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 •J N c/5 4> 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) ------- |