Document 9yO27VkKknOn0j2kqNz8b6z6

296 WALLACE ET AL. Fig. 1- Tenex cartridge and Dupont pump used in collecting personal air samples. 1). All volatile organics except vinyl chloride were collected using glass tubes 10 cm long and 1.5 cm i.d. containing 1.5 g Tenax-GC adsorbent, using DuPont P-125 pumps at a flow rate sufficient to collect 20-25 liters of air during each of the three periods of the day in the monitoring schedule (Table 2). Samples were analyzed using thermal desorption and purging by helium into a liquid-nitrogencooled nickel capillary cryogenic trap, followed by high resolution glass column gas chromatography-mass spectrometry (GC-MS) techniques (Krost et al., 1982). Identities of sample constituents were established by comparing their mass spectra with those in the National Bureau of Standard (NBS) reference libraries. Concentrations of the target compounds were determined by loading an external standard (hexafluorobenzene or perfluorotoluene) at a known concentration on each sample cartridge before analysis. The absolute peak height of the chemical to be measured was then compared to the peak height or area of the standard, using selected characteristic ions in the mass spectra. These heights or areas are proportional to the number of moles of each substance. The proportionality con stant, called the relative molar response (RMR), is known or can be determined experimentally for any compound. The mass of the chemical on the cartridge can then be determined from the known mass of the standard, the molecular weights of each chemical, the RMR, and the height or area. The calculated mass can then be corrected for recovery efficiency. The average concentration during the time the subject wore the monitor is then the quotient of the calculated mass (corrected for recovery efficiency) divided by the measured volume of air pulled across the URL 03318 EXPOSURE TO VOLATILE ORGANIC COMPOUNDS 297 cartridge, provided that this volume does not exceed the breakthrough volume of the target chemical. Most of the target chemicals had breakthrough volumes well above the 20- 25-liter typical sampling volume. Concentrations of those chemicals whose sample volumes occassionally exceeded their breakthrough vol umes were calculated by dividing the mass by the breakthrough volume rather than the sampling volume. This procedure gives an estimate of the mean concen tration during the latter part of the sampling period (i.e., the time required to sample the breakthrough volume). Vinyl chloride was collected using a permeable membrane (dimethylsilicone rubber) badge attached to the subject's lapel. The activated charcoal element was desorbed by carbon disulfide and analyzed using gas chromatography-electron capture detection techniques. Breath. Breath samples were collected on the second and third day of each trip using a specially designed spirometer (Fig. 2). The subject inhaled pure air from Tedlar Bag A while exhaling into Bag B. Roughly 45 liters of exhaled breath were pumped across duplicate Tenax cartridges, using MSA pumps with gas me ters to measure the volume. The breath samples were then analyzed by the same methods as the air samples. Some breath samples were taken before and after dinner, to test the contribution of food or beverages to breath concentrations. Water. Samples of drinking water from each subject's home and workplace were collected by the subject. A 5-ml aliquot was transferred to a purge device (Bellar and Lichtenburg, 1974) via a glass syringe. The water was purged onto a Tenax GC trap at a flow rate of 40 ml/min. The compounds were then desorbed onto an analytical column (1.8 x 2.0 mm0.2%carbowax 1500 on 60/80 Carbopack C). Both a Hall Electrolytic Conductivity Detector and flame ionization detector were operated simultaneously to detect the halogenated and aromatic compounds. ULTRAPURE AIR TANK o TENAX CC CARTRIDGES TEOLAR BAG A TEDLAR BAG B DOUGLAS VALVE ANO MOUTHPIECE FiC. 2. Schematic diagram of the spirometer. 298 WALLACE ET AL. Food. Because no validated protocols exist for collecting, preparing, storing, and shipping food items from the home, only samples collected from community grocery stores were analyzed. The objective was to determine whether exposures through food were sufficiently important to require the collection of individual food samples-- and the concomitant validation of protocols--in the large-scale study to follow. About 40 food items were collected from a community grocery on four separate occasions--two in New Jersey and two in North Carolina. All foods were selected and prepared according to U.S. Food and Drug Administra tion (FDA) Total Diet Study Adult Market Basket specifications (Entz et al., 1982). Four composites (dairy, meats, fatty foods, and beverages) from each of the four sampling trips were analyzed for volatiles at the FDA laboratory in Washington, D.C. using FDA automated head space equipment and protocols. Portions of each individual food item used in the composites were held in reserve for additional sampling to determine the source of elevated concentrations in the composite. QUALITY ASSURANCE The primary laboratory for the air, breath, and water analyses was^Research Triangle Institute (RTI); and for the food analyses, FDA. All analyses except those for food were repeated at independent laboratories to estimate the uncer tainty of the data. From each batch of 30-50 Tenax cartridges, about 10% were stored at the laboratory (lab blanks) and another 10% were shipped to the field but not exposed (field blanks). These cartridges were then analyzed at the same time as the exposed samples to determine background levels on the Tenax batch and also any contamination that may have occurred during transportation to or from the field. Another 10% of the cartridges were spiked with known amounts of each of the target compounds. Again both laboratory and field spikes were prepared and analyzed with the field samples to determine the loss of the compounds with time and travel. Replicate samples were also obtained for 100% of the breath samples and 10% of the air and water samples. Each sample train was internally audited before, during and after sampling in each geographic area. External audits were also performed by a team from EPA's Environmental Monitoring Systems Laboratory in Research Triangle Park, North Carolina. Flow rates were checked with a bubble meter and battery charges on personal monitors were verified. A chain-of-custody record was kept for each sample. All quality control procedures, including preparation of blank and spiked lab oratory and field samples and duplicate samples for a total of 70 air, water, and breath samples, are documented in the TEAM Study quality assurance (QA) report (Sparacino et al., 1982b). Air. Equipment problems at the designated QA laboratory, the Environmental Monitoring Systems Laboratory at Research Triangle Park, NC (EMSL-RTP), prevented completion of the planned number of 10 QA samples. Four samples were eventually analyzed, but these had been stored longer than the recom mended upper limit of 4 weeks. Reasonable agreement was obtained, but more , fRL 03820 EXPOSURE TO VOLATILE ORGANIC COMPOUNDS 299 samples analyzed in a more timely fashion would be required to estimate interlaboratory precision. Contamination of the Tenax by benzene during prepara tion, storage, shipment or analysis was evident on many air and breath blanks and spikes; thus no quantitative results for benzene in air or breath are presented. Duplicate air analysis for three prevalent target chemicals resulted in typical coefficients of variation of 25-50% (Table 3). The QA laboratory for the vinyl chloride analyses was REAL, Inc. of New Orleans, La. AH 24 field QA samples were below the limit of detection, although spiked, laboratory and field samples indicated that the analysis method was ca pable of detecting levels above 10 ftg/m3. Breath. The QA laboratory was IIT Research Institute of Chicago, 111. Thirteen paired QA breath samples agreed to within 60-70% for chloroform and tetrachloroethylene. Duplicate analyses for three prevalent chemicals agreed to within 25-40% (Table 3). Percentage recoveries ranged from 80 to 100%, with the ex ceptions of chloroform (--65%) and ethylene dichloride (--40-60%) (Table 4). Water (purge and trap). Twenty QA samples of drinking water were analyzed by the Environmental Monitoring Systems Laboratory at Cincinnati, Ohio. QA analyses agreed with the primary lab analyses both qualitatively and quantita tively with average errors of 30%. RESULTS Occurrence and Concentration of Target Chemicals The broad-spectrum analyses identified over 200 chemicals, of which 100 were C to C|5 hydrocarbon isomers. One hundred twenty other chemicals were iden tified in the 8 air samples, 12 breath samples, and 1 water sample collected in New Jersey (Table 5). Of these, 38 appeared in all three media, 23 in breath only, 19 in air only, and 12 in water only. The average number of chemicals identified in nine personal air samples from both locations was 149 (36); in two water samples was 116 (24); and in 13 breath samples was 101 ( 18). Median levels and ranges of the target chemicals in air, breath, and drinking water are provided for the New Jersey subjects (Table 6) and the North Carolina TABLE 3 Coefficients of Variation of Duplicate Air and Breath Samples for Three Prevalent Target Chemicals Compound Breath Na CV* Air N CV Chloroform 1,1,1 -Trichloroethylene Tfetrachloroethylene 46 42 14 52 38 38 17 34 50 28 17 26 * Number of duplicate pairs. 1n <M * Calculated by -- Y---------------x 100, where D-. is the ith set of duplicates. N t Mean (D) URL 03822 300 WALLACE ET AL. t URL 03823 302 WALLACE ET AL. TABLE 5--Continued Chemical n-Butanol Chlorobenzene 2,2,4-Trimethylpenta-1,3-diol-) -di-isobutyrate 2-Butanol (tent.) Indan Iripropylene alycol (tent.) /t-Butyl-*-butanoaie (tent.) Diphenyl ether Diethyl phthaiate Ethyl acetone Isopropanol Thiapentane Chlorodifluotomethane Cyclopentane 3-Methyl-2-pentanone Phenyl n-propyl ether 1,2-DichloitM, 1,2,2-teu-afluoroethane Benzoic acid , Methylcyclobutane 1-Butanol Benzene-cf6 Cyclohexane-J10 Bromodichloromethane 2-Hexane Isopropoxyethaaol 2-Methylfuran 2-Undecane n-Pentanal Vinylidine chloride Silane 3-MethyIheptane Dibromochloromethane 2,3-Dimethylpentanal (tent.) Bromoform Benzonitrile n-Octanol 2-Ethyl-l-hexanol n-Nonanal n-Dodecanal (tent.) 2,6-Dw-butyl-bromomethylphenol Frequency of occurrence" Air - 8) Breath it* - 12) Water (N - 1) 0 3t 0 31 0 I1 1 01 l 00 1 00 l 00 1 00 l 2I 0 10 2 70 0 10 0 10 0 X0 0 10 0 l ,,0 1 1 ?0 0 1 -0 1 00 1 00 0 10 0 10 0 01 0 10 0 1 0 c 73 0 1 0 r~ 0 1 0 a OJ 0 1 0 CO 0 0 1 ro 0 01 0 01 0 01 0 0I 0 01 0 0I 0 01 0 01 0 01 0 01 0 01 a In summarizing the frequency of occurrence, those chemicals which were detected twice in the same sample but at different elution temperatures are assigned a frequency of occurrence of l for that sample. N = total number of samples. EXPOSURE TO VOLATILE ORGANIC COMPOUNDS 303 TABLE 6 Summary Statistics by Medium for Nine New Jersey Subjects Medium Air (jig/m3) Breath (jig/m3) Water* (home) (ng/ml) Water* (work) (ng/ml) Compound Vinyl chloride Vinylidene chloride Chloroform 1,2-Dichloroethane 1,1,1-Trichloroethane Carbon tetrachloride Bromodichloromethane Trichloroethylene 1,1,2-THchloroethane Tetrachloroethyleoe Chlorobenzene Benzene* Dichlorobenzenc(s) Vinyl chloride Vinylidene chloride Chloroform 1,2-Dichloroethane 1,1,1-Trichloroethane Carbon tetrachloride Bromodichloromethane Trichloroethylene 1,1,2-Trichloroethane Tfetrachloroethylene Chlorobenzene Benzene* Diehlorobenzene(s) Vinyl chloride Vinylidene chloride Chloroform 1,2-Dichloroethane 1,1,1-Trichloroethane Carbon tetrachloride^ Bromodichloromethane Trichloroethylene 1,1,2-THchloroethane/ Tetrachloroethylene Chlorobenzene Benzene* Dichk>rabeozene($) Vinyl chloride Vinylidene chloride Chloroform 1,2-Dichloroethane 1,1,1 -Trichloroethane Carbon tetrachloride Bromodichloromethane No. of samples* 106 No data 165 163 165 165 164 164 161 163 160 162 No data No data 14 49 49 48 49 49 49 49 49 4* 47 18 No data No data 75 75 75 13 75 75 13 75 .75 75 No data No data No data 45 45 45 No data 45 No. below limit of detection* 106 31 41 0 2 136 l 151 1 45 13 14 7 38 1 22 49 6 44 0 22 8 18 0 75 41 1 0 40 13 36 75 69 1 45 22 1 No. at trace* Median Range 0 1.1 0.63-2.84 2 2.1 0.03-129,00 88 0.65 0.10-10.20 1 9.1 0.72-708.00 95 1.0 0.19-736.00 16 0.44 0.10-13.40 38 2.8 0.43-127.00 7 0.35 0.14-34.70 14 7.2 0.57 - 250.00 64 0.61 0.05-26.50 0d __d 0 2 3.5 0.09-53.00 11 0.1 0.12-0.69 8 4.8 0.12-85.00 22 0.3 0.10-46.50 0 0.17 0.17-0.20 24 0.7 0.10-24.50 4 0.2 0.07-5.13 7 L0.5 0.64 - 202.00 21 0.2 0.07-8.15 0d 0 0.1 0.10-0.10 0 128.0 11.0-225.00 0 0.02 0.005-0.05 2 0.02 0.02-3.50 3 0.05 0.01-0.34 0 16.0 4.40 - 42.00 3 0.02 0.02-5.40 0 0.01 0.01-0.01 7 0.07 0.02-6.50 0 0.03 0.02-0.02 0 __d __d 0 127.0 0.02 - 230.00 0 0.02 0.02-0.02 3 0.07 0.02-1.60 0 13.0 0.02-37.00 304 WALLACE ET AL. TABLE 6--Continued Medium Compound No. of samples4 No. below limit of detection* No. at trace* Median Range Trichloroethylene 1, l ,2*Trichloroethane Tetrachloroethylene Chlorobenzene Benzene'' Dichlorobenzene(s) 45 No data 45 45 *5 No data 19 17 45 41 3 0.11 0.02-3.50 6 0.07 0.02-6.50 0 0.03 0.03-0.03 0 __y ___4 " Maximum number of individuals =* 9; multiple observations per individual. Duplicates averaged. * For duplicates, both had to be below the LOD to be counted. c For duplicates, at least one had to be at trace and the other at trace or below to be counted. d Due to difficulties in chemical analysis, data were not analyzed. * By purge and trap method. ! By Master Analytical Scheme analysis. subjects (Table 7). Results for the food composites collected in each community are summarized (Table 8). Air. Ten of eleven target compounds were present in air; only vinyl chloride was never found, perhaps because its limit of detection (LOD) was*>20 times the LOD of the other chemicals. Six of the ten were found in more than 75% of both the New Jersey samples (Fig. 3) and the North Carolina samples. Typical geometric means were 1-10 |xg/m3 in both locations. However, the range was often large; maximum concentrations exceeded 100 p.g/m3 for five chemicals in personal air samples from the New Jersey participants (Fig. 4) and for two chem icals in the North Carolina samples. Breath. Nine of twelve target compounds were found in breath samples from each group of participants. Six of these chemicals were present in more than 50% of the breath samples from New Jersey (Fig. 3) and North Carolina. Again the geometric means were in the 1 -10 fig/m3 range, with the exception of chloroform in breath samples of the North Carolina group, which exceeded 30 p.g/m3. Water. Municipal surface water sources were used by all subjects but one, who used bottled water. Two chemicals were predominant in drinking water samples; chloroform and bromodichloromethane. Median levels of chloroform exceeded 100 p.g/liter in both groups; median levels of bromodichloromethane were in the 15-20 M.g/liter range. No other chemicals attained median concentrations as High as 1 fig/liter. Summer concentrations of each trihalomethane were --five times the fall levels (Fig. 5). Food. Of the eight target compounds, chloroform and bromodichloromethane were found in most beverage composites and in a dairy composite; 1,1,1-tri- chloroethane and trichloroethylene were found in several fatty food composites. Further investigation of the individual food items comprising the fatty food com posite showed that the source of the trichloroethylene was margarine (Fig. 6). Beverages are likely to be an important route of exposure of chloroform and bromodichloromethane. However, the solid food samples appeared to have little EXPOSURE TO VOLATILE ORGANIC COMPOUNDS 305 TABLE 7 Summary Statistics by Medium for Thkce North Carolina Subjects Medium Air (*xg/mJ) Breath (wj/m5) Water* (home) (ng/ral) Water* (work) (ng/ml) Compound No. of samples* Vinyl chloride Vinylidene chloride Chloroform 1,2-Dichloroethane 1,1, l-Trichloroethane Carbon tetrachloride Bromodichloromelhane Trichloroethylene 1,1,2-Tric hloroethane Tetrachloroethylene Chlorobenzene Benzene* m + p-Dichlorobenzene 32 No data 59 60 61 61 60 60 60 61 60 56 No data Vinyl chloride Vinylidene chloride Chloroform 1,2-Dichloroethane 1,1, l-Trichloroethane Carbon tetrachloride Bromodichloromelhane Trichloroethylene 1,1,2-Thchloroethane Tetrachloroethylene Chlorobenzene Benzene* m + p-Dichlorobenzene No data 5 17 17 17 17 17 17 17 17 17 17 5 Vinyl chloride Vinylidene chloride Chloroform 1,2-Dichloroethane 1,1, l-Trichloroethane Carbon tetrachloride Bromodichloromethane Trichloroethylene 1,1,2-Thehkiroethane Tetrachloroethylene Chlorobenzene Benzene m fp-Diehlorobenzene No data No data 30 30 30 No data 30 30 No data 30 30 15 No data Vinyl chloride Vinylidene chloride Chloroform 1,2-Dichloroethane 1,1,1 -Trichloroethane Carbon tetrachloride Bromodichloromethane Trichloroethylene No data No data 18 18 18 No data 18 18 No. below limit of detection* 32 9 31 4 7 48 14 59 4 47 12 4 1 16 0 17 17 8 17 0 13 5 1 0 29 23 0 18 21 28 15 1 17 16 1 13 No. at trace' Median Range 0 0.70-2.26 0 3.4 0.09-17.60 27 0.10-1.58 2 4.5 0.13 - 236.00 45 0.16-4.66 7 0.10-3.66 28 0.66 0.09-5.54 0 0.14-0.94 13 2.9 0.29-125.00 7 0.05-3.66 0 1 1.25-6.25 0 50 0.11-685.00 1 0.12-0.50 8 0.6! 0.29-7.65 0 0.10-0.30 0 0.14-2.20 9 0.13-2.02 0 0.21-0.20 3 4.2 1.19-60.00 2 0.09-3.00 0 __d 4 0.09-0.76 0 120 75.0-191.0 0 0.02-0.19 0 0.02-1.90 0 16 12.0-20.0 0 0.02-1.18 2 0.02-0.44 0 0.03-3.10 0 0.04-0.04 0 no 0.02-148.00 0 0.02-0.13 0 0.02-0.80 0 14 0.02 - 20.00 0 0.02- 1.48 306 WALLACE ET AL. TABLE 7--Continued Medium Compound No. of samples0 No, below limit of detection* No. at trace' Median Range 1, t 2-TrichJoroethane Tetrachloroethylene Chlorobenzene Benzene No data IS 18 9 12 18 8 2 0 0 0.02-0.43 0.25-0.25 0.04- 24.00 " Duplicates been averaged. * For duplicate*, both measures had to be below the LOD in order to be counted. e For duplicates, at least one measure had to be at trace and the other at trace or below, in order to be counted. d Due to methodological difficulties in chemical analysis, statistical analysts was not performed. * By purge and trap method. potential for exposure, with the possible exception of trichloroethylene in mar garine. Temporal and Occupational Effects Air and breath levels varied considerably from day to day and with the season (Table 9). For the nine New Jersey residents, chloroform in breath was much higher (P < 0.01) in July than in December, corresponding to similar differences in air and drinking water concentrations. Tetrachloroethylene and 1,1,1-trichloroethane were significantly higher in breath in July and December than in Sep tember. Air levels of carbon tetrachloride and 1,2-dichloroethane were signifi cantly lower in December than in July. Weekday air exposures were significantly (/ < o.Ol) higher than weekend exposures for trichloroethylene, tetrachloroethy lene, and 1, 1,1-trichloroethane, indicating a possible source in the working or commuting environment. Time of day was a relatively unimportant variable, with only 1,1,1-trichloroethane showing a significant decrease between daytime and night-time values. Occupational exposures were evident for one individual (the sewage treatment plant worker), who had many of the highest values noted for chloroform, carbon tetrachloride, tetrachloroethylene, and trichloroethylene n air and breath. However, the mean exposures for all five workers in oil, chemical, and sewage treatment plants were not greatly different (about a factor of 2) from the exposures of the four "unexposed" New Jersey subjects. Total Exposures For the four most prevalent compounds, mean daily exposures were estimated for each subject, assuming an air intake of 10 cubic meters/day and a water intake of 1 liter/day ( Table 10). (Of course, air ar.d water intakes vary by activity le\ ti the values selected are conservatively low for sedentary adults. Values of 20 m-; day and 2 liters/day have been employed in EPA risk assessments.) Three of the four chemicals -- trichloroethylene, tetrachloroethylene, and 1,1,1-trichk'ioethane__were transmitted almost exclusively through air. The fourth (chloroform! URL 03828 EXPOSURE TO VO LATILE ORGANIC COMPOUNDS UftL 03829 TABLE 8 VOMTILE HAI.OCARDONS W FOOD AND BEVERAGE COMPOSITES FROM NORTH CAROLINA AND NEW JERSEY9 Sample location NC NC NJ NJ Composite Chloroform Dairy Meats Fats Beverages Dairy Meats Fats Beverages ' __ * -- _ n -- -- 12 Dairy Meats Fats Beverages -- -- -- 10 Dairy Meals Fats Beverages -- -- 6 LLl-Tyichloroethane -- -- -- -- -- -- -- ---- -- -- -- -- -- Bromodichloromelhane -- -- -- 1.0 -- -- -- -- -- -- -- -- -- -- Carbon tetrachloride -- -- -- -- -- -- -- -- -- -- -- -- -- -- Trichloroethylene -- -- 920c -- -- -- -- -- -- 55 -- -- Tetrachloro* ethylene -- -- -- -- -- -- -- -- -- -- -- -- ---- 9 Noi corrected for method recovery. b Not detected. r AH values in ppb. Ethylene dibromide -- -- -- -- -- -- -- -- -- -- -- -- -- 308 WALLACE ET AL. |-j~i sh: VJ'oS*F Fig. 3. ftrcentage frequencies of detection for target compounds (10 volatile organics) in personal air and breath samples for the nine New Jersey subjects. Key: PERC tetiachloroethylene, 111TR1 = 1,1,1-trichloroethane, TRI - trichloroethyleite, BENZ * benzene, CARBTET - carbon tetra chloride, CHLO = chloroform, EDC ** 1,2-dichloroethane, CLOBENZ = chlorobenzene, BDCM ** bromodichloromethane, and 112TRI = 1,1,2-trichloroethane. 3D OCO CO oCO Fig. 4. Range of concentrations of nine volatile organics for New Jersey participants (165 personal air samples, 49 exhaled breath samples). For key to chemical names sei caption to Fig. 3. > O* EXPOSURE TO VOLATILE ORGANIC COMPOUNDS v V v* ` Fig. 5. Cumulative distribution of trihalomethane concentrations in drinking water, by season, for the nine New Jersey participants. Plotted on iog*normal probability paper. was transmitted mainly through water, but a significant fraction (32 1896) came through the air. If the contribution of chloroform from beverages is included, the relative contribution from air would decrease. Mean exposures ranged over more than an order of magnitude for each chemical. URL 03831 URL 03832 TABLE 9 Adjusted Geometric Means for Volatiles by Medium, Trip, Exposure, Type of Day, and Time of Day from New Jersey Data Medium Compound Trip (Season) July Sept. Dec. Occupational Exp. Unexp TVpe of day WD WE Time of day Work Even Sleep No. of samples Breath (jig/m3) Chloroform 1,1, 1-Trichloroethane Tetrachloroethylene Air (fig/m3) Chloroform Carbon tetrachloride 1,2-Dichloroethane 1,1,1 -Trichloroethane Trichloroethylene Tetrachloroethylene Chlorobenzene Water (home) (ng/ml) Chloroform 1,1,1 -Trichloroethane Bromodichlorornethane Trichloroethylene 6.2 4.4 10.1 3.1 1.9 .82 9.4 3.9 6.8 .60 148.4 .12 25.5 .05 3.7 (0.01) 1.7 (0.05) 4.0 (0.05) 1.1 (0.01) 1.4 (0.05) .81 (0.05) 9.1 (NS) 3.0 (NS) 6.7 (NS) .72 <0.10) (0.01) (0.01) (0.01) (0.05) .40 6.2 8.7 1.1 1.1 .56 10.7 2.7 8.1 .40 28.8 .08 8.8 .09 2.1 2.1 (NS) 2.6 5.3 (0.10) 9.8 4.6 (0.05) 2.2 1.1 (0.01) 1.5 1.3 (NS) .74 .71 (NS) 8.6 11.4 (0.10) 3.4 3.0 (NS) 8.3 5.9 (0.05) .69 .44 (0.05) * i *M 1.8 2.5 (NS) 4.5 2.9 (NS) 10.0 4.8 (0.10) 1.4 1.7 (NS) 1.4 1.4 (NS) .81 .65 (0.05) 13.7 6.9 (0.01) 4.6 2.2 (0.01) 10.1 5.1 (0.01) .57 .56 (NS) 70.8 60.9 (0.01) .06 .06 (NS) 16.4 13.5 (0.01) .06 .07 (NS) 1.6 1.6 .60 17.3 3.2 8.2 .59 1.8 (NS) 1.4 (NS) .85 (0.05) 5.8 (0.01) 3.2 (NS) 6.8 (NS) .66 (NS) 64.1 67.4 (NS) .07 .07 (NS) 14.0 15.5 (0.10) .06 .06 (NS) 1.3 1.2 .74 9.0 3.2 6.7 .44 49 (9) 48 (9) 49 (7) 165 (33) 165 (97) 163 (129) 164 (1) 163 (39) 163 (15) 159 (109) 75 (0) 75 (43) 75 (0) 75 (43) EXPOSURE TO VO LATILE ORGANIC COMPOUNDS URL 03833 Telrachloroethylene Water (work) (ng/ml) Chloroform I, l,l-THcfiloroerhene Bromodichloromethane Trichloroethylene Telrachloroethylene .09 94.6 .15 16.0 .05 .07 (N.S) (O.Ol) (NS) (0,10) (0.05) (NS) .06 27.1 .08 8.2 .15 .11 40.0 68.0 (NS) .06 .25 (.05) 10.6 12.5 (NS) .08 .09 (NS) .11 .08 (NS) .07 .08 (NS) 83.9 30.6 (0.05) .10 .13 (NS) 17.6 7.9 (0.05) .10 .07 (NS) .10 .08 (NS) .06 .10 (0.10) 75 (43) 45 (1) 45 (25) 45 (1) 45 (22) 45 (23) a Significance level of the test of the difference between geometric means. NS = not significant. * () - Number of samples below LOD or at trace levels. (In general, data are based on only nine respondents and should not be extrapolated to larger populations.) r Kit URL 03834 UJ SJ WALLACE ET AL. Subject 1 2 3 4 5 6 7 8 9 10 11 12 Mean Standard deviation TABLE 10 Estimated Mean Daily Air Exposures of Twelve Subjects to Four Volatile Organics Occupation Mean daily air exposure. Hg {%)" N* 1,1,1-THchloroethane Tctrachloroelhylene Trichloroelhylenc Statistician Statistician Statistician Activities coordinator Oil plant worker Oil plant worker Housing inspector Chemical plant operator Chemical plant operator Police officer Police officer Sewage treatment operator 10 10 10 9 9 10 4 8 9 4 8 9 360(100) 120 (100) 55 (99) 210(100) 110(100) 910(100) 720(100) 86 (98) 130(100) 54 (100) 1000(100) 680(100) 370 (99.8) 160(100) 72(100) 72 (100) 78(100) 78 (100) 340(100) 60(100) 100 (9$) 92 (100) 24(100) 450 (100) 610(100) 178 (99.6) 14 (100) 12 (97) 15(97) 48 (99) 66(100) 140(100) 77(100) 21 (87) 120 (100) 14 (100) 170 (100) 300(100) B3 (98.3) 357 (0.6) 186(1.4) 87 (3.7) I_ Exposure in air x 100 Percentage of total exposure provided by air only, calculated by -- 2, 1 -------------------------- ----- --........... N (Exposure in air plus exposure in water) b Number of days sampled. I ' i Chloroform 25 (16) 33 (22) 73 (37) 82 (46) 38 (34) 53 (39) 32 (17) 61 (44) 45 (37) 31 (14) 7(5) 260 (70) 62 (32) 66 (18) EXPOSURE TO VOLATILE ORGANIC COMPOUNDS 313 DISCUSSION Review of Previous Studies Although a number of investigators have measured these or similar chemicals in ambient air (Brodzinsky and Singh, 1982) or in drinking water (Boland, 1981), few have measured individual exposures in breathing-zone air and drinking water simultaneously. The first such effort occurred in March 1979 (Wallace et al., 1982; Zweidinger et al., 1982). Eleven college students at Lamar University in Beau mont, Texas and six students at the University of North Carolina carried personal monitors for 5-9 hr, collected drinking water samples, and gave breath, blood, and urine samples to be analyzed for 15 volatile organics. Findings included high frequencies of detection for seven chemicals in air and wide variability in expo sures. Later studies were carried out for benzene in Houston and St. Louis (Zwei dinger et al., 1980) and for a number of volatiles in Baton Rouge/Geismar, Lou isiana, in Houston, Texas, and in Greensboro, North Carolina (Petlizzari et at., 1983). Field work in the large-scale follow-up to the pilot study described here has been completed on 362 volunteers in Bayonne and Elizabeth, New Jersey, 25 in Greensboro, North Carolina, and 25 in Devils Lake, North Dakota (Wallace et al., 1983). A major finding of the last two studies has been the significantly higher fre quencies of detection and concentrations of these chemicals in personal air sam ples than in simultaneous outdoor samples. For example, 90% of personal air samples in Greensboro had detectable levels of chloroform compared to only 40% of the outdoor samples. Similar significant differences were noted for benzene, styrene, 1,1,1-trichloroethane, bromodichloromethane, trichloroethylene, paradichlorobenzene, and trichlorobenzene isomers in Greensboro, Baton Rouge/ Geismar, and Bayonne/Elizabeth. Thus reliance on outdoor ambient air sampling may grossly underpredict the frequency and severity of exposure to these toxic and carcinogenic compounds, which have a number of indoor and in-transit sources. One previous study of exhaled breath (Krotozynski et al., 1979) has provided valuable baseline information on concentration ranges in Chicago residents. Two field studies of volatile organic levels in food have appeared. One (Pearson and McConnell, 1975) dealt with foodstuffs in the United Kingdom; the other (Bauer, 1981) dealt with levels in ambient air, water, food, and human tissue in West Germany. Volatile organic levels in the U.K. foodstuffs tended to agree in the aggregate with levels found in this study; however. West German levels of chloroform in drinking water were far lower, and tetrachloroethylene in food (particularly pork sausage) far higher, than the levels reported here. URL 03835 The Present Study Air. From Table 5, we find that 46 chemicals were present in at least half of the air samples selected for broad-scale analysis; and 16 were present in all sam ples. Thus it is clear that these urban and suburban subjects were inhaling a complex mixture of compounds including known human carcinogens, animal car 314 WALLACE ET AL. cinogens, mutagens, and cocarcinogens. The significance with respect to health of inhaling these compounds at low levels for long periods of time is not known. Distributions of most compounds were closer to log normal than normal, as determined by the Kolmogorov--Smirnov modified D-statistic. The five com pounds with the highest geometric mean concentrations for the 12 respondents were 1,1,1-trichloroethane (8.6 p.g/m3), tetrachloroethylene (6.3), trichloroethy lene (2.3), chloroform (1.9), and carbon tetrachloride (1.3). The relative concentrations in air of the three most prevalent compounds were the same for the North Carolina subjects as for the New Jersey subjects (Figs. 7 and 8). In both cases, the highest concentrations were, in descending order, 1,1,1trichloroethane, tetrachloroethylene, and chloroform. Breath. From the broad-spectrum analyses of 12 breath samples, 37 chemicals were present in at least half the samples, and 8 were present in all samples. These figures prove that many, probably most, of the chemicals being inhaled or ingested are in fact crossing the body membranes into the blood stream. Fig. 7. Cumulative frequency distributions of three volatile organics in air in New Jersey, plotted on log-normal probability paper, of 8-hf exposures for the New Jersey residents (N - 170) show 1000-fold ranges. EXPOSURE TO \OLATlLE ORGANIC COMPOUNDS 315 FlO. 8. The frequency distribution of three volatile organics in air in North Carolina indicates that exposures of the North Carolina residents have a similar range to those of the New Jersey residents. Again the observed distributions of the chemicals were closer to log normal than normal, as judged by the Kolmogorov-Smimov modified D-statistic. No significant difference was observed between the breath concentrations of three chemicals measured before and after dinner. This is further indication that food was not an important source of volatile organics for the subjects tested. The four chemicals with the highest geometric mean concentrations were tet~ rachioroethylene (7.9 jjig/m3), .chloroform (4.5), 1,1,1-trichloroethane (2.6), and trichloroethylene (0.7). The relative breath concentrations of the three most prevalent compounds were sharply different for the North Carolina subjects from the New Jersey subjects (Figs. 9 and 10). Chloroform, which showed the lowest concentrations of the three chemicals in air in both locations, and was also the lowest in the breath of the New Jersey subjects, was highest in the breath of the North Carolina subjects. The five highest values occurred during the second trip, which (unfortunately) was the only time that drinking water concentrations were not measured. Spearman correlation coefficients were calculated for all possible pairs of the chemicals in air and in breath for the New Jersey subjects. The three strongest correlations for both air and breath were between trichloroethylene, tetrachlo- C 3D r-- o C--OCOOJJ 316 WALLACE ET AL. URL 03833 Fio. 9. Concentration levels of three volatile organics in exhaled breath in New Jersey paralleled the air exposures. roethylene, and 1,1,1-trichloroethane (Table il). These correlations suggest a common source for at least a portion of the subject's exposures to these solvents. The lack of significant correlations between chloroform and the other chemicals in air or in breath is further indication of the essentially dissimilar exposure pathways for chloroform compared to the other three prevalent vcUtTe erganics. Water. The subjects' tap water, drawn from municipal surface wtirer supplies, contained only two of the target chemicals in significant amounts: crJoroform and bromodichloromethane, each of which contributed at least T5T- the sub jects' total exposure to each chemical. Although several other haior-irrons were present in more than half of the samples, their contribution to the subjects' total exposure was nearly always less than 1%. Food. Chloroform and bromodichloromethane in beverages such cs sort drinks may supply a significant fraction of the daily intake of these cheirJc-Xjs.. The solid foods tested did not appear to contribute importantly to intake of un;* ruLocarbon. However, only four samples were tested for each of four food so no general conclusions can be reached. EXPOSURE TO VOLATILE ORGANIC COMPOUNDS 317 FtC. 10. Concentration levels of three volatile organics in exhaled breath in North Carolina showed high chloroform levels. TABLE 11 Significant (P < 0.01) Spearman Correlations of Volatile Organics in Personal Air and Breath of Nine New Jersey Subjects Breath (N - 49) Air (N = 160) fttUutant I Pollutant 2 rPrP Trichloroethylene Trichloroethylene Tetrichloroethylene Tetrachloroethylene Trichloroethylene Chloroform Carbon tetrachloride Chloroform Tetrachloroethylene 1,1,1-Trichloroethane 1.1.1-Trichloroethane Carbon tetrachloride Chlorobenzene Chlorobenzene Trichloroethylene Carbon tetrachloride .69 .63 .60 .44 ` .41 .41 .37 -- 0 0001 0.0001 0.0001 0.002 0.004 0.004 0.009 NS .68 0.0001 .60 0-0001 .58 0.0001 -- NS -- NS -- NS .25 0.002 .43 0.0001 318 WALLACE ET AL. CONCLUSIONS The following conclusions are based on several hundred environmental samples collected over a period of time from only 12 persons, and should not be extra* polated to larger populations. (1) Monitoring individual air and water exposures and breath concentrations for a number of volatile organic compounds simultaneously proved feasible (pro vided careful attention is paid to sources of contamination* particularly benzene) and gave useful information on the concentrations encountered in normal daily activities. (2) Six of the target compounds were found in more than half of the air and breath samples, but seldom or never in food or water. Two compounds were transmitted primarily through drinking water or beverages. None was transmitted primarily through food. Two target compounds (vinyl chloride and 1,1-2-trichIoroethane) were virtually never found. (3) Median concentrations in air and breath were usually in the 1-10 tig/m3 range, but air concentrations were highly variable, with maxima exceeding 100 p.g/m3 for five chemicals. On the other hand, little day-to-day or person-to-person variability within a given geographic area was noted for trihalomethanes in drinking water, but a strong seasonal component was present, with summer values much greater than winter values. Beverages appeared to be an important route of exposure for chloroform and bromodichloromethane. (4) Mean daily intakes of the four most prevalent chemicals were conservatively estimated to range from about 50 to 1000 pg/day for 1,1,1-trichloroethane, 25 to 600 M-g/day for tetrachloroethylene, 10 to 300 p.g/day for trichloroethylene, and 140 to 370 pg/day for chloroform. ACKNOWLEDGMENTS We thank Df. Robert Ziegenfus for providing computer graphics for this article, and Mrs. Brenda Hurley for seemingly endless retyping of text and tables. REFERENCES Bauer, U. (1981). Human exposure to environmental chemicals--Investigations of volatile organic halogenated compounds in water, air, food, and human tissues (text in German). ZemraSbl. Bakteriol. Parasitenkd. Infektionskr. Hyg. Abt. I Orig. Reihe B 174, 200-237. Bellar, T. A., and Lichtenberg, J. (1974). Determining volatile organics at microgram-per-litre levels by gas chromatography. J. Amer. Water Works Assoc. 66, 739-744. Boland, P. A. (1981). `'National Screening Program for Organics in Drinking Water." SRI Interna tiona], EPA Contract No. 68-01-4666. Brodzinsky, R., and Singh, H. (1982). "Volatile Organic Chemicals in the Atmosphere: An Assess ment of Available Data." Environmental Sciences Research Laboratory, USEPA, Research Tri angle Park, N-C. Entz, R,, Thomas, K.. and Diachenko, G. (1982). Residues of volatile hatocarbons in food using headspace gas chromatography. J. Agric. Food Chem. 30, 846. Helms, C., et al. (1981). "Evaluation and Classification of the Potential Carcinogenicity and Muta genicity of Chemical Biorefractories Identified in Drinking Water." U.S. Dept, of Health & Human Services, Washington, D.C. Krost, K. J., Pellizzari, E. D., Waibum, S. G.. and Hubbard, S. A. (1982). Collection and analysis of hazardous organic emissions. Anal. Chem. 54, 810. URL 03840 EXPOSURE TO VOLATILE ORGANIC COMPOUNDS 319 Krotozynski. C. K., Bruneau, G., and O'Neill. H. J. (1979). J. Anal. Toxicol. 3, 225-34. Pearson. C. R.. and McConnell. G. (1975). Chlorinated C( and C* hydrocarbons in the environment. Proc. R. Soc. London Ser. B 189, 305. Pellizzari, E. (1979). "Analysis of Organic Air Pollutants by Gas Chromotography and Mass Spec troscopy " U.S. Environmental Protection Agency, Research Triangle Park, N.C. Pellizzari, E. D., Erickson, M. D., Guiguere, M. T., Hartwell, T. D., Williams, S. R., Sparacino. C. M., Zelon, H., and Waddell, R. D. (1980). "Preliminary Study on Toxic Chemicals in Environmental and Human Samples: Work PJan," Vols. I. and II, (Phase I)- U.S. Environmental Protection Agency, Washington, D.C. Pellizzari, E. D., Hartwell, T. D., Leininger, C., Zelon, H., Williams, S., Breen J. J., and Wallace, L. (1983). Human exposure to vapor-phaSe halogenated hydrocarbons: Fixed-site vs personal exposure. In "Proceedings, National Symposium on Recent Advances in Pollutant Monitoring, Raleigh, N.C." EPA-600/9-83-007. Pellizzari, E. D., Hartwell, T. D., Sparacino, C. M., Sheldon, L. S., Whitmore, R-, Leininger, C., and Zelon, H. (1984). "Total Exposure Assessment Methodology (TEAM) Study: First Season-- Northern New Jersey, Interim Report." U.S. Environmental Protection Agency, Wash. D.C., EPA Contract 68-02-3679. Sparacino, C., Leininger, C., Zelon, H., Hartwell, T., Erickson, M., and Pellizzari, E. (1982a). "Sampling and Analysis for the Total Exposure Assessment Methodology (TEAM) Prepilot Study: Research Triangle Park, Final Report." U.S. Environmental Protection Agency, Wash. D.C., EPA Contract 68-02-2688. Sparacino, C., Pellizzari, E., and Erickson, M (1982b). "Quality Assurance for the Total Exposure Assessment Methodology (TEAM) Prepilot Study, Final Report." U.S. Environmental Protection Agency, Wash. D.C., EPA Contract 68-02-3146. Wallace. L. A., Pellizzari, E. D., Hartwell, T. D., Sparacino, D., and Zelon, H. (1983). "Personal Exposure to Volatile Organics and Other Compounds Indoors and Outdoors--The TEAM Study." Presented at 76th Annual Meeting of the Air Pollution Control Association, Atlanta, 1983. Wallace, L. A., Zweidinger, R., Erickson, M., Cooper, S,, Whitaker, D., and Pellizzari, E. D. (1982). "Monitoring individual exposure: Measurements of volatile organic compounds in breathing-zone air, drinking water and exhaled breath." Environ. Int. 8, 269-282. Zweidinger, R. A., Cooper, S. D., Harris, B. S. H., Ill, Hartwell, T. D., Folsom, R. E., Jr., Pelliz zari, E. O., Sherdon, A. W,t Wong, T. K., and Zelon, H. S. (1980), "Measurement of Benzene Body Burden for Populations Potentially Exposed to Benzene in the Environment." U.S. Envi ronmental Protection Agency, Wash. D.C., EPA Contract 68-01-3849. Zweidinger, R., Erickson, M., Cooper, S., Whittaker, D., Pellizzari, E. D., and Wallace L. (1982). "Direct Measurement of Volatile Organic Compounds in Breathing-Zone Air, Drinking Water, Breath, Blood, and Urine." U.S. Environmental Protection Agency, Washington, D.C., NTIS No. PB-82-186-545. Government Toxic chemical levels higher indoors than out To the surprise of Environmental Protection Agency scientists, a five- year study has revealed that people are exposed to far greater concen trations of common toxic organic chemicals indoors than they are out doors, even in cities where plants manufacture or use these chemicals. This finding, which has been con firmed by European studies, has vast implications for the way toxic pol lutants are measured and regulated. The study found that air 19 the main exposure route for the most commonly found 11 volatile organ ic chemicals. But contrary to gener al belief, people living in areas teem ing with petrochemical, paint, or plastics processing plants are not subjected to greater exposures than are people living in less industrial ized, or even rural, areas, The sci entists reached this conclusion af ter studying 355 people in the high ly industrialized cities of Bayonne and Elizabeth, N.J., 25 people in the lightly industrialized city of Greensboro, N.C., and another 25 in rural Devils Lake, N.D. Using personal and stationary out door air monitors, the investigators found median indoor levels to be two to five times greater than me dian outdoor levels for the 11 vola tile organic chemicals. At the high est exposures, indoor levels exceeded outdoor levels up to 70 times. Pre cise indoor sources of these pollu tants are not known. The scientists, however, suspect building materi als and consumer products. The study's principal investiga tor Lance A. Wallace, a physicist in EPA's Office of Research & Devel opment, says he "doesn't know whether the levels detected indoors cause cancer." However, he adds, one of the 11 chemicals, benzene, is a known human carcinogen, and five others are suspected cancer- causing'agents. He suspects that the 11 may be contributing to the "sick- building syndrome" phenomenon. As buildings have become more air tight, complaints of sleepiness and general malaise have increased. Wallace says that the National Aeronautics 4c Space Administration is an untapped repository for infor mation on the contribution of ma terials to indoor air pollution. The agency has tested 10,000 materials used in the shuttle for their "off gassing" characteristics. Because as tronauts were experiencing the sickbuilding syndrome, NASA needed to know which materials were re leasing volatile organic chemicals. Xylene, a toxic chemical at high con centrations and one of the 11 most often measured chemicals indoors, was detected by NASA in nearly 800 of the 10,000 materials tested. NASA found that marking penf re leased the highest levels of xylene, Wallace says. In the EPA study, another of the Our 24th annual chromatography catalog is packed with application know-how. Our 1985 Analabs Chromatography Catalog is virtually the only source you'll ever need to precisely find the right GC, LC, and/or TLC accessories and chemical standards. You'll probably find yourself and your staff using it more than ever because we've expanded our capillary and packed column sections to help you keep pace with the latest in separation technologies. And, you'll find many new and exclusive environmental standards. If our new catalog doesn't have everything you need--don't be afraid to ask. Our 24 years of serving the industry means we understand the chromatographer's needs. So don't think of us as just a supply house--we're in business to help the practicing chromatographer. For fast service send in the coupon and we'll send you the source. The Foxboro Company, 80 Republic Drive, North Haven, CT 06473, (203) 288-8463. I'll take one! Send my copy of the Analabs Chromatography Catalog to: I I Company. Address_ City_____ .State. -Zip. The Foxboro Company,80 Republic Or., No. Haven, uo*wj I I I 03842 1 Government Many toxic chemicals are prevalent indoors 1,1,1-Triehloroethane, solvent Tetraehloroethylene, solvent Benzene, paint, tobacco, gasoline o-Xylene, paint, gasoline rn,p-Xylene, paint, gasoline Ethylbenzene, paint, gasoline Carbon tetrachloride, solvent Trichloroethylene, solvent Chloroform, tap water Styrene, insulation, plastics p-DIchlorobenxene, moth crystals, deodorants 11 chemicals, benzene, was found to be 30 to 50% higher in the air of homes of smokers than in the homes of nonsmokers. Using a breath anal ysis technique developed for this study, smokers' breath was found to contain twice the concentration of benzene than did the breath of nonsmokers. Wallace says one of the most sig nificant findings of the EPA study is that "breath values are reflective of personal exposure" to toxic vola tile organic chemicals. The test is simple, sensitive, noninvasive, but expensive--about $500 per sample. It would be most useful in monitor ing the exposure of people to a chemical spill or release, especially when air monitors could not be used in a timely fashion. The findings of the EPA study, known as the TEAM (total exposure assessment monitoring) study, have been replicated by European scien tists using different methodology. Phillip J. Walsh, an environmental health scientist at Oak Ridge Na tional Laboratory, reviewed the TE^\M study protocols. He says TEAM "is a competent study, con ducted by a competent team, so the results are probably valid results." A special review panel of EPA's Sci ence Advisory Board (SAB) was even more positive. After reviewing the TEAM study, the SAB panel wrote: "This series of studies, more than any other, pro vides compelling evidence of the necessity of evaluating total human exposure and of riot basing expo sure estimates on pollutant concen trations measured only by fixed-sta tion monitors." At present, under the Clean Air Act and other rele vant laws, epidemiological studies are designed, and regulations are set based on measurements from fixed-station monitors. EPA staff director for SAB Terry F. Yosie says that Wallace and his TEAM collaborators "identified spe cific microenvironments and corre lated human activity patterns with pollutant exposures." Since people spend up to 95% of their time in doors, and indoor air contains high er levels of 11 toxic chemicals than outdoor air, scientists are beginning to think that legislation ought to be drafted to reflect these realitiesV Lois Ember, Washington Tax Deferred Retirement Annuity 7.55%and individual Retirement Accounts (IRA's) The ACS Board of Trustees, Group Insurance Plans for ACS Members, announces a new guaranteed rate, compounded daily, equivalent to an effective annual rate of 7.55%. This rate wilt be applied to all contributions made from July 1-31,1985, and is guaranteed for a full three years. The minimum contribution requirement is $100. Please mail this coupon lor more information and enrollment forms. Or telephone 1-800-243-4896 (in Connecticut 1-467-2887) Professional Pensions, Inc. 444 Foxon Road East Havsn, Connecticut 06513 (umJarwrlttan by MHrepofrW Ufa Insoranct Company) Name _____________________________ Street _____________________________ 24 June 24. 1985 C4EN CIRCLE 29 ON REAOER SERVICE CARO