 INCREMENTAL SURFACE AREA DISTRIBUTION (DESORPTION):
XAD-4
-11-
-------
A. Chromosorb 102
o
X
01
o
•o
01
60°C
I
0 50 100 150 200 250 300 350 400 450 500
T. °C (Corrected For Chromel Alumel Thermocouples)
B. XAD-2
0 50 100 150 200 250 300 350 400 450 500
T. °C (Corrected For Chromel Alumel Thermocouples)
C. Tenax - GC
o
•o
c
01
60° C
I
0 50 luO 150 200 250 300 350 400 450 500
T. °C (Corrected For Chromel Alumel Thermocouples)
FIGURE 5 DIFFERENTIAL SCANNING CALORIMETRY DATA
-12-
-------
A. Chromosorb 102
10
8
6
0>
£ 4
OJ
'3
5 ?
0
,60°(
\
^
S,
\
^
^X
\.
^
0 50 100 150 200 250 300 350 400 450 500
T. °C (Corrected For Chromel Alumel Thermocouples)
B. XAD-2
10
8
| 6
+•?
f 4
2
0
KPC
<*—*
\
\
^
"^
10
8
' 6
0 50 100 150 200 250 300 350 400 450 500
T. °C (Corrected For Chromel Alumel Thermocouples)
C. Tenax — GC
,60°C
I
0 50 100 150 200 250 300 350 400 450 500
T. °C (Corrected For Chromel Alumel Thermocouples)
FIGURE 6 THERMOGRAVIMETRIC ANALYSIS DATA
-13-
-------
Thus, while Tenax-GC clearly has superior thermal stability, each of
the resins appears capable of meeting the 60°C requirement of the SASS
train. *
Each of the purchased resins considered must be cleaned up before use
in sampling trains. The cost of this procedure has been estimated so
that a true cost of ready-to-use resin can be used for comparison. At
ADL, about 1000 cc of resin is cleaned at a time in giant Soxhlet ex-
tractors. After cleanup the resin must be dried. It is estimated that
about 4 labor hours (A cost of $30 per hour has been used for estimating
purposes) are required for Tenax-GC extraction and 4 for drying, packing,
etc. Thus, 1000 cc of Tenax-GC (140 g) costs about $448 for purchase,
$240 for preparation, for a total of $688/140 g or about $500/100 g
(714 cc). XAD-2 cleanup is more extensive and takes about 8 labor hours
for cleanup and 4 for drying. Thus, 1000 cc of XAD-2 (380 g) costs
about $3.34 for purchase, $360 for preparation, for a total of $363/380 g
or about $100/100 g (263 cc).
In the SASS train, a 7 cm dia x 9 cm deep sorbent trap will hold 343 cc.
Therefore, a cost/trap for XAD-2 and Tenax-GC would be about $130 and
$240 respectively. The resins should be reusable in each case, and the
recycle cleanup costs are expected to be the same for both.
D. Collection Efficiency Studies
ADL has a joint EP.. program with TRW^1*) to evaluate efficiencies of in-
cinerators in des-tr •; ing industrial waste. For that program, Chromosorb
102, XAD-2 and -4 ana Tenax-GC were evaluated for their potential use in
the sorbent trap section of a modified Method 5 train. Trap geometry,
sorbent particle size and quantity were studied to find an optimum for
the train. The final trap had to have a minimum pressure drop, good
initial collection efficiency and good capacity for pollutants.
Initial studies were done using diesel exhaust as a challenge. Diesel
exhaust is hot (120 - 150°C), wet, and contains a reasonable level of
test pollutant. These tests were conducted by running the hot exhaust
into a trap held at ambient temperatures with no provision for heating
or cooling. Final studies at lower challenge levels were done with
dilute mixtures of hexane and decane.
The studies were run as breakthrough experiments. The pollutants were
initially measured at the entrance to the trap to determine the challenge
level, and the exit from the trap was then continuously monitored with a
hydrocarbon analyzer. Before discussing the data obtained, it is useful
to define some terms in describing collection efficiency and breakthrough
(or capacity). Figure 7 shows an idealized sorbent bed test experiment.
A typical sorbent trap -will not collect with 100.00% efficiency and thus
there will be a finite initial exit concentration. When the bed has
begun to reach its capacity, the pollutants will start to break through.
-14-
-------
Challenge
Concentration |
Final
Breakthrough
Total
Hydrocarbon
ppm C
Exit
Concentration
Initial (5%)
Breakthrough
L
Time
FIGURE 7 SORBENT TRAP BREAKTHROUGH EXPERIMENT
-15-
-------
The point at which these pollutants equal 5% of the challenge concentra-
tion is frequently taken as the bed capacity for analytical purposes.
Final 100% breakthrough occurs sometime later.
The data obtained in the screening experiments is given in Table 2.
Sample was pulled through the sorbent traps at 28.3 Apm (1.0 cfm) using
a standard RAC control module. For these tests, the sorbent was held
in a piece of standard pipe. Bed dimensions were about 4.0 cm ID x 7 cm
deep in all but one case but were not measured accurately. The sorbent
quantity (which was weighed accurately) gives an index of bed depth.
It was never possible to purify XAD-4 acceptably as indicated by GC
analysis of washings, and it was thus dismissed as a real candidate.
XAD-2 and Chromosorb 102 appear to behave quite similarly in collection
efficiency and capacity. Diesel exhaust contains a wide distribution
of chemicals, including some very low molecular weight species (methane,
formaldehyde, etc.) which are not collected in these traps. The XAD-2
and Chromosorb 102 have an apparent capacity of about 1300 - 300 = 1000
ppm C or 500 mg/cu m for 20 - 25 minutes. This amounts to the collection
of 350 mg trapped from the 707 liters collected over the 25-minute period.
The same amount seems to be collected on either 20 or 40 grams of XAD-2,
so it has a capacity of about 9 mg/g - 18 mg/g (350 yg/40 g - 350 yg/80 g)
of sorbent for diesel. components.
None of the sorbents of interest has any useful capacity for hexane, an
observation report* i by others and observed in the previous diesel ex-
haust studies. The best comparison to Tenax-GC and XAD-2 comes from
the decane experiments. XAD-2 in two runs showed no breakthrough after
three hours of a 180 ppm C challenge. The Tenax-GC trap broke through
after 5-10 minutes.
Based on these experiments, it was felt that XAD-2 represented the best
general choice of sorbent for use in a wide variety of tests.
Particle size did show measurable differences in the trap pressure drop,
but these differences are minor for the most part, except for Chromosorb
102, whose pressure drop was too high for reliable train operation.
To further test the thermal stability of XAD-2 resin for trace organic
analysis, two blank experiments were run. Laboratory air was heated to
100°C and 120°C and pulled through 40 g beds of XAD-2, held at those
same temperatures for three hours at 28 £pm (1 cfm). The sorbent traps
were extracted overnight with pentane in the continuous extractor, the
extracts concentrated to 0.2 mil and aliquots analyzed by GC/FTD. No
species could be observed in the chromatograms, except for a minor sol-
vent impurity.
-16-
-------
TABLE 2
Sorbent Trap Breakthrough Experiments
Run
1
2
3
4
5
6
7
8
9
10
11
12
13
14
Reslna
C102
C102
X2
X2
X2
X2
C102
X2
X4
T
T
X2
X2
T
Amount
-JsL
20
20
20
40
40
40
22
37
33
12
11
34
30
11
[All experiments run at 28
Size AP
Q
(Mesh) In. Hg. Source
48-60
60-80
20-50
20-50
30-50
30-40
48-60
30-40
20-50
35-60
35-60
20-50
30-50
35-60
5.6- 7.4
5.7-11.1
3.5
2.1- 3.0
4.0- 4.5
2.5
4.0
2.3
2.6
3.5
3.1
2.5- 2.8
2.7- 2.8
3.6
Diesel
Diesel
Diesel
Diesel
Diesel
Diesel
C6
C6
C6
C6
C6
CIO
CIO
CIO
.3 Apm (1
Total
cfm) flow rate]
Hydrocarbon Levels (ppm C) 5% Break
Run Time Inlet
(mln) Challenge
55
70
60
135
133
80
40
40
70
20
5
180
180
55
1300
1250
1300
1350
1200
1300
160
140
180
160
160
140
140
200
Initial
265
210
320
270
210
260
60
4
6
115
19
4
4
6
Exit
1 Hr.
_
180
800
390
320
360
-
-
42
-
-
-
2
_
Final
470
180
800
790
640
340
120
82
70
160
103
-
2
88
Time
(min)
20
None
20
25
25
25
0-5
5-10
20
0-5
0-5
None
None
5-10
a C102 = Chromosorb 102, X2 - XAD-2, X4 = XAD-4, T - Tenax GC
b Bed dimensions for all were about 4.0 cm ID x 7 cm deep, except run 3 was 4 cm deep
c Diesel «= diesel exhaust, C6 = hexane in air, CIO = decane in air
-------
E. Evaluation of Other Literature References
A number of papers have been published in the last few years dealing
with methods for sampling and analysis of organic pollutants. The
novel methods involving the macroreticular resins have used both the
styrene (Chromosorb, XAD) and phenyl ether (Tenax-GC) type materials.
Unfortunately, very few papers have dealt with the comparative perfor-
mance of each type of resin. Many of the papers have dealt with re-
search approaches using thermal desorption for GC analysis, and for
these studies Tenax-GC clearly has been a choice material. A more
general examination of these papers, however, shows that several resins
work quite well for the approach being considered in the SASS train.
The paper by Pellizari'9' demonstrates comparable efficiency of the
Chromosorbs and Tenax-GC. The Chromosorb 101 used in that study has
only 25 m2/g. Pellizari'10^ also found similar results in quantitative
thermal desorption from Chromosorb 101 and Tenax-GC. (see page 559 of
article)
Russell^11' has also shown similar behavior in the comparative perfor-
mance of the PoropaV series polymers (equivalent to the Chromosorbs)
and Tenax-GC.
Junk, v1' et_ al^ hs -f published results in the recovery of trace organics
from water at the pym-ppb level using XAD-2 resin.
F, Conclusions and Recommendations
Based upon the information available in April 1976, XAD-2 and Tenax-GC
appeared to be suitable resins to consider using in the sorbent trap of
the SASS train for some applications. The pressure drop in an XAD-2
bed is less than the Tenax-GC, and the final use cost of XAD-2 is about
one-half of that for Tenax-GC. For general use, XAD—2 appears to have
much greater capacity than Tenax-GC. XAD-2 is readily available in
large quantities.
The use of XAD-2 was recommended for the sorbent module of the SASS
train. Further studies were recommended to describe the quantitative
behavior of the resins under sampling conditions.
New data should include information on collection efficiency (at SASS
velocities), capacity and recovery. Tests should be done at trace levels
In streams which realistically simulate process streams.
-18-
-------
III. EVALUATION PROGRAM
A. Approach
As discussed in the introduction, two basically different approaches
have been taken to the collection of data for this program. One in-
volves the straightforward method of collecting pollutants from a gas
stream using the critical components of a sampling train. The other
uses a gas chromatographic method. The first method is time-consuming
and difficult to use with hazardous compounds, especially Agent Orange
and TCDD but is necessary to verify the correlation with the second
method. The GC method allows a more rapid screening, enables one to
work with hazardous compounds in a safer laboratory experiment, and
potentially allows the evolution of a general correlation between col-
lection efficiency and compound volatility.
To properly use sorbent traps for the collection of organic vapors in
sampling trains, such as the SASS train, it is necessary to character-
ize the resins used in the traps for their initial collection efficiency
and their capacity for the compounds being collected. These factors
are affected by the concentration of the organic vapors in the stream
being sampled and by the vapor pressures of the specific compounds.
Collection efficiency is defined in the conventional manner as
(Inlet-Outlet) concentration
Inlet concentration x
for most resins, with enough effectiveness to be of interest, the initial
collection efficiency is almost always 98 - 100%. In the course of sam-
pling, a sorbent trap will lose its efficiency by exceeding the capacity
of the trap. This may happen in two ways.
For streams with a high concentration of organic vapors, the pores of
the resin will become filled and the trap will, in essence, overflow.
This phenomenon may be thought of as a weight capacity. For low organic
vapor concentration streams the capacity, or holding power, of the trap
is exceeded by virtue of the species being stripped out of the trap by
the air being sampled, as in a gas chromatography experiment. This
capacity breakthrough phenomenon is a volumetric (gas) capacity (Vg)»
The retention volume (Vg) obtained from the GC type elution analysis
is equivalent to the same value obtained from a frontal analysis ex-
periment where one uses a steady state challenge concentration. This
similarity has been used by many researchers to study the fundamental
properties of chromatography and is discussed in several references
including those by Purnell^21'22) and Hildebrand.*23'
-19-
-------
It has been shown^13) that the retention time of a chemical on a sor-
hent is directly proportional to the equilibrium adsorption capacity
of the sorbent. Other studies^1**) have shown that retention times
are a regular function of a homologous series of compounds giving
linear relationships with such simple parameters as carbon number.
Based upon the results of these previous studies and the results pre-
sented in the Background (Section II), both the GC and sampling train
experiments were conducted with a series of n-hydrocarbons in order
to determine the basic vapor pressure (boiling point) relationship
with trap capacity and breakthrough. A wider range of compound types
were studied in the GC experiments, including examples of the pollutants
of direct interest for the shipboard incineration studies.
B. Experimental
1. Chromatographic Apparatus
Special stainless steel columns were constructed to hold the resins
for mounting in a conventional gas chromatograph. The dimensions to
be used were arbitrary, but a minimum of resin was used in order to
keep the retention times short. Since the SASS sorbent trap was 9 cm
deep (it is now 7 cm deep), a 9-cm column was constructed from 1/4 in
(0.635 cm) O.D. tut ^, of 0.020 in (0.51 mm) wall thickness. The
column was terminated by a dead volume 1/4 x 1/16 in Swagelok fitting
containing a 2-ym s jinless steel frit to hold the resin in place.
The column was connected to the injector and detector by use of short
lengths of 1/16 in stainless steel tubing. The volume of the column
was 2.50 m£. The column weight was determined before and after each
time it had been packed with either XAD-2 or Tenax-GC so that the
weight of resin was accurately known for each experiment.
A Varian 2100 GC was used equipped with a flame ionization detector.
Helium carrier gas was used and maintained at a flow rate determined
to be 1.12 mi/sec. This rate gives about one-tenth of the linear
velocity obtained in a SASS trap. Higher gas flows extinguished the
flame.
Solutions were prepared in carbon disulfide solvent and injected in
the normal way. The time for the compound being studied to reach
its maximum in the elution profile was recorded as the value of in-
terest. This point of peak maximum corresponds with the 50% break-
through point for a continuous flow stream containing a low concen-
tration of the pollutant. A set of typical elution curves obtained
for octane (Ce) and phenol on XAD-2 are shown in Figure 8.
The volumetric capacity (Vg) in mA/g was calculated from the elution
time data, flow rate and resin quantity as
t x F
V = -F-
g g
-20-
-------
0.2"/min
I
1"
CS2
Inject
a. C8
b. Phenol
FIGURE 8 GC ELUTION PROFILES ON XAD-2
-21-
-------
where tr = retention time in seconds to peak maxima
F = carrier gas flow rate in mfc/sec corrected to STP
273°C, 760 mm Hg)
g = resin (XAD-2 or Tenax-GC) weight
It is necessary to conduct the volumetric capacity studies in a range
where the elution time is independent of sample size. To experimentally
determine this point, a series of injections of octane of varying quantity
either neat or in carbon disulfide, were made on both the XAD-2 and
Tenax-GC columns. The variation of retention time observed for XAD-2
is shown in Figure 9. Any quantity below 0.01 yfc of octane gave identi-
cal times. For Tenax-GC, the corresponding value was about 0.003 yA.
For all of the subsequent studies, 1 y£ of a 0.1% solution of the com-
pound in carbon disulfide solution was injected, or 0.001 y£ of compound.
An arbitrary cut-off time of 4 hours was chosen for termination of any
particular experiment, except in a few cases where observations were
carried out up to 20 hours. Preliminary experiments indicated that the
4 hour time at 135°C was a factor of 5 - 10 times longer than needed to
correspond to the SASS train sampling at 60°C for 4 hours.
2. Steady Stat* Concentration Apparatus
The experiments de .gned to test the weight loading capacity of the
sorbent resins invoiced the continuous sampling of air containing a
known concentration of chemical through sorbent traps and measuring
the trap breakthrough. The equipment used for these experiments is
shown schematically in Figure 10.
Hydrocarbon and particulate free air was fed to the sampling system at
about 100 &pm. A branch in the air line allowed water to be added so
that the effect of moisture in the air could be studied. Test levels
of about 12 - 20% RH were generated. The water was vaporized with the
tube furnace.
A precalibrated syringe drive allowed the addition of controlled amounts
of hydrocarbons to the air stream. Test concentrations ranging from
10 mg/cu m to 1000 mg/cu m were generated.
The XAD-2 and Tenax-GC sorbent resin was held in glass sorbent traps
which had previously been designed for and used on an RAG Method 5
type train. The trap, shown schematically in Figure 11, contained a
volume of approximately 100 m& of resin. Sampling was maintained at
28 £pm (1 cfm) using the control module of the KAC Staksampler.
Breakthrough was determined by monitoring the air concentration levels
using a heated Beckman 402 flame ionization hydrocarbon analyzer.
After a steady state concentration of hydrocarbon had been achieved
in the delivery line, as determined by the HC analyzer, the HC analyzer
was switched to the trap exit and sampling through the trap initiated.
-22-
-------
1.0
0.1
a
E
0)
ti
£
LU
0.01
0.001
\
\
\
\
\
\
\
\
Column XAD-2, 135°C
\
\
\
\
x \
\
X
\
r
\
yy\
H
X
:
(
—
I
—
-
—
_
—
—
—
—
—
—
—
—
—
—
—
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
.10
.09
.08
.07
.06
.05
.04
.03
.02
.01
.009
.008
.007
.006
.005
.004
.003
.002
.001
150
200 250
Peak Retention Time (Seconds)
FIGURE 9 VARIATION OF PEAK ELUTION TIME WITH SAMPLE SIZE
-23-
-------
Syringe
Drive
Constant Temperature Oven
Charcoal
Canister
Tube Furnaces
Hepa Filter
HXh
-FT "H"
j i
TC
Water Reservoir
Vent
Proportional Controller
Peristaltic Pump
Heated Lines
Beckman 402
Hydrocarbon
Analyzer
ITC
I §
^ V
^> ^
^ v«
^> ^
^' ' I
!
Charcoal Canister
5-Way Ball Valve
Valve
Charcoal Canister
Control
Module
Impinger
Train
Condenser
(Optional)
X
^*!£
i
-
• 1
8 SS
%j$l Sorbent N
^ Tranc JC
1
%%2*2
\bove)
i ^
1
0000$
Vent
FIGURE 10 SORBENT TRAP EXPOSURE APPARATUS
-------
NJ
Ln
~ 1-11/16" or 45 mm
Gas Flow
40 EC Glass Frit
Glass Wool Plug
FIGURE 11 SORBENT TRAPS
-------
Although the system was designed for the study of two traps simultaneously,
we found that it was only possible to keep track of one trap at a time.
Sampling was continued either until the trap had reached 100% breakthrough
or for four hours, whichever came first. A typical curve obtained from
the data for a 100 mg/cu m challenge of n-decane on Tenax-GC is shown
in Figure 12.
A 4-hour sampling period with these traps is about equivalent to the
same time period with the SASS trap. Although the SASS train samples
at 5 cfm (vs 1 cfm here), the SASS train sorbent trap contains 4.5 times
as much resin.
For these experiments, the sorbent trap was maintained at 60°C, the
original SASS train sorbent trap design temperature. (As indicated
earlier, the sorbent trap operating temperature has been lowered to
20°C.)
C. Results of GC Experiments, Vg's
Experimental data have been obtained on a number of compounds. The
compounds were chosen partly to represent a smooth homologous series
for examination of vapor pressure relationships and partly to reflect
those compounds of direct interest to the incineration studies.
Compounds studied extensively were:
Compound Molecular Formula Text Code
n-Hexane
n-Octane ^8^18
n-Decane C j ()H2 2
n-Dodecane Cl2H2g ^12
n-Tridecane C13H28 Cl3
Benzene CgHg
Naphthalene C10H8
Dichlorobenzene CgH^C
Phenol C6H5OH C6H5OH
Aniline C6H5NH2 C6H5NH2
Limited data were obtained
on:
Agent Orange 2,4-D + 2,4,5-T AO
Aroclor 1242
Pyrene ^16^10 ~™
Vinyl chloride monomer C2H3C1 VCM
-26-
-------
150
0)
c
CO
.c
4-1
o
Q.
o
8
I
O)
a.
CD
100
50
0.2
FIGURE 12
0.4 0.6 0.8
Sampled Volume (m3) at STP
1.0
BREAKTHROUGH OF TENAX GC CHALLENGED
WITH n-DECANE IN DRY AIR STREAM AT 60°C
-27-
-------
The paraffins were chosen for the systematic vapor pressure relation-
ship study. Benzene, naphthalene and pyrene represent a start into
the POM's. Dichlorobenzene and Aroclor 1242 represent the PCB's.
Phenol and aniline were chosen to examine compound polarity effects.
Agent Orange and vinyl chloride monomer are species to be burned in
incinerator tests.
For the first set of compounds, the GC Vg determination tests were each
run as a set of four replicates. A fresh column was packed for each set
of determinations. The data from the individual experiments are given
in Appendix A. Thus, the data obtained represent a reliable estimate
of Vg and provide a measure of the variability which can be expected.
The specific retention volume data are summarized for XAD-2 and Tenax-
GC in Tables 3 and 4, respectively. On the average, the Vg data have
about a 10% relative standard deviation.
Experiments were run at 135° and 96°C in order to allow an extrapolation
to the SASS train operating temperature of 60°C using the conventional
Arrhenius relationship
Vg = ae~Ea/RT, log Vg = a1 - ^a
RT
Direct observation at 60°C took too long for most of these compounds,
but it was possible ^ obtain data at 60°C, 96°C and 135°C for hexane
and benzene. These data are given in Table 5.
The data from Tables 3-5 are plotted in Figures 13 and 14 for XAD-2
and Tenax-GC, respectively. Most of the compounds on these plots only
have two points to establish the extrapolation to 60°C. To demonstrate
the validity of this approach, the data from replicate points for ben-
zene and hexane on XAD-2 in Table 3 were used to establish the lines for
these compounds. The additional data given in Table 5 were then plotted
on Figure 13 after the lines had been plotted. These data demonstrate
the linear relationships between Vg and reciprocal temperature.
For XAD-2 (Figure 13) most of the compound types show similar heats of
adsorption (slope). The low molecular weight benzene and hexane show
a higher heat of adsorption (smaller slope) which may have to do with
their size and diffusion into the smaller pores of the resin. A range
of heats of adsorption seem apparent for Tenax-GC (Figure 14). As for
XAD-2, benzene shows the highest heat of adsorption. The paraffins
(€9, CIQ) show the lowest heat of adsorption, while the polar and aro-
matic compounds show an intermediate heat of adsorption. These heats
of adsorption differences between compounds of different polarity might
have been expected for the medium polarity Tenax-GC, a polyphenylether.
However, one might also have expected comparable heats of adsorption
differences between polar and nonpolar compounds on the low polarity
XAD-2, a cross-linked polystyrene. The fact that many compounds exhibit
-28-
-------
Table 3
Specific Retention Volumes (Vg) for XAD-2
(average of replicates)
Compound BP(°C) 96°C 13JL°C
Hexane , C,
D
Benzene ,
Octane, C
Benzene , C,H£
o o
g
Decane, CL.
Phenol, C,HCOH
D J
Aniline, C,HCNH0
o 5 2.
Dichlorobenzene, CgHifCl2
Naphthalene, C-..H,.
69
80
126
174
182
184
180
218
Vg
252
273
2930
—
4990
5460
9850*
—
Std
dev.
54
42
335
—
645
565
—
—
login
,* J-U
Vg
2.40
2.44
3.47
—
3.70
3.74
3.99
—
Vg
43
49
271
1840
477
566
924
3860
Std.
dev.
6
5
33
200
53
72
108
467
log 0
„ iu
Vg
1.63
1.69
2.43
3.26
2.68
2.75
2.96
3.59
*
single value
Vg = specific retention volume of carrier gas in units
of m£/g of resin.
Average weight of XAD-2 per column was 0.734 + 0.017 g.
-29-
-------
Table 4
Specific Retention Volumes (Vg) for Tenax-GC
(average of replicates)
96°C
Compound
Hexane, C,
o
Benzene, C,H,
o o
Octane, Cg
Decane, CIQ
Dodecane, GI~
Tridecane, C.,
Phenol, C,HCOH
o _>
Aniline, C,HCNH0
o j i
Dichlorobenzene,
Naphthalene, C-
BP(°C)
Std.
dev.
69
80
126
174
216
235
182
184
180
218
—
156
659
6500
—
—
3080
4380
6020
__
—
21
38
329
—
—
98
179
134
__
—
2.19
2.82
3.81
—
—
3.49
3.64
3.78
_.
135 °C
vg_
13
27
56
308
1540
3330
298
418
498
1870
Std.
dev.
3
3
8
28
264
600
15
28
34
192
^10
1.11
1.43
1.75
2.49
3.19
3.52
2.47
2.62
2.70
3.27
Vg = specific retention volume of carrier gas in mA/g of resin
Average weight of Tenax GC per column was 0.321 + 0.027 g.
-30-
-------
Table 5
Specific Retention Volumes (Vg) of Benzene and
Hexane on XAD-2 at Three Different Temperatures
Compound 60°C 96°C 135 °C
Hexane, Cg Vg 2100 211 38
3.32 2.32 1-58
Benzene, C6H6 Vg 2390 255 47
log10Vg 3.38 2.41 1.67
-31-
-------
6.0
5.0
4.0
oi
O)
o
3.0
2.0
1.0
Pyrene
Aroclor
\
\
\ Range of
^\- Agent Orange
' Estimates
/
/
C10H8
Ho
CrH4CI2
C6ri5NH2
C6H5OH
C6H5
40°C 30°C
I I I
2.2
2.5
1/Tx103(°K*1)
3.0
3.4
FIGURE 13 TEMPERATURE DEPENDENCE OF SPECIFIC
RETENTION VOLUME FOR XAD-2
-32-
-------
E 4.0 -
01
o
O)
o
\ Range of
) Agent Orange
2.2
2.5
1/Tx103(°K-1)
3.0
3.4
FIGURE 14 TEMPERATURE DEPENDENCE OF SPECIFIC
RETENTION VOLUME FOR TENAX GC
-33-
-------
the same heats of adsorption dependence on XAD-2 is useful in establishing
a Vg-vapor pressure relationship and an elevated temperature screening
test.
The Vg data obtained at 135°C on XAD-2 and Tenax-GC are shown plotted
against boiling point (vapor pressures at 760 mm Hg) in Figures 15 and
16, respectively. On XAD-2 (Figure 15) a good linear relationship be-
tween Vg and boiling point is seen for the n-paraffins. Each of the
other compounds falls somewhat below the line established by the paraffins,
On Tenax GC (Figure 16), all of the data fall on a smooth curve, and most
of it fits on a straight line. However, even though the apparent re-
lationship of Vg to BP is simpler for Tenax GC, the specific retention
volumes for each compound are greater on XAD-2 than on Tenax-GC. This
is true even for phenol which shows the greatest departure from the paraf-
fin line in Figure 15.
One purpose of examining these Vg-BP relationships was to look for the
basis of establishing a simple rapid screening test as a measure of the
retention of a given compound based upon limited data. Clearly, more
experimental results will have to be obtained before these relationships
can be completely defined on a quantitative basis. The basis for such
a relationship does indeed exist, as demonstrated by these data and par-
ticularly by the results obtained by Kiselev^11*' and his co-workers.
Kiselev recently published an extensive series of Vg values obtained on
several r- .ins, including Chromosorb 101. Chromosorb 101 is chemically
identical to XAD-2 (and Chromosorb 102) and differs primarily in that
its surfac3 area is only about 30 m2/g vs. the 350 m*/g surface area
for XAD-2. That data nicely support the Vg-BP relationship and are
based on the study of a large number of compounds and have, therefore,
been examined in greater detail for this report.
The Chromosorb 101 data itself from the Kiselev paper are given in Ap-
pendix B. The logarithm of relative specific retention volume Vg/VgCgHe
(all Vg data were expressed relative to Vg benzene equal to 1.00) vs.
boiling point of the compounds have been plotted versus boiling point
in Figure 17. An excellent linear relationship with boiling point is
observed, encompassing all the data within a single line. The fur-
thest outlier in the data is hydroquinone, which is a very polar com-
pound. A similar relationship is to be expected for XAD-2. We may
expect two to four sets of lines for XAD-2 for the grossly different
polarity classes of compounds, where the major difference in the sets
would be the intercepts of the lines.
The other purpose of establishing the Vg-BP relationships is to enable
the prediction of the behavior of compounds for which Vg data are not
available. This is especially pertinent for Agent Orange and other
compounds which take so long to elute from the resin that it is diffi-
cult to determine their Vg value. It also eliminates the need to work
with highly toxic compounds such as TCDD.
-34-
-------
6.0
5.0
J. 4.0
o
oT
o
3.0
2.0
1.0
Agent Orange: 2,4-D 2,4,5-T
»
70 100
200
300
BP, °C
FIGURE 15 RELATIONSHIP OF SPECIFIC RETENTION VOLUME
WITH BOILING POINT: XAD-2 Vg DATA AT 135°C
-35-
-------
Agent Orange: 2,4-D
70 100
200
300
BP,°C
FIGURE 16 RELATIONSHIP OF SPECIFIC RETENTION
VOLUME WITH BOILING POINT: TENAX GC Vg
DATA AT 135°C
-36-
-------
2.0
Ol
CC
O
01
3
1.0
n
A
o
t
Benzenes
Chlorobenzenes
Bromo, lodo benzenes
Phenols & Chlorophenols
Nitrobenzenes
Aniline
100
300
FIGURE 17
200
BP, °C
RELATIVE SPECIFIC RETENTION VOLUMES (Vg/VgC6H6) FOR AROMATIC
COMPOUNDS ON CHROMOSORB 101 VS. BOILING POINT.Vg DATA AT 190°C
(GVOSDOVICH, KISELEV AND YASHIU)
400
-------
In order to extrapolate the relevance of these data to the Agent Orange
species, 2,4-D and 2,4,5-T, some additional data were obtained. Using
a separate calibrated boiling point GC column and temperature program
(per ASTM procedures^15-!), the retention times and, therefore, boiling
points of 2,4-D and 2,4,5-T were obtained. These values were:
2,4-D (ethylester) 317°C
2,4,5-T (ethylester) 356°C
These points are indicated on the BP axis of Figures 15 and 16. In
order to estimate the Vg's for XAD-2, the line for the paraffins
(Figure 15) was extended and a second lower limit was estimated and
drawn in for polar compounds. The Vg range for the two primary species
in Agent Orange can then be estimated from these lines, and similarly
for Tenax-GC in Figure 16. The estimates obtained are
(at 135°C) on
XAD-2 Tenax-GC
2,4-D 4.2-5.4 4.9
2,4,5-T 4.6-6.0 5.5
These Vg values are in turn plotted on the Vg - 1/T curves in Figures
13 and 14.
In order to relate these data to retention in the SASS or modified
Method 5 -rain, we must have the comparable Vg data for these trains.
Each train holds a certain amount of resin in the sorbent traps and
pulls a certain volume of gas in a 1 or 4 hour period. From these
data one can calculate the value of Vg (or logi0Vg) which a compound
must have to be completely retained by the trains. Table 6 contains
the relevant data for each sampling train and resin combination. The
Vg (LogigVg) values calculated are lower limits. The compounds must
have a Vg equal to or greater than this value (at 60°C) in order to be
completely retained by the resin. These Vg limit values for the 4 hour
sampling period have been plotted in Figures 13 and 14. Using a slope
for these values most similar to the non-paraffins, the limit values
can be extrapolated to 135°C. The stippled line generated by this
exercise represents the bounds of the SASS train collection efficiency,
for low input level challenge concentrations.
For XAD-2, compounds boiling above CIQ would be completely retained in
the SASS train, while those boiling below may be partially or completely
lost. For Tenax-GC, the compounds must have a volatility lower than
CIB. Clearly for both resins, each of the Agent Orange species would
be completely ictained, in terms of volumetric breakthrough. As dis-
cussed later, a compound could always exceed the weight capacity re-
gardless of its Vg if a large enough quantity is collected. The
-38-
-------
Table 6
Relationship of SASS and Modified Method 5 Trains
to Specific Retention Volume (Vg) Data
XAD-2
TENAX-GC
SASS
141 fcpm (5 cfm)
Modified Method 5
28.3 Apm (1 cfm)
Sampling
Time (Hrs)
4
1
5 4
1
Sampling
Volume (M3)
34
8.5
6.8
1.7
Sorbent Trap
Volume (mfc)
445
100
Resin
Cap(g)
130
29
Breakthrough
Vg(m*/g) log10Vg
260,000 5.42
65,000 4.81
234,000 5.37
59,000 4.77
Resin Breakthrough
Cap (g) Vg log1QVg
57 596,000 5.78
149,000 5.17
13 523,000 5.72
130,000 5.12
density XAD-2
density Tenax-GC
0.293 g/rafc
0.128 g/rnfc
-------
less volatile TCDD would be retained even longer than the primary Agent
Orange species.
In order to conclude this preliminary study, Vg experiments were con-
ducted (Table 7) directly with Agent Orange, POMfs (pyrene) and PCB's
(Aroclor 1242) on XAD-2 at 135°C. Neither the Aroclor nor pyrene showed
breakthrough after about 16 hours. Agent Orange showed an indication
of breakthrough corresponding to the reported Vg. This value should
be related to the 2,4-D component and the Vg value is shown plotted in
Figures 13 and 15. The observed approximate value falls just in the
middle of the projected range of Vg's.
Vinyl chloride monomer was studied at 60°C because of its volatility.
As expected, vinyl chloride monomer broke through the trap rapidly,
and the sorbent trap is not an acceptable means to collect this material,
D. Results of Steady State Challenge Experiment, Weight Capacities
The primary purpose of the steady state challenge experiments was to
obtain capacity data in a manner as close as possible to that which
would represent the SASS train sorbent trap operating conditions. In
addition to volumetric capacity data, these experiments made it possible
to obtain weight capacity data from an overloading of the trap.
The data "u .lined on XAD-2 and Tenax-GC are given in Tables 8 and 9.
Experiments were done at a low concentration level, 10 mg/cu m, to
obtain tho volumetric capacity data and at higher levels for the weight
capacity data. Most weight capacity experiments were run at a 100 mg/cu m
challenge. However, some additional experiments were run at higher
levels of 500 and 1000 mg/cu m either to test the effect of input con-
centration on capacity or to obtain breakthrough for those compounds
(hexadecane) which did not break through at a lower challenge level.
The volumetric breakthrough point was taken as the volume when the
breakthrough had reached 50% of the input challenge level. The weight
capacity was taken at that same point by calculating the amount of
hydrocarbon that had been removed from the air stream. The decane ex-
periments on both resins show the transition from a volumetric break-
through to a capacity breakthrough as one goes from an input level of
10 mg/cu m to 100 mg/cu m.
The data show that Tenax-GC has a low weight capacity and even for the
least volatile compound studied (Cie) may not be adequate for many
sampling situations. The capacity for XAD-2 becomes quite good at Ci3
and appears to retain material reasonably well down to a Cj Q volatility.
The experiments on both Tenax-GC and XAD-2 show that the weight capacity
is a function of input challenge level.
The weight capacity data obtained from the higher concentration challenge
levels in these tables is plotted in Figure 18 vs. boiling point of the
-40-
-------
Table 7
Specific Retention Volumes (Vg) of
Other Selected Pollutants on XAD-2
Compound
Aroclor 1242
Pyrene
Agent Orange 2,4-D
2,4,5-T
Vinylchlorlde monomer
BP(°C)
-
393
317
356
-14
Column
Temp.(°C)
135
135
135
60
Vg*
> 64,800
> 64,800
-v 60,500
* 60,500
62
4.82
1.79
'each value is from a single experiment
-41-
-------
Table 8
Steady State Challenge Capacity; XAD-2
Hydrocarbon Condition Cone, (mg/tn )
Weight
Capacity (mg/g)
n-octane
n-octane
n-octane
n-octane
n-decane
n-decane
n-decane
n-decane
n-decane
n-decane
n-tridecane
n-tridecane
n-hexadecane
dry
dry
dry
dry
dry
wet1
dry
dry
dry
wet
dry
dry
dry
9.2
10.2
11.4
26.2
10.9
10.6
11.6
20.4
107
102
110
510
1040
±u —
4.13
4.56
4.18
4.44
4.99
5.10
5.09
4.73
4.85
4.95
5.11
0.09
0.25
0.15
0.57
0.49
0.52
0.29
>2.8
3.5
6.5
43
125
1. 16.7% RH
2. 13.4% RH
-42-
-------
Table 9
Steady State Challenge Capacity; Tenax-GC
Weight
Hydrocarbon Condition Cone, (mg/m3) Log10Vg(m£/g) Capacity (mg/g)
n-octane dry 9.2 3.97 0.08
n-decane dry 10.2 4.56 0.21
n-decane dry 99.5 4.17 1.36
n-tridecane
n-tridecane
n-hexadecane
n-hexadecane
dry
wet1
dry
wet2
104
110
99.6
96.7
4.67
4.59
4.84
4.90
4.2
3.5
5.3
7.0
n-hexadecane dry 491 4.48 12
1. 13% R.H.
2. 20% R.H.
-43-
-------
O)
150
140
130
120
110
100
90
80
u
a
§ 70
60
50
40
30
20
10
0
100
• =XAD-2
O Tenax GC
(510mg/m3
Challenge)
Dry
(1,000mg/m3
Challenge) *Dry
(491 mg/m^
Challenge)
C1Q 200
C13 C16 300
BP.°C
400
FIGURE 18 SORBENT RESIN CAPACITY VS. BP 100 mg/cu m
CHALLENGE CONCENTRATION
-44-
-------
compounds. A linear relationship appears to exist at least for the Tenax-
GC where there is sufficient data at a single concentration. Comparable
data at a single concentration does not exist for XAD-2 because break-
through did not occur. Estimates may be made from the higher concen-
tration level data for this resin. The boiling points of the Agent
Orange species 2,4-D and 2,4,5-T have been indicated on this chart in
order to extrapolate what the capacities of the resins might be for
these species. Tenax-GC calculates to have a capacity of about 8 mg/g
while XAD-2 projects to a capacity of greater than 100 mg/g.
Humidity does not appear to have any deleterious effect on the collec-
tion capacity. In fact it may improve the capacity. For Tenax-GC,
the tridecane capacity was decreased by water, but the hexadecane
capacity was increased. In the case of XAD-2 the capacity was in-
creased in both cases (10 and 100 mg/cu m) in the presence of water.
This effect might be rationalized in terms of the hydrocarbons having
a higher affinity for the organic resins than the wet (now polar) air
stream. The data do indicate that there is no significant interference
in the physical adsorption process such as blocking of the pores. In
a study on Tenax-GC published by Janak and coworkers'16' in 1974, they
found no measurable effect of water on the retention of several com-
pounds, even when using 100% relative humidity.
The experiments run on XAD-2 (Table 8) with both 10 and 20 mg/cu m
challenge levels give similar Vg values indicating that Vg values
taken from the 10 mg/cu m experiments are in the Henry's law region
and provide a basis for comparison with the GC derived data. The data
from both experiments available for direct comparison are given below.
logio Vg (mfc/g)
XAD-2 Tenax-GC
GC_ Steady State GC_ Steady State
Octane 4.65 4.2 4.05 4.0
Decane (5.4)* 5.0 5.3 4.5
* extrapolated
Considering the marked difference in the experiments, the agreement
between the Vg's derived from these two approaches is very good. On
the average, the Vg's from the GC experiment are higher than those
from the steady state apparatus. These data suggest that an average
value of about 0.5 logio Vg should be subtracted from the GC derived
Vg's in order to provide a direct comparison with SASS or modified
Method 5 train conditions.
-45-
-------
Although many workers have found a constancy of Vg with gas velocity'17»1«" t
the tenfold velocity difference between the GC and steady state experi-
ments reported here may be the basis of the difference in Vg's between
the two experiments. The GC experiment with its lower velocity has a
greater likelihood of achieving true equilibrium. More recent studies
by others^19*20) have shown some dependence of Vg on gas stream velocity.
The effects have been attributed primarily to pore volume diffusion and
gas-solid (or surface) interactions.
E. Applicability to Agent Orange Shipboard Incineration Tests
Part of the purpose of these experiments has been to relate the perfor-
mance of the collection efficiency of the sorbent traps to the sampling
associated with the Agent Orange shipboard incineration tests. Some
relevant parameters for the Mathias III are
waste feed rate 65 metric tons/hour
exit gas flow 76.4 cu m/sec
exit velocity 0.5 m/sec
exit temperature 1200°C
excess air 19%
Table 10 h. ws some values calculated from these data for Agent Orange
containing 50/50 2,4-D and 2,4,5-T and 3 ppm of TCDD. Assuming the des-
truction efficiencies indicated, then the data following in the table
would be expected for emission rates. If the gas containing those con-
centrations were sampled by a Method 5 train or the SASS train under
the conditions indicated, then the indicated quantities of material would
be collected.
Collection efficiency for TCDD should not be a problem. However, very
low quantities are expected, and there may be some problems in quanti-
tative recovery for analysis. The amount of TCDD available could range
anywhere from 1 to 240 ug.
The complete collection of the primary Agent Orange species could be a
problem. If the incinerator is only operating at 99% efficiency, there
could be 82 g of Agent Orange, plus other products, in the SASS train
sorbent trap.
We do not know what the maximum resin weight capacities might be for
Agent Orange at high challenge concentrations, but we can arrive at
an upper and low limiting estimate. The steady state experiments shown
in Figure 18 give a lower level estimate of about 10 mg/g for Tenax-GC
and 100 mg/g fox XAD-2 for the equilibrium capacity at low concentrations.
An upper limit estimate of the trap capacity can be calculated assuming
that, due to the low volatility of Agent Orange, the capacity is equal
to the pore volume of the resin based on condensation and filling of the
-46-
-------
Table 10
MATHIAS III - AGENT ORANGE BURN
2. 4-D + 2. 4. 5-T TCDD
Destruction Efficiency
Emission Rate, mg/cu m
Quantities Collected, mg
Method 5, 28 1pm (1 cfm) - 1 hr,
1.7 cu m
SASS, 142 1pm (5 cfm) - 1 hr,
8.5 cu m
SASS, 142 1pm (5 cfm) - 4 hr,
34 cu m
99%
2,400
4,100
20,000
82,000
99.9%
240
410
2,000
8,200
99.99%
24
41
204
820
99%
0.007
0.012
0.060
0.24
99.9%
0.0007
0.0012
0.006
0.024
-------
pores. The pore volumes are 0.85 cc/g for XAD-2 and 0.053 cc/g for Tenax-
GC. The following estimate can then be arrived at for the SASS train sor-
bent trap capacity for Agent Orange
Agent Orange Trapping Capacity
*
SASS Sorbent Trap Steady State Pore Volume
Resin Quantity (gj^ Lower Limit (g) Upper Limit (g)
57 0.6 3
130 13 110
* Assuming unit density for Agent Orange
By comparison of these values with the data in Table 10, one can see that
Tenax-GC would only be acceptable for the most efficient burn. XAD-2
would have sufficient capacity for all of the conditions listed. Using
XAD-2 in tho. SASS sorbent modules would give a maximum holding capacity
for Agent ±ange of 100 g when sampling from a high effluent concentration.
When sampling at low effluent Agent Orange concentrations, the capacity
would be '-> g. If the destruction efficiency from Agent Orange was 99.99%,
corresponding to a stack concentration of 24 mg/cu m, it would be possible
to sample 540 cu m or for up to 64 hours at the SASS sampling rate of 5 cfm
(142 liters/minute) before exceeding the trap capacity. These values
will be altered as other species compete for the adsorption capacity of
the resin.
F. Recovery of TCDD
Because of its low volatility, collection of tetrachlorodibenzodioxin
(TCDD) on the sorbent resins will not be a problem. However, because
the potential quantities involved in the incineration tests are so small,
quantitative recovery could be a problem. In order to investigate this
issue, XAD-2 traps were spiked with TCDD and its recovery determined.
The potential quantity of TCDD to be collected in the shipboard tests
could range from 1 - 240 yg (Table 10). XAD-2 traps (Figure 11) con-
taining 40 g of XAD-2 were spiked in duplicate with 5 yg and 50 yg of
TCDD. The traps were then extracted overnight with pentane in a continuous
extractor. The pentane solutions were concentrated and analyzed by GC.
The GC studies were done using a Ni-63 electron capture detector and pro-
perly prepared calibration curve. The glass column was a 10% OV-17 on
100/120 Supelcoport, 2.5 ft. x 1/8 i.d., operating at 200°C.
-48-
-------
The results were as follows:
TCDD Added (ug) Recovered (ug) % Recovery
5.0 3.4, 4.4 68,88
50 32,33 64,66
Based upon the average of these results, the recovery of TCDD should be
at least 65%. Recovery might actually be expected to be higher in the
tests where the resin would be loaded with Agent Orange and other species
which would help in co-eluting the TCDD from the resin.
-49-
-------
IV. CONCLUSIONS AND RECOMMENDATIONS
XAD-2 has been shown to be more efficient than Tenax-GC as a collector
for use in the SASS train sorbent trap. The volumetric breakthrough
capacity for low input challenge levels is about three times greater
for XAD-2 than for Tenax-GC. The weight capacity breakthrough is about
ten times greater for XAD-2 than Tenax-GC, as determined from the steady
state challenge experiments at high concentrations.
The volumetric capacity (Vg) of both XAD-2 and Tenax-GC shows a regular
dependence on the volatility (boiling point) of the pollutant. There
is an indication that this relationship may be different for compounds
of different polarity classes.
A SASS train sorbent trap operating for four hours will collect all
materials boiling above 190°C (> CIQ) when using XAD-2 and above 240°C
(* Ci3) when using Tenax-GC. Both materials will efficiently collect
POM's, PCB's and Agent Orange initially, but the XAD-2 will have a much
greater capacity for these materials. Neither material will be adequate
in collecting vinyl ihloride monomer.
The effort associa a with this task has enabled the beginning of only some
of the systematic studies which are needed for a complete quantitative
understanding of tl i sorbent trap behavior. Several areas should con-
tinue to be explored for further understanding. Some of these are:
• Establish volumetric capacity (Vg) - volatility (boiling
point) relationships for various compound classes.
• Develop the quantitative relationship between volumetric
capacity derived from the two different types of experi-
ments: elution (gas chromatography) and frontal (steady
state) analysis.
• Determine the collection efficiency of sorbent traps
for aerosols (formed by cooling in the heat exchanger).
• Determine compound recovery as a function of class type
and concentration.
• Completely characterize resin blanks for possible inter-
ferences in each of the analytical steps used to deter-
mine pollutants collected by the traps.
-50-
-------
V. REFERENCES
1. Technical Manual for Process Sampling Strategies for Organic Materials,
Report No. EPA-600/2-76-122, April 1976.
2. P.W. Jones, R.D. Giammar, P.E. Strup and T.B. Stanford, Environ. Sci.
Technol., 10, 806 (1976).
3. P.L. Levins, D.A. Kendall, A.B. Caragay, G. Leonardos and J.E. Oberholtzer,
SAE Paper 740216 presented at the Automotive Engineering Congress, Detroit,
Michigan, February 26, 1974.
4. Destruction of Chemical Wastes in Commercial Scale Incinerators. Corre-
spondence during July - Sept. on EPA Contract No. 68-01-2966 sponsored by
US EPA Office of Solid Waste Management Programs.
5. Selection of Sorbent Trap Media. Communication to IERL/EPA, April 23, 1976
on EPA Contract No 68-02-2150
6. A. Dravnieks, et^ aJL. Environ. Sci. Technol., 5_, 1220 (1971).
7. Analysis of the Odorous Compound in Diesel Engine Exhaust. Final Report
to CRC and EPA (Contract 68-02-0087). June 1972.
8. Zlatkis, et al^ Chromatographia 6^, 67 (1973).
9. Pellizari et al, Enviro. Sci. Technol., 9_, 552 (1975).
10. Pellizari et^ Ed, Enviro. Sci. Technol. V_, 556 (1975).
11. Russell, Enviro. Sci. Technol. £, 1175 (1975).
12. Junk, et^a^. (J. Chromat. 99, 745 (1974) and other papers (Junk and Svec
principal authors.))
13. a. A.B. Littlewood, "Gas Chromatography" Second Edition, Academic Press,
Inc. London, 1970, p. 33
b. A.V. Kiselev and Y.I. Yashiu, "Gas-Adsorption Chromatography", Plenum
Press, New York, 1969, p. 120.ff.
14. T.N. Gvosdovich, A.V. Kiselev and Y.I. Yashiu, Chromatographia, 6^, 179
(1973).
15. R.J. Leibrand, Hewlett-Packard Applications Laboratory Report 1006,
Avondale, P.a., March, 1966.
-51-
-------
V. References - continued
16. J. Janak, J. Ruzickova and J. Novak, J. Chromatography, 99, 689
(1974).
17. P.E. Porter, C.H. Deal and F.H. Stress, J. Am. Chem. Soc., 78,
2999 (1956).
18. A.J.B. Crulckshank, M.L. Windsor and C.L. Young, Proc. Royal Soc.
A295, 271 (1966).
19. J.E. Oberholtzer and L.B. Rogers, Anal. Chem., 41, 1590 (1969).
20. O.K. Guha, J. Novak and J. Janak, J. Chromatog., 84, 7 (1973).
21. H. Purnell, "Gas Chromatography," John Wiley & Sons, Inc., New
York, 73 (1962).
22. J.H. Purnell, Ed., "Progress in Gas Chromatography," Interscience,
New York, 209 (1968).
23. G. Hildebiand, Response as a Function of Sample Input Profile and
the use of Combination Columns in Gas Chromatography, University
Microfi: ., Ann Arbor (1963).
-52-
-------
Appendix A
Individual Specific Retention Volume (Vg) Data
-53-
-------
Specific Retention Volume (ml/g)
XAD-2
Compound
Hexane
Benzene
Octane
Decane
Phenol
Aniline
Dichloroben »ie
Napthlene
47
56
319
2122
556
643
1080
4535
38
45
249
1663
442
502
834
3502
135°C
Vg
47 47
252 263
1746 1818
449 460
552
883 898
3602 3802
Vg_
43
49
271
1837
477
566
924
3860
S.D,
6.4
4.9
32.7
200
53.4
71.5
108
467
Mass Packing (g) 0.7326 0.7506 0.7113 0.7411
-54-
-------
Specific Retention Volume (nl/g)
XAD-2
96°C
Compound y_g Yjg. S.D.
Hexane 290 213 - - 252 54.4
Benzene 328 258 227 227 273 42.4
Octane 3299 2862 2519 3032 2928 327
Decane > 16,000 -
Phenol 5614 5009 4099 5246 4992 645
Aniline 6019 5461 4889 - 5456 565
Napthalene - 9851 9851
Mass Packing (g) 0.7326 0.7506 0.7113 0.7411
-55-
-------
Specific Retention Volume (ml/g)
TENAX GC
Compound
Hexane
Benzene
Octane
Decane
Dodecane
Tridecane
Phenol
Aniline
Dichlorobenzene
Napthalene
15
26
49
275
1355
2907
289
398
482
1846
135°C
vg
11
25
56
324
1728
315
438
537
2067
-
30
64
324
-
3754
288
-
474
1685
13
27
56
308
1542
3331
297
418
498
1866
S.D.
2.8
2.6
7.5
28.3
264
599
15.3
28.3
34.3
192
Mass Packing (g) 0.3611 0.3064 0.3054 0.3100
-56-
-------
Specific Retention Volume (ml/g)
TENAX GC
96°C
Compound Yj Vg S.D.
Benzene 149 170 130 175 156 20.7
Octane 641 633 703 659 38.3
Decane 6497 6090 6495 6895 6494 329
Phenol 3092 3137 2931 3140 3075 98.5
Aniline 4443 4521 4180 4381 179
Dichlorobenzene 5872 6136 5939 6128 6019 134
Mass Packing (g) 0.3611 0.3064 0.3054 0.3100
-57-
-------
Appendix B
Relative Specific Retention Volumes (Vg) on Chromosorb 101
Vg Relative to Benzene at 190°C
Name
Benzene
Toluene
Ethylbenzene
o-Xylene
m-Xylene
p-Xylene
Propylbenzene
Mesitylene
n-Butylbenzene
Durene
Chlorobenzene
o-Dichloroben2 re
m-Dichlorobenzene
p-Dichlorobenzene
1,2,4-Trichlorobenzene
1,2,4,5-Tetrachlorobenzene
Bromobenzene
m-Dibromobenzene
p-Dibromobenzene
lodobenzene
Phenol
Pyrocatechol
Resorcinol
Hydroquinone
o-Chlorophenol
p-Chloropheno1
BP(°C)
80
111
136
144
139
138
159
165
183
190
132
180
172
174
214
243
155
220
218
188
182
246
276
285
175
217
Vg/Vg(Cg^)
1.00
1.86
3.48
3.94
3.36
3.27
5.26
5.76
8.58
11.75
3.31
10.35
8.41
8.41
18.35
47.7
6.86
26.3
27.4
12.8
7.16
31.2
58.7
43.2
7.9
24.8
log1Qrel Vg
0.000
0.270
0.542
0.595
0.526
0.515
0.721
0.760
0.933
1.070
0.520
1.015
0.925
0.925
1.264
1.679
0.836
1.420
1.438
1.107
0.855
1.494
1.769
1.635
0.898
1.394
-58-
-------
Appendix B - continued
Relative Specific Retention Volumes (Vg) on Chromosorb 101
Vg Relative to Benzene at 190°C
BP(°C) Vg/Vg(C,H.) login rel Vg
D D 1U
2,4,6-Trichlorophenol 246 69.9 1.844
o-Cresol 191 12.25 1.088
m-Cresol 203 14.0 1.146
p-Cresol 202 13.85 1.141
2,3-Dimethylphenol 218 23.5 1.371
2,5-Dimethylphenol 212 19.55 1.291
2,4,6-Trimethylphenol 221 27.7 1.442
2,4,5-Trimethylphenol 231 38.4 1.584
2,3,4-Trimethylphenol ? 46.8 1.670
o-Propylphenol 223 38.7 1.588
Aniline 184 10.90 1.037
Nitrobenzene 211 18.95 1.278
m-Dinitrobenzene 301 112.0 2.049
o-Nitrotoluene 222 22.4 1.350
m-Nitrotoluene 233 27.2 1.435
Source: T.N. Gvosdovich, A.V. Kiselev and Y.I. Yashiu,
Chromatographia 6., 179 (1973).
-59-
-------
TECHNICAL REPORT DATA
(Please read fnurucrions on the reverse before completing}
i. REPORT NO.
EPA-600/7-77-044
2.
3. RECIPIENT'S ACCESSION-NO.
4. TITLE AND SUBTITLE
Selection and Evaluation of Sorbent Resins for
5. REPORT DATE
April 1977
Collection of Organic Compounds
6. PERFORMING ORGANIZATION CODE
7. AUTHORlS)
8. PERFORMING ORGANIZATION REPORT NO.
J. Adams, K. Menzies, and P. Levins
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Arthur D. Little, Inc.
20 Acorn Park
Cambridge, Massachusetts 02140
10. PROGRAM ELEMENT NO.
EHE623
11. CONTRACT/GRANT NO.
68-02-1332, Task 24
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 PERIOD COVERED
Task Final; 6/76-3/77
14. SPONSORING AGENCY CODE
EPA/600/13
is. SUPPLEMENTARY NOTES T£RL-RTP Task Officer for this report is Larry D. Johnson, Mail
Drop 62, 919/549-8411 Ext 2557.
16. ABSTRACT
The report gives results of an experimental program to characterize the
behavior of resins which can be used in the sorbent trap module of a sampling train
used for environmental assessment studies. Experimental design considerations
were based on th° ^orbent canister in the new source assessment sampling system
(SASS) train. BC..I XAD-2 and Tenax-GC resins were studied. Investigated compounds
represented both . regular homologous series and compounds of direct interest to
shipboard inc'nei-aiion studies. Two experimental approaches were used: a gas
chromatograpny method using elution analysis to determine volumetric capacity (Vg)
at low pollutant concentrations: and a steady state apparatus for frontal analysis to
determine weight capacities of the resins. The studies showed that XAD-2 has a
greater volumetric and weight capacity than Tenax-GC and is, therefore, preferred
for use in the SASS train sorbent canister. A regular relationship was observed
between the capacity of the resin and the volatility of the compounds studied. Under
normal SASS train sampling conditions, materials such as POMs, PCBs, and Agent
Orange would be completely retained by either the XAD-2 or Tenax-GC resin.
7.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
Air Pollution
Sampling
Organic Compounds
Polymers
Sorbents
Gas Chromatography
Air Pollution Control
Stationary Sources
Sorbent Resins
Environmental Assess-
ment
SASS Train
13B
14B
07C
07D
11G
13. DISTRIBUTION STATEMENT
Unlimited
19. SECURITY CLASS (ThisReport)
Unclassified
21. NO. OF PAGES
65
20. SECURITY CLASS (Thispage)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
------- |