Document MJ9a0yXjm9BxnXXrEmjjOoV3k

Hr/crvnccs ' Dutliy, H.T., Kaplan, }\.,Appl. Microbiol., 16, 900 (1968). Ucm` K. 0., Chindgien, 0. !.. Valdez, E. 0., "Innovations in lU'cyclins Automotive Scrap," Ann. Meeting, Institute or Scrap Iron and Steel lnr,, Washington. DC, January 15-18,1972. Demi, K. C.. Mahoney. L. K.t Valdez, E. G., manuscript in prep aration, 1973. L*r*c, . N., Sixth International Gel Permeation Chromatogra phy Seminar, pp 11-129, Miami, Fla., October 1968. Mahoney. L. It.. Weiner, S. A., unpublished data, 1972. Saunders, J. H., Frisch, K C"Polyurethanes, Chemistry and Technology," Part li Technology, Interscicncc, New York, N.Y., 1967. ' Warner, A. J., Parker, C. It., Itauvn, B., "Solid Waste Manage ment of Plastics," p A-58, Project 1440.2, Ih-Bell and Kichard- 80i>, Hnzardvillc, Enlield, Conn. (1970a>. Warner, A. J., Parker, C. M , Baum, B., ibid , p A 81 (1970b). llcctioed for review December 20, 1972. Accepted September 7 1973. . -4 1 > Aqueous Solubility, Adsorption, and Vapor Behavior of Polychlorinated Biphenyl Aroclor 1254 Rlzwanul Haque,' David W. Schmedding, and Virgil H. Freed Dapartmenl of Agricultural Chemistry and Environmental Health Sciences Center, Otegon State University, Corvallis. Ore. 97331 The water solubility, adsorption from aqueous solution, and the vapor behavior of the polychlorinated biphenyl Aroclor 1204 have been determined. The aqueous solubili* ty of Aroclor 1254 h<t* been found to be ~5G ppb. The ex tent of adaorption of the PCB's is highly dependent on the nature of the adsorbent. Del Monte sand adsorbs very lit tle ns compared to a Woodburn soil. The vapor loss of Aroclor is significant from the sand but negligible from the soil. The va|K>r loss increases with increasing tempera ture. In general, the isomers containing fewer chlorine atoms show greater loss as compared to other isomers hav ing more chlorine atoms*. The occurrence of polychlorinated biphenyls (PCB's) in the environment and their consequence on environmental quality have been well emphasized in recent years (Nel son, 1972; Gustafson, 1970, 1972; Peakell and Linrer, 1970). Duta describing PCB residues in water and animal tissue and their toxicological effects are accumulating at a rapid rate (Nelson, 1972; Gustafson, 1972). However, little is known about the mechanism of transport of PCB's in the environment. The question of how PCB's are trans ported in the biosphere still remains. To answer this, one must know the basic properties of PCB's. In the present manuscript, wc shall describe tire behavior of the PCB, Aroclor 1254, in some laboratory water, soil, and air sys tems. The chemical properties discussed arc the water solubility, vapor loss, and adsorption. Experimental The PCI), Aroclor 1254, was a sample from Monsanto Chemical Co. and was used without further purification. The clay, sand, silica gul, and soil samples used in this study and their sources arc given in Table I. All adsor bents were used as received, aside from screening (00/80 mesh), except Ottawa semi (20/30 mesh), which was used oh received. Woodburn soil was determined to have 3.1% organic matter and 16.2% Hays. The distilled water used 1 To whom cnrn-spomtcacc should Ijc addressed. throughout the studies was extracted with hexane and boiled for 1 hr to remove the residual hexane. This proce dure was necessary to be sure test water was free from dissolved organic materials. The procedure for each exper iment is outlined as follows: Solubility Studies. Approximately 5 grams of the chemical was placed in a large Erlenmeycr flask (6-liter), stoppered, and equilibrated with water. To prevent dis persion of PCB, the sample was slowly stirred with a mag netic stirrer. To eliminate heat transfer, the stirrer and flask were separated by a %-in. polystyrene sheet. An ali quot of aqueous sample was removed, centrifuged at 3000 rpm for 10 min, extracted with hexane, and analyzed for the PCB. The analysis was carried out at 1-weck intervals and was continued until it was determined that the con centration had reached the saturation point. A typical gas-liquid chromatogram of a standard and a water extract is shown in Figure 1. The individual peaks in the chromatogram corresponding to different numbers of chlorine substitutions were identified by mass spec trometry. The distribution of chlorine components in the mixture was established initially by microcoulomctric gas chromatography. The gas chromatogram shape changed owing to the differential rates of solubilization of the vari- Table I. Sources of Materials Under Investigation SurfacB Source lllitc day Kaolinite clay Montmoriilonite day Silica gel Del Monte sand Woorilmrn soil Ottawa sand lllite r'35-- Fithian. III., Wards Natural Science Establishment Inc, Roch ester, N.Y. Kaolinite ('4--Oneal Pit. Macon, Ga., Wards Natural Science Establishment Inc., Rochester, N.Y. Montmoriilonite ^26--Clay Spur, Wyo . Wards Natural Science Establishment Inc , Rochester, N.Y. Silica Gel, Powder !?3405, J. T. Baker Chemical Co.. Phillipsbury, N.J. Del Monte Sand, EI.-.40, Del Monte Properties Co., San Francisco. Calif. Woodburn Silt Loam, 0 4 in , Hyslop Farm, Corvallis, Ore. Scientific Supplies Co.. Seattle. Wash. !.! i* I 1, -r 3 </> -Z o 'i X * t \ Figure 1. Typical gas chromatogram* of standard and a watar extract Aroclor 1264 In hoxana out iaomer* before equilibrium. The envelope was divided into 4, 5, 6, and 7 chlorine components by drawing a line perpendicular to the baseline at the minimum of the trace between various chlorinated portions. Further division was not attempted owing to insufficient separation of indi vidual isomers. Since many of the peaks were not sharp, the cut and weigh method was used to determine peak areas. Isomers having different numbers of chlorine atoms posses* varying electron-capture detector response, thus a correction was needed to quantities the data. The differ ential response for various chlorine containing isomers was referenced relative to p,p'-DDE. The final concentration of the I'CB was determined by comparing with a known standard. Adsorption Studies. The adsorption studies were car ried out by oquitibraling a saturated aqueous solution of PCB with a known amount of the adsorbent material. The saturated aqueous solution was previously equilibrated as described for approximately 6% months. Since the solu bility of Aroclor 1204 is in the ppb range, the equilibrium xjwiment* were performed with the concentration of PCB constant and the amount of adsorbent material var ied rather than vice versa. A known quantity of adsorbent material was weighed in a 250-ml centrifuge bottle, and a saturated solution of the PCB (125 ml) was added and kept on a shaker for 24 hr. Earlier studies showed that thix period of time waft sufficient to reach equilibrium for the adsorption of other pesticides on surfaces from aque ous solution (Haque and Sexton, 1968; Haque and Cosh ow, 1971). The samples were then centrifuged at 3000 rpm, and 100 ml of (lie supernatant liquid were extracted with three 10-ml portions of hexene. The extract was blown down,to approximately 3 ml under a nitrogen stream and the sample made up to 5 ml in n volumetric flask. The concentration of PCB remainmg in solution was determined on an Aerograph 550 gas cbroinatogrnph, equipped with an electron capture detec tor. The adsorption of Aroclor 1254 on the surface of the container was accounted for bv running n blank through the analysis. We have determined the adsorption of Aro clor 12M on surfac es by Inking into account the sum of all the gas chromatographic peaks. No attempt was mode to determine the adsorption nf each constituent, on individu al materials. The amount of chemical adsorbed, x, wan calculated by the formula: x - V(C3 - C,) where V is the volume of the adsorbate, Ct and C'| are the original nnd final concentrations of adsorbate (ppb). Vapor Loss Studies. Two different kinds of investij-ations were carried out to observe the vapor loss of I'CB. In one experiment the loss was studied from Aroclor 1254 it self, whereas in the other case the vapor loss was moni tored from Ottawa sand and Woodburn soil. In the first set a known quantity (~0.G gram) of the chemical was placed in planchets (exposed area 5.07 cm8) and the vapor loss was monitored gravimetrically as a function of time. The experiments were carried out in duplicates and at two different temperatures (26 and 60C). The second experi ment was performed as follows: Sand and the soil surfaces were treated with the PCB. An ether solution of Aroclor 1254 of known concentration was mixed with sand or soil and dried on a rotary evapo rator for 2.5 hr. Ottawa sand (130 grams) containing PCB was placed in petri dishes. Entire samples were taken at 0 , 1- and 4-week intervals and extracted in Soxhlet ex tractors. The chemical was extracted with 250 mi of acetone for 24 hr and the volume of acetone was reduced to 100 ml by evaporating under nitrogen stream, partitioned into hexane by adding 200 ml of 2% NaaS04 and extracted three times with hexane. The sample was concentrated and analyzed on a Tracor gas chromatograph equipped with NiM de tector. Similar experiments were done under wet conditions by adding approximately 30 ml of water every day. The ex periments with soil were carried out by placing (25 grams) of Woodburn silt loam containing 10 ppm of Aroclor 1254 in I30-S Lilly cups. The cups were equipped with an alu minum support wrapped with 2-Kimwipes (No. 3115) and placed in the groove of the cup. The Kimwipes acted as a wick for the wet soil loss study. A filter paper (No. 42) was placed on the Kimwipe wick to act as a soil barrier, The samples were taken at 0-, 1-, 2-, 3-, and 4-week inter vals and the PCB was extracted with acetone using the procedure described earlier in this paper. Similar experi ments were performed under wet conditions by adding -50 ml of water at 48-hr intervals through a small funnel inserted in the cup. Alt the above experiments were car ried out in duplicate at room temperature (26*C). Rosults and Discussion The equilibration of Aroclor 1254 in water was achieved in approximately 2 months, the concentration at this point was approximately equal to that at 6 months. It was expected that the total solubilization process would be a slow one, although it takes only a week to achieve a major part of the equilibrium. The total solubility of Aroclor 1254 was --56 ppb. This value is within the range re ported for many chlorinated hydrocarbon insecticides (DDT-type compounds) having similar structure (Bow man el al., 1960; Bigger ct al., 1967; Gunther cl a!., 1967). However, it is significantly lower than the values reported by Zitko 0971), who obtained solubilities in the range of 0.3-3 ppm in "fresh water." These high values ma\ be clue to the presence of foreign materials in the "fresh water." From the intensities of various peaks in the gas chro matogram, it may be said that the water solubility of PCB isomers in genera) decreases with increasing chlorine atoms. The percent decrease in the concentration of Aroclor 1251 by the addition of increasing amounts of adsorbent mutcriul is shown in Figure 2. As expected, sand nnd sili- MUMS 0 0 1 5 4 6 1*0 environments Science A Tcctiriology FI Figure 2. Percent decrease In the concntrallon d Aroclor 1254 by the addition ot Increasing amounts of adsorbent. Original coneantiatlon of Aroclor 1254.56 ppb ea gel adsorb vary little of the chemical whereas iliile clay and Woodburn soil showed greater adsorption. The ad sorption behavior of montmorillonite and kaolinite clays was intermediate between those extremes. The high-ad sorbing capacity of Woodburn soil is tentatively attrib uted to the presence of organic matter. The adsorption equilibrium of PCB on iliile and Woodburn soil was treat ed by a Freundlich-type isotherm Equation 2: -m- - AX,1" (2) where K and n are constants and m the mass of the adsor bent in grams. A typical Freundlich plot is shown in Fig ure 3. The values of K and n for illite clay and for Woodhum soil arc 63.1 and 1.1 and 26.3 and 0.81, respectively. The Freundlich-type isotherm shown in Figure 3 suggests a physical-type adsorption. The Freundlich isotherm for other surfaces is not shown because of their extremely low adsorption capacity. t The loss of Aroclor 1254 from itself is shown in Figure 4. Although the loss of the chemical is rather small at 26*C, it is sulMtlantial at 60*C. It is interesting to note that the weight loss change ie linear as a function of time. Mathe matically. such a behavior is characteristic of diffusion in a plane sheet (Crank. 1967), and could be expressed as in Equation 3: Mass loss (t) TM A F t (3) whore A is the exposed area, t the time, and F the flux defined by Equation 4: Figure 3. Freundlich Isotherm showing adsorption of Aroolor 1254 on iliile clay and Woodburn soli (C in ppb and x/m In ne/o) - *V!P " .-o-____o.. --?........... ...T , loti* 20 i 'O gfr-r/aoj/tm* '' a PCS 1294 Voor Lot* s^SO'C ''xLot 0 6 i (6* gmiiUoirAf** 1 10 20 T.m* |0oy) V(> 3(5 Figure 4. Loss ot Aroclor 1254 from Itself as a function ol time where I) is the diffusion coefficient, {dc/<tx) i the concen tration gradient, and A the thickness of the shed. Since it is not possible to estimate the concentration gradient from the available data, it is difficult to calculate IJ However, from the slope of ihc line (Figure 4), wc can calculate the flux. The calculated value of the flux has been found to be 2.0 X 10 gram day 1 cm' 2 at 26C and 8.6 10' 8 gram day 1 -cm"2 at 60*C. The loss of PCM from a sand surface is shown in Figure 6. As expected, the higher chlorine-containing isomers show the least loss and vice versa. This is mainly due to the fact that the vapor pressure for PCB isomers dccrenses with increasing number of chlorine atoms Vapor loss under wet conditions was similar and comparable in mag nitude. Similar experiments dealing with the Woodburn Figure 5 Loss ol Aroclor 1 ?M from an Ottawa sand noils showed that vapor low of Aroctor 1254 from the soil smlure a.s iicfjlisiWe, The hiRh loss from the send as fompart-d to the soil mny be attributed to the marked dif ference in the absorbing capacity of these two surfaces. The vapor loss behavior of J'CB from sand and soil surfac es is qualitatively similar to DDT (Cliath and Spencer, 1972; Gucnzi and Heard, 1970). The findings of this paper give a qualitative picture of the behavior of the PCB Aroclor 1254 in water, soil, and air. It is difficult to extend these findings directly to the field conditions. There is always a danger of drawing erro neous conclusions. However, some speculations about the environmental significance of the work could be made. In general, once PCB is introduced in the environment, the isomers will be transported in air, soil, water, and biota. The concentration in each phase will highly depend upon the partition coefficients. The water solubility will deter mine the distribution in water. It appears that PCB in pph range could be transported in water. The environ mental contamination of water will highly depend upon tint number of chlorine atoms present in the isomer. The lower chlorine-containing isomers will show higher con centration in water and vice versa. The concentration of PCB in water will be reduced whenever the water comes in contact with a solid surface or particulate matter. The reduction in the concentration of PCB in water will de pend upon the surface characteristics. Such properties as the surface area, organic content of the material, nature of the surface (clay, sand), and pH of the medium will greatly influence the adsorption. The trnni>ort of PCB in air will be governed by temperature, vapor pressure, and the surroundings of the chemical. Sig nificant amounts of HOB could be transported in air at higher temperatures if PCB is present, in the environment where it is loosely bound or adsorbed to a surface. The loss could also be significant if PCB is evaporating from its own surface, especially at higher temperatures. The number of chlorine atoms present in the particular PCB isomer will also influence the loss to a great extent. The larger the number of chlorine atoms present in the isomer, the smaller the loss. However if PCB, on the other hand, in bound strongly to a surface, the vapor loss will be much smaller. In general, the environmental contamination and exposure of PCB to living species will be functions of the surrounding temperature, moisture, surface, and so forth. 'Hie findings of this paper may be summarized as fol lows; The solubility of Aroclor 1254 at room temperature is 'v56 ppb and, although the solubilization is a slow pro cess, the bulk of the process is achieved very rapidly. Aro clor 1254 is readily adsorbed from aqueous solution onto available surfaces. The amount of chemical adsorbed de pends upon the nature of the surface. A sand surface with few sites adsorbs relatively little, whereas a soil surface with high clay and organic content adsorbs a much larger amount. The vapor loss of Aroclor 1254 from its own sur face depends on the area exposed and the temperature. At room temperature it is small, but by approximately dou bling the temperature, loss is enhanced by a factor of 40. The vapor loss from a sand surface is significant whereas from a more tightly bound situation on a soil surface, it is extremely small. The generally more volatile lower chlo rine isomers show a greater loss than those of higher chlorine content. Acknowledgment We thank Susan K, Randall for her technical assistance and Donald A. Griffin for his help in identifying various isomers via mass spectrometry. Literature Cited Bigger, J. W., Dutt, G. R, Higgs, R L., Bull. Em-iron. Contain. Toxicol.. 2,90(1967). Bowman. M. C., Acree, F., Corbett, M. K., J. Agr Food Chcm, 8, 406 0960). Cliath, M. M., Spencer, W. F.. Environ. Set. Technvl, 6, 910 0972). Crank, J., "The Mathematics of Diffusion," Oxford at the Clar endon Press, U.K., 1967. Cuenzi, W. D., Beard, W. E., Soil Sri. Sor. Airier /Vor , 34, 443 (1970). Gunther, K. A., Westlake, W. E., Jagten, P. S.. Ben. Rev.. 20. 1 (1967). Gustafson, C. G., Kd., Proc. Symp. on I'CH's-Still Prevalent Still Persistent, 164th American Chemical Society Meeting, New York, N.Y., September, 1972. Gustafson, C. (., Environ. Set. and Tech , 4. 814 (1970). Haque, R, Coshow, W. it., Environ. Sri 4 Tcchnoi, S, 1,19 (1971). Haque, K., Sexton, H., J. ColloidIntcrfac. Set., 27, 818 (1968). Nelson, N., Chairman, panel on "Hazardous Substance*, Poly chlorinated Biphenyls-- Environmental Impact," Env Hen., 5, 249(1972). Penkal), 1). B,, t.incer. J. Riasci., 20,958 (1970). Zitko, V., Bull. Environ. Contain Toxicol., 5, 279(1971). Received (or review January 29. 1973. Accepted October 3. 1973. The research described here is supported 6\ U.S l\tbhc Health Service Grant No. ES 00040 442 Environmental Sc'ortco 4 Toclmoloyy obi.550 rtUNS