Evaluation of Polyurethane Foam for Sampling of Pesticides, Polychlorinated Biphenyls and Polychlorinated Naphthalenes in Ambient Air Robert G. Lewis,' Alan R. Brown, and Merrill D. Jackson U S . Environmental Protection Agency, Health Effects Research Laboratory, Research Triangle Park, North Carolina 277 I 1
Polyurethane foam has been evaluated for use in a higkvolume air sampler to collect a broad spectrum of pesticides, polychlorinated biphenyls (PCBs) and polychlorinated naphthalenes (PCNs). The sampler draws air through a glass module equipped with a particulate filter and a polyurethane foam vapor trap at flow rates whlch can be controlled from 100 to 250 L/min. Up to 360 m3 of air can be sampled in a 24-h day, providing theoretical detection limits of less than 0.1 ng/m3 for some indlvidual compounds. Extraction and clean-up methodology for gas chromatographic analysis are uncomplicated. Collection efficiencies have been determlned for several organochlorine and organophosphate pestlcldes, PCBs and PCNs.
During 1970-1972, t h e U.S. Environmental Protection Agency had developed under contract t o Syracuse University Research Corporation (SURC) a high-volume air sampler for pesticides (1-3). Basically, t h e sampler was a modification of the standard Hi-Vol sampler ( 4 ) ,which is widely used by state and federal agencies for collection of airborne particulate matter. T h e modification consisted primarily of replacing the Hi-Vol filter assembly with a glass sampling module and altering the air flow control system to provide for an optimum sampling rate of 280 L/min. T h e sampling module was constructed from commercial glass process pipe (10-cm x 5-cm reducer). A 10-cm glass fiber filter was positioned at t h e opening of t h e module and t h e lower compartment was charged with 180 cm3 of 3-mm glass beads coated with a viscous liquid for entrapment of pesticide vapors. Cottonseed oil was the most efficient trapping medium evaluated, but was found to suffer from several disadvantages i n practical use ( 5 ) . The major disadvantage was that separation of collected pesticides from t h e oil required acetonitrile-hexane partitioning and Florisil chromatography; Le., t h e modified Mills-Onley-Gaither procedure (6). Reports of the successful application of polyurethane (PU) foam for high-volume air sampling of polychlorinated biphenyls (PCBs) and pesticides (7-g), lead to an investigation of this medium for use in the SURC sampler. P U foam offers a distinct advantage over cottonseed oil in that i t is reusable a n d requires little or no post-sampling cleanup. A 5.5-cm X 7.6-cm foam plug can be readily substituted for t h e coated glass beads in t h e lower compartment of t h e SURC sampler head and presents n o greater resistance t o air flow.
EXPERIMENTAL Materials. Polyurethane (PU) foam, open-cell polyether-type, density 0.021 g/cm3, was obtained in 3-in. (7.6-cm) sheet stock from a local upholstery shop. Cylindrical sampling plugs were cut from the stock using 5.5-cm circular metal templates. Extractable impurities were removed prior to use by successive Soxhlet extractions of 24-h each with acetone and n-hexane, respectively. Plugs were allowed to air dry while loosely wrapped in aluminum foil. Periodic checks of pre-extracted blank plugs showed no evidence of contamination. After use, plugs were Soxhlet extracted for 16-24 h with 5% diethyl ether in n-hexane, 1668
ANALYTICAL CHEMISTRY, VOL. 49, NO. 12, OCTOBER 1977
air dried as above, and re-used. Glass fiber filter mat (Gelman Type A) and wool felt (weight 14.9 mg/cm2; thickness 0.56 mm) were cut into 10.2-cm circular disks and pre-extracted for 24-h with 5% diethyl ether in hexane in a Soxhlet extractor. The filters were stored in a clean, air-tight container until used. The solvents were pesticide quality or equivalent. All test pesticides, polychlorinated biphenyls, and polychlorinated naphthalenes (PCNs) were obtained from the Pesticide Repository, U S . Environmental Protection Agency, Research Triangle Park, N.C., and were reported to be 9%100% pure (10). Pesticides used in this study were: 7-BHC (la,2a,3a,4e,5e,6e-hexachlorocyclohexane); aldrin (1,2,3,4,10,10-hexachloro-1,4,4a,5,8,8ahexahydro-exo-1,4-endo-~,8-dimethanonaphthalene); p , p '-DDE [l,l-dichloro-2,2-bis(p-chlorophenyl)ethylene]; p,p'-DDT [ 1,l,l-trichloro-2,2-bis(p-chlorophenyl)ethane]; mirex (1,la,2,23,3a,4,5,5a,5b,6-dodecachlorooctahydro-l,3,4-metheno-2~cyclobuta[cd]pentalene); diazinon [ O,O-diethyl-O-(2-isopropyl4-methyl-6-pyrimidinyl)phosphorothioate];methyl parathion (0,O-dimethyl-O-p-nitrophenol); parathion (0,O-diethyl-O-pnitrophenol); and malathion [O,O-dimethyl-S-(1,2-dicarbethoxyethyl)phosphorodithionate]. Air Sampling Studies. Foam plugs were held under slight compression in the 5-cm (id.) X 7.6-cm lower chamber of the glass sampling module. A 10.2-cm glass fiber or felt filter was clamped over the opening of the upper chamber for prefiltration of the incoming air or introduction of vapors of the test compounds. Four air samplers were operated simultaneously, each of which had been calibrated directly with a Roots rotary positive displacement meter or indirectly with a calibrated venturi device. Samplers were located out-of-doors in a rural, nonagricultural environment and positioned a t least 10 m apart. For penetration studies, a separate 5-cm X 7.6-cm glass pipe was attached to the bottom of the sampling module to contain a back-up plug. In some cases, it was necessary to employ a PU foam plug as a pre-filter to remove airborne interferences. For this purpose, two sampling modules were clamped together at the large (10.2-cm) ends. Retention efficiencies were determined by multiple injections of microliter volumes of test compounds in n-hexane directly into the PU foam sampling plug. After a l-h drying period, the fortified plug was placed in front of a second plug in the sampling system. Ambient air was pumped through the train for 24 h at 225 L/min to determine penetration to the second plug. Airborne particulate matter was excluded by a glass fiber pre-filter. Collection efficiencies were measured by fortifying the glass fiber or felt filter, then pulling ambient air through the filter and, subsequently, the foam plug(s). Hexane solutions of the test compounds were added dropwise to the filters in amounts of 2 mL or less, and the solvent was allowed to evaporate before the filter was attached to the sampling module. After 24 h of air flow, the filter and foam plug(s) were analyzed individually as described below, At least one blank determination with unfortified filters was made simultaneously to correct for airborne interferences and possible contamination or losses from the analytical methodology. Isolation and Clean-up Methods. Foam plugs and filters were removed from the sampling module with forceps and immediately placed into 500-mL and 250-mL Soxhlet extractors, respectively. Both were extracted with 5% diethyl ether in n-hexane (total volumes: 750 mL and 350 mL, respectively) for 16 to 24 h (rate: 4 cycles/h). Extracts were concentrated to ca. 5 mL under vacuum in a rotary evaporator, rinsed into 15-mL graduated centrifuge tubes, and adjusted to 10-mL final volume.
Organophosphate pesticides were determined without cleanup. For organochlorinepesticides, PCBs and PCNs, the extracts were further reduced in volume to 1 mL by careful evaporation under nitrogen a t room temperature. The concentrated solution was then subjected to column chromatographic cleanup on alumina, activity grade IV. Column size was 2-mm X 15-cm. Elution was with 10 mL of n-hexane a t a drop rate of ca. 0.5 mL/min. The eluate was adjusted in volume to 10 mL for gas chromatographic (GC) analysis. Blank values for unused plugs were essentially zero (equivalent t o 95%, except for p,p’-DDE (90%). For the test mixtures of PCBs and PCNs, recoveries were 96 t o 98%. Organophosphorus compounds were retained by the alumina column, necessitating their analysis by GC-FPD and GC-ECD without cleanup. Recoveries of all test compounds and mixtures from unexposed fortified foam were essentially quantitative except for the small losses cited above. Collection efficiencies were corrected for losses from column chromatographic cleanup only in t h e case of p,p’-DDE. Some trials were performed using an additional foam plug for prefiltration of the intake air. However, with this arrangement (described in the Experimental section), air flows could not be increased above 180 L/min without over-loading the Hi-Vol motor. Typical GC-ECD responses to alumina-cleaned extracts of two foam plugs connected in tandem and exposed for 24 h to ambient air a t 225 L / m i n are shown in Figure 2, A and B. T h e last eluting major peak in A is 2.03 min. Chromatogram C is from an unused P U foam plug processed in t h e same manner. In order to determine the ability of P U foam to hold adsorbed pesticides, PCBs and PCNs, microgram quantities of the test compounds in n-hexane were injected into the plug and air was drawn through it into a second plug a t 225 L/min for 24 h. Except for aldrin, 88 to 106% of pesticides were retained by the fortified plug, with only 2 to 6% recovered from the second plug (see Table I). T h e apparently poor retention of aldrin may be the result of partial oxidation to dieldrin. Retention values for Aroclor 1242 and Halowax 1001 were poorer t h a n expected, a t 79% and 5770, respectively. ANALYTICAL CHEMISTRY, VOL. 49, NO. 12, OCTOBER 1977
1869
TABLE 1 1 . COLLECTION E F F I C I E N C I E S O F P O L Y U R E T H A N E FOAM A T 225 lirnin FOR C H L O R I N A T E D PESTICIDES VI A I R C O N C E N T R A T I O N S
-
Sta;isrical
Collecrion Efficiency
Air Conc
( n g m3) BHC
Aldrin
0 15 0 08 0 03
53 2 38 4 55 0
150
58 5
14 0 20 7 2
21 23 1
9 3
90
7
163
104 8 101 0
7 3
24 5 27 6
3 8 7 2
126 134
0 92 0 37
114 7 94 6 93 6 83 0
9 20 0 60 0 30 0 12
103 7 1004 98 7 105 4
6 5
P P 3DE
n p DDT
?,I, rex
90 21 2 1'6 71
3 7
5 6
57
TABLE I l l . L
I
1
>
d
L - L _ L--
I 9
1
17
1
A V E R A G E COLLECTION E F F I C I E N C I E S OF P O L Y U R E T H A N E FOAri' FOR O R G A N O P H O S P H O R U S P E S T I C I D E S A T 2 2 5 1 r i i n AND 184 I min
~
1
Figure 2. Gas chromatograms of polyurethane foam extracts after alumina cleanup. (A) upper trap, 24-h ambient air collection; (6) lower trap, 24-h ambient air collection; (C) unused foam plug. ChromaECD. All chromatograms tographed on 3 % OV-1 at 180 O C with 63N~ made at 10' X 2 attenuation
^ r
Pexi
IP
--
Fj
_r i
i
I
- 2 i
7
-
TABLE I . A V E R A G E RETENTION EFFICIENCIES O F P O L Y U R E T H A N E FOAM FOR PESTICIDES, PCBr AND PCNs A F T E R 24 hr AT 225 I/min AIR F L O W
Pesticides
Fortification Level of Upper Trap (ug)
Remaining n Upper Trap ( O O Retention
j :
% Found i n Lower Trap
Aldrin
1 .oo
61
4
P . P ' DDE
2.00
89
6
p,p'-DDT
6.00
106
3
Mirex
6.00
88
4
86
2
101
4
Parathion
3 00
96
4
Aroclor 1242
2 00
79
7
2 0 00
57
10
Halowax 1001
Penetration t o the second plug was also more significant in the latter cases. Some 33% of Halowax 1001 was totally lost. Collection efficiencies were determined for separate mixtures of organochlorine and organophosphorus pesticides by volatilization from a wool felt filter into a tandem pair of foam plugs. T h e sampler was allowed to operate for 24 h for all collection efficiency studies. Flow rate was 225 L/min, except as noted. Residues averaging ca. 25% of the original application were retained on the felt after 24 h, thereby permitting mass balance calculations. 1670
ANALYTICAL CHEMISTRY, VOL. 49, NO. 12, OCTOBER 1977
Table I1 gives the collection efficiencies for the upper foam trap for several concentrations of a mixture of organochlorine pesticides. As expected, there was little dependency of trapping efficiency on apparent air concentrations in the subnanogram to nanogram per cubic meter range. T h e less volatile members of the series were essentially quantitatively trapped by the first foam plug. About 20% of the aldrin was found in the second trap, which indicated an overall collection efficiency of 65 to 75%. Little or no y B H C was trapped by the second plug, however. The poorer collection efficiencies of the latter two compounds may be due in part to oxidation or isomerization, although no dieldrin (epoxidized aldrin) or other BHC isomers were detected. Collection efficiencies for a mixture of organophosphate pesticides a t three concentration levels are given in Table 111. Measurements a t the two lower concentrations were made at a reduced flow rate (184 L/min) because of the use of a 5.5-cm X 7.6-cm P U foam pre-filter. Use of 11-cm X 7.6-cm foam
TABLE Iv. A V E R A G E COLLECTION E F F I C I E N C I E S FOR PCBs A N 0 P C N r ON P O L Y U R E T H A N E FOAM Total $;'
, x !u r t
mg
I 1 minl
Collection E'ficiency I hI
"13,
Aroclor 1 2 4 2
i
m
I\
A r m or 1 2 4 2
1 40
204
l . o c l o r 1254
225
326
225
326
225
3 26
140
204
225
326
ha1o'::a.v
1001
ralo,'.ax
100'
-alo:.ax
1001
r i a nvuax 1013
'
I 1
7% 98
EO
54 303
61
TABLE V COMPARATIVE COLLECTION EFFICIENCIES FOR AROCLOR 1242 A N D H A L O W A X 1001 O N P O L Y U R E T H A N E F O A M A T 225 l/min
G C Peak
No.
% Collected Efficiency (Normalized) Halowax 1001 b
Aroclor 1 242a
J I
I 2
1
1
1
4
6
8
I I?
17
1
_A
1:
43
41
TINF
Figure 3. Gas chromatograms of Aroclor 1242. (A) standard fortification solution diluted 1OX to 200 pg/pL; (B) residue on felt filter after 24 h at 225 L/min.; (C) residue in upper foam trap after 24 h at 225 L/min. Extracts B and C have been subjected to alumina cleanup. Numbers
58
58
65
56
indicate peaks used for quantification
a4
72
pre-filter would have permitted a flow rate of 225 L/min, but apparatus needed to mount the larger plugs on the sampling module intake was not available for this study. T h e results of the study, however, show no significant change in collection efficiencies with sampling rate. T h e trapping efficiency for diazinon was found to be higher than expected in light of its relatively high volatility. Little or no intact pesticide was found in the second foam trap. Again, oxidation (to the corresponding oxon) may have accounted for less t h a n quantitative recoveries of the organophosphates from the dual plug sampler. No evidence of such conversion was detected, however. Measurements made with two commercial mixtures both of PCBs and PCNs are given in Table IV. All data presented are averages of replicates (2 to 8 independent trials). At 225 L / m i n total 24-h collection efficiencies (single plug) for the PCB mixture averaged from 70 to 85%,while the more volatile PCN mixtures were collected less efficiently a t 44 to 61%. Reducing the air flow rate to 140 L/min appears to have little or no effect on trapping efficiencies. Several experiments were made with each mixture using a second foam trap. In all cases, however, only trace quantities of the mixtures were observed in the second of the two tandem traps. Examination of the GC-ECD chromatograms of the lesser chlorinated mixtures, showed evidence t h a t the more volatile congeners were less efficiently trapped than those with lower vapor pressures. For example, chromatogram C in Figure 3 shows considerable alteration of the relative peak heights in collected Aroclor 1242 compared to the original (chromatogram A). A similar loss of t h e more volatile components is evident in the residue remaining on the fortified felt filter (chromatogram B) after
86
100
80
91
95
92 100 a Normalization factor 1.05;Avg. overall collection efficiency 75% ( n = 1 5 ) ;Absolute retention times 2.4 min 1 1 ) to 6.2 min ( 9 ) .
Normalization factor 1 . 1 1 ; Avg. overall collection efficiency 57% ( n = 8 ) ;Absolute retention times 2.1 min ( 1 ) to 4.5 min 1 6 ) . 24 h a t 225 L/min. Underlying extraneous material collected from the atmosphere further contributed to the alteration of the appearance of the collected Aroclor 1242 in chromatogram C, however. This background was substracted prior to computation of the collection efficiency in this study, which involved introduction of an equivalent of 18.4 ng/m3 of Aroclor 1242. Some of t h e background material has been shown by mass spectrometric analysis to be PCBs already present in the ambient atmosphere. Individual collection efficiencies were calculated for all of the components used for quantification of Aroclor 1242 and Halowax 1001 and are presented in normalized form in Table ANALYTICAL CHEMISTRY, VOL. 49, NO. 12, OCTOBER 1977
1671
V. The less volatile components of both mixtures were found to be trapped with more than twice the efficiencies of the more volatile components. At 225 L/min, collection efficiencies ranged from 41% to 95% for Aroclor 1242 and 37% to 90% for Halowax 1001, averaging 75% (for 15 independent determinations) and 57% (for 8 determinations), respectively. During the course of these studies, weather conditions varied widely. Sampling was conducted in temperatures ranging from -18 OC to +30 OC and during periods of rain and snow, as well as dry, windy conditions. A survey of the data collected shows no measurable effect of weather on sampling efficiency. CONCLUSIONS It is apparent from these studies that while polyether-type polyurethane foam is an efficient high-volume air sampling medium for low levels of chlorinated pesticides and polychlorinated biphenyls of low volatility, it is inadequate or only marginally adequate for 24-h sampling of the more volatile members of these families. The greater polarities of organophosphate pesticides assist in their collection, so that acceptable trapping efficiencies (75%) are feasible for many of these compounds despite their higher vapor pressures. The use of tandem traps does not always improve trapping efficiency, despite logical expectations. This latter finding points out the potential fallacy in using “break-through” data (penetration to a secondary trap) for estimation of sampling efficiencies. I n this case, as for all environmental monitoring, the difficulties inherent in the use of nonspecific GC detectors (such as the ECD) to identify and quantify organic compounds are obvious. A considerable advantage in using the high volume sampler for ambient air monitoring, however, is derived from the fact that sufficient quantities of compounds may be collected to permit unambiguous GC/MS confiimation of results for a single sample.
LITERATURE CITED J. Bjorklund, E. Compton, and G. Zweig, “Development of Methods for Collection and Analysis of Airborne Pesticides”, report prepared under contract CPA 70-15 for the National Air Pollution Control Administration, U.S.Department of Health, Education, and Welfare, Durham, N.C., September 1970. E. Compton, E. Bazydlo, and G. Zweig, “Field Evaluation of Methods of Collection and Analysis of Airborne Pesticides”, report prepared under contract CPA 70-145 for the U S . Environmental Protection Agency, Research Triangle Park, N.C., May 1972 (NTIS No. PB-214008). B. Compton, and J. Bjorklund, Design of a High-volume Sampler for Airborne Pesticide Collection”. paper presented at the 163rd National Meeting of the American Chemical Society, Boston, Mass., April 1972. G. A. Jutze, and K. E. Foster, J . Air Polluf. ControlAssoc., 71, 1, 1967. R. G. Lewis, in “Air Pollution from Pesticides and Agricultural Processes”, R. E. Lee, Jr., Ed., CRC Press, Cleveland, Ohio, 1976, pp 51-94. P. A. Mills, J. H. Onley, and R. A. Gaither, J . Assoc. Off. Anal. Chem., 46, 186 (1963). T. F. Bidleman, and C. E. Oiney, Bull. fnvlron. Contam. Toxlcol., 11. 442 (1974). M. M. Orgill, M. R. Peterson, and G. A. Sehmel, “Some I n M Measurements of DDT Resuspension and Translocation from Pacific Northwest Forests”, Report No. BNWL-SA-5126, Battelle Pacific Northwest Laboratories, Richland, Wash., September 1974. E. C. Turner and D. E. Glotfelty, Anal. Chem., 49, 7 (1977). “Analytical Reference Standards and Supplemental Data for Pesticides and Other Organic Compounds”, J. F. Thompson, Ed., Publication No. EPA-600/9-76-012, U S . Environmental Protection Agency, Research Triangle Park, N.C., May 1976. (a) R. G. Lewis and N. J. Zimmerman, unpublished results, 1976; (b) N. J. Zimmerman, “An Initial Investigation to Determine the Effectiveness of Poiyurebne Foam as a Collecting Medium for Atmospheric Pesticide”, M.S. Thesis, Department of Environmental Sciences and Engineering, University of North Carolina, Chapel Hill, N.C., August 1976. R. G. Lewis and N. J. Zimmerman, Anal. Qual. ConfrolNewsl.,U.S. EPA, No. 28, 7, January 1976. C. E.Rodes, M. D. Jackson, and R. G. Lewis, “Monitoring fw Po)ychlorinated Biphenyl Emissions from an Electrolytic Capacitor Disposal Project”, submitted to J . Air Polluf. Control Assoc., 1977. F. W. Kvtz, A. R. Yobs, and H. S. C. Yang, in “Air Pollution from Pesticides and Agricultural Processes”, R. E. Lee, Jr., Ed., CRC Press, Cleveland, Ohio, 1976, pp 95-136.
RECEIVED for review May 23, 1977. Accepted July 11, 1977. Presented a t the 173rd National Meeting of the American Chemical Society, New Orleans, La., March 25, 1977 (Paper No. 78, Division of Pesticide Chemistry).
Stable Free Radical Reagent and Solid Phase Suitable for a Nitric Oxide Dosimeter J. S. Nadeau and D. G. 8. Boocock* Department of Chemical Engineering and Applied Chemistry, University of Toronto, Toronto, Ontario, Canada M5S 1A4
A new reagent combined with a solid phase support has been developed as a system for a nitric oxide dosimeter. The free radical reagent, 2-phenyld,4,5,5-tetramethylimidarollne 30xide 1-oxy1 Is applied In chloroform solution to silica gel plates. Reaction with nitric oxide produces a second stable radical which may be quantified by ESR. Plates were exposed to 0.05-1.00 ppm nitric oxide in nitrogen for 1 h inside a 1000-L non-rigid chamber. Commonly occurring ambient gases showed no serious interferences. Sufficient sensitivity remains to allow the incorporation of a membrane for the purpose of buffering mass transfer effects due to wind velocity.
I n t h e early 195O’s, it was demonstrated t h a t oxides of nitrogen together with hydrocarbons were the prime ingre1672
ANALYTICAL CHEMISTRY, VOL. 49, NO. 12, OCTOBER 1977
dients for photochemical smog formation ( I , 2 ) . This knowledge focused attention on nitric oxide and nitrogen dioxide. In 1954 Saltzman published a wet chemical technique for the colorimetric determination of nitrogen dioxide ( 3 ) and later the reagent was modified for recording air analyzers ( 4 ) . The method was extended to the determination of nitric oxide by the use of permanganate as oxidant ( 5 ) . Recently, chemiluminescence methods for the two oxides have been widely adopted. In our studies we had need of an analytical system which would allow simultaneous measurement of time weighted concentrations of ambient nitric oxide a t several locations. We were also mindful of the interest in occupational health and dosimeters and thus directed our efforts toward a system which might serve as a dosimeter. Solid phase systems are obviously attractive for small collectors or monitors. During