Envlron. Sci. Technol. W Q P , 26, 2381-2387
Burford, J. R.; Bremner, J. M. Soil Biol. Biochem. 1975,
Yoshinari, T.; Hynes, R.; Knowles, R. Soil Biol. Biochem. 1977,9,177-183.
7,389-394.
Stumm, W.; Morgan, J. J. Aquatic Chemistry; John Wiley and Sons: New York, 1981. SAS User's Guide: Statistics; SAS Institute, Inc.: Cary, NC, 1985. Morris, J. T.;Whiting, G. J.; Chapelle, F. H. Environ. Sci. Technol. 1988,22,832-836. Slater, J. M.;Capone, D. G. Appl. Environ. Microbiol. 1987,
Bijay-Singh; Ryden, J. C.; Whitehead, D. C. Soil Biol. Biochem. 1988,20,737-741. Van Cleemput, 0.; Patrick, W. H., Jr. Soil Biol. Biochem.
53,1292-1297. Smith, R.L.;Duff, J. H. Appl. Environ. Microbiol. 1988, 54, 1071-1078. Ekpete, D.M.; Cornfield, A. H. Nature 1965,208,1200. Bowman, R.A.;Focht, D. D. Soil Biol. Biochem. 1974,6, 297-301. decatanzaro, J. B.; Beauchamp, E. G. Biol. Fertil. Soils 1985,1,183-187. Myrold, D. D.; Tiedje, J. M. Soil Biol. Biochem. 1985,17, 819-822. Drury, C. F.;McKenney, D. J.; Findlay, W. I. Soil Biol. Biochem. 1991,23,751-755.
246-251.
Bouasaid, F.; Martin, G.; M o N ~J.;, Collin, J. J.; Landreau, A.; Talbo, H. Environ. Technol. Lett. 1988,9, 803-816. Gayle, B. P.; Boardman, G. D.; Sherrard,J. H.; Benoit, R. E. J. Environ. Eng. 1989,115,930-943. Manoharan, R.; Liptak, S.; Parkinson, P.; Mavinic, D.; Randall, C. W. Environ. Technol. Lett. 1989,10,701-716. Hiscock, K. M.; Lloyd, J. W.; Lerner, D. N. Water Res. 1991, 9,1099-1111.
1974,6,85-88.
Koakinen, W. C.; Keeney, D. R. Soil Sci. SOC.Am. J. 1982, 46, 1165-1167.
Waring, S.A.;Gilliam, J. W. Soil Sci. SOC.Am. J. 1983,47, Weier, K. L.; Gilliam, J. W. Soil Sci. SOC.Am. J . 1986,50, 1202-1205.
Parkin, T. B.; Sexstone, A. J.; Tiedje, J. M. Appl. Environ. Microbiol. 1985,49, 1053-1056. Delwiche, C. C.; Bryan, B. A. Annu. Rev. Microbiol. 1976, 30,241-262.
Meiberg, J. B. M.; Bruinenberg, P. M.; Harder, W. J. Gen. Microbiol. 1980,120,453-463. Lloyd, D.; Boddy, L.; Daviea, K. J. P. FEMS Microbiol.Ecol. 1987,45,185-190. Bengtason, G.; Annadotter, H. Appl. Environ. Microbiol. 1989,55,2861-2870. Christensen,P. B.; Nielsen, L:P.; Revabech,N. P.; Sorensen, J. Appl. Environ. Microbiol. 1989,55, 1234-1241. Received for review April 14,1992.Revised manuscript received July 20, 1992. Accepted July 30, 1992.
Sensing the Fugacity of Hydrophobic Organic Chemicals in Aqueous Systems John Resendes, Wan Ylng Shlu, and Donald Mackay"
Department of Chemical Engineering and Applied Chemistry, University of Toronto, Toronto, Ontario, Canada M5S lA4
It is suggested that sensing the fugacity of hydrophobic chemicals in aqueous systems by measuring their concentrations in an equilibrated air headspace can provide valuable information about the nature and extent of interactions between the chemicals and dissolved and particulate phases present in the water. An experimental air-water closed system is described into which sorbing materials can be titrated and the fugacity response determined by continuous air circulation with periodic gas sampling and GC analysis. The system has been used to measure the cosolvency of octanol and the sorption of five chlorobenzenes and a PCB to humic materials and sediment. The results are in accord with theory based on established partitioning principles, with organic carbon partition coefficients varying from 10 to 100% of the octanol-water partition coefficient. No "solids' concentration effect" was detected.
Introduction Hydrophobic chemicals, notably the organochlorines, when present in aquatic systems may bioaccumulate to high concentrations and cause toxic effects and are subject to evaporation, sedimentation, and degrading reactions. These processes are profoundly affected by whether the chemical is dissolved or sorbed to particulate matter and the extent to which its activity is modified by interactions with other dissolved materials, such as the naturally occurring fulvic and humic acids, and electrolytes. Although measurements of total concentration of the chemical are relatively straightforward, it is difficult to discriminate between dissolved and sorbed chemical by techniques such as filtering or centrifuging because of uncertainties about 0013-936X/92/0926-2381$03.00/0
the particle size "cutoff". Indeed, it is possible that no distinct discrimination is possible between dissolved and sorbed states because there may be a continuum of particulate matter ranging from truly dissolved low molecular weight fulvic acids to filterable particles of humin. Conventional measurements of the "dissolved" coneeahdcion may affect particle aggregation, disturb the sorption process, or displace the equilibrium during separation of the phases. Accordingly, it is desirable to supplement conventional approaches with a nondisturbing or noninvasive analytical technique to sense the condition of the chemical. Such a technique is headspace analysis, which is routinely used for analysis of volatile organic chemicals and has been employed to probe or sense the condition of organic chemicals in aqueous solutions by Yin and Hassett ( I ) , Sproule et al. (2))and Perlinger (3). In this study, we describe a headspace system that is essentially a modification of one devised by Hussam and Carr (4) and similar to one used for environmental studies by Perlinger (3). Measurements are made of changes in the air-phase concentration (and hence the partial pressure or fugacity) of a chemical in equilibrium with an aqueous solution into which potentially interacting materials such as cosolvents or humic materials are titrated. From the response of the fugacity to this titration, the nature and extent of the interactions that occur in solution can be deduced. The approach is explored and illustrated by examining the response of the selected chlorobenzenes to the addition of octanol as cosolvent and humic acids and sediment as sorbents. Of particular interest is the ability of the system to probe the "solids' concentration effect" first noted by O'Connor and Connolly (5) and discussed by DiToro (6)and others.
0 1992 American Chemical Society
Environ. Sci. Technol., Vol.
26, No. 12, 1992
2381
It has been observed that when sorbent concentrations increase beyond levels causing about half the sorbate to be sorbed, the amount of sorbate in solution remains relatively constant a t about half. DiToro has interpreted this as being caused by a reduction in the partition coefficient to maintain about half of the sorbate in solution. Several workers have reported this effect and it appears to be real, but is puzzling from a thermodynamic viewpoint and is still controversial. Most studies have involved “disturbing” techniques, although Perlinger (3) has detected the effect for sorption on alumina by headspace analysis. No headspace analysis studies are known in which this effect is observed with organic matter. A simple test is to observe the chemical’s fugacity drop as organic matter is titrated into the solution. If there is no solids’ concentration effect, the drop will continue indefinitely because of the constant partition coefficient. If the effect occurs, there should be a drop in fugacity to approximately 50% of the initial sorbent-free value then no further drop as more sorbent is added.
Theoretical Principles The system described later establishes an equilibrium mixture of known composition of air, water, volatile solute, and sorbent or cosolvent. Concentrations of the solute in the air are measured by gas chromatography (GC).The concentration can be converted into a partial pressure (Pi) or fugacity f A i in the air phase. This fugacity can be equated to the fugacity of the chemical in the aqueous phase fwi. fAi = Pi = fwi = xiyifR, = HiCw; (1) where xi is the mole fraction of the solute chemical in the water, yiis its activity coefficient, fRiis the reference fugacity, i.e., that of the pure liquid chemical at the system temperature and equal to the liquid vapor pressure, Cwi is the dissolved concentration of solute in water (mol/m3), and Hi is the Henry’s law constant (Pa m3/mol). Cwi is x i / u w , where vw is the molar volume of the solution (approximately 18 X lo* m3/mol);thus Hiis also ylfRiw. The dimensionless air-water partition coefficient KAW is given by KAW = CAi/Cwi = (PAi/RT)/Cwi = H;/RT = yifRivw/RT (2) where R is the gas constant (8.314 Pa m3/mol K) and T is the absolute temperature. It follovrs that if Hi or KAW is known, and C, or PAiis measured, then Cwi and xi can be deduced. We suggest that two general types of association of chemical with other materials in the water column may occur: “second-phase” and “dissolved-phase”association, but no claim is made that all behavior will fall into these classes. Second-Phase Association. Here a second solid or liquid phase is added to the water, into which the chemical sorbs or dissolves. The volume of water is denoted by VW, the volume of air by VA,and the volume of sorbing phase by Vs. The total amount of solute in the system before and after addition of the sorbing phase will be constant; thus VwCwo + VAoCAo = VwCw + VACA+ VsCs (3) where Cwo and Cw are the initial and final concentrations in water, VAoand VAare the initial and final volumes of air, CAoand CA are the initial and final concentrations in the air, and Cs is the concentration in the sorbing phase. If the sorbent displaces an equal volume of air, VA will be ( VAo- Vs). If we invoke partition coefficients such that 2382
Envlron. Sci. Technol., Vol. 26, No. 12, 1992
Cs/Cw is Ksw and CA/Cw and C~o/Cwoare KAw, then
+ VAO)= CAO(VW/KAW CA[VW/KA +~(VAO- Vs) + VSKSW/KAW~ (4) and cAO _ -1+
CA
vS(KSW - K A W ) Vw + KAWVAO
(5)
When KsW>> KAW and Vw >> VA&AW, this reduces to CAO/CA = 1 + VsKsw/Vw = 1 + FKsw (6) where F is the ratio Vs/V,. If CAo is measured, followed by CA as a function of F, a plot of CAo/CAwill be linear with an intercept of 1and a slope of Ksw. The partition coefficient, Ksw, can then be determined without analyzing or disturbing the water phase, provided that there are no other sorbing phases or interfaces present. Obviously, it is necessary to add sufficient sorbent that CAO/ CA increases significantly above experimental error. A second approach, as exemplified by Chiou et al. (7), i s to maintain Cw at the solubility limit C% (with an excess of solid solute present), and the total (dissolved plus sorbed) concentration of the solute CT is measured as a function of F. The total amount in solution is then CT(vW + vs) =
vWc$ + v s c s = cb(vW + V&sW)
(7)
then
cT/cf$
(VW +
v&SW)/(vW
+ VS) =
1 + F’(Ksw
- 1)
(8)
where F’ is the volume fraction of the sorbent and approaches F, the volume ratio, when Vw >> V,. The partition coefficient, Ksw, can again be measured without separating the water phase. Inherent in this approach is the assumption that the sorbent has no effect on the solubility of the chemical or on the crystal structure of the solid solute. If the solute is liquid, the excess solute may be diluted and “undersaturation” may occur. Dissolved-Phase Association. If a truly dissolved “cosolvent” such as an alcohol is added and no second cosolvent phase is present, the effect is to change y from its initial value yo to a new (probably lower) value yM reflecting greater ideality. This causes a corresponding change in the air-water partition coefficient (KAw) from its pure water value of yOfRuw/RT to a new air-solution value (KAM)of ~ & R u M / R T . A mass balance on this system before and after the addition of a volume of cosolvent (VS) gives as before CwoVw + CAOVAO= Cw(Vw + Vs)
+ CA(VAO - Vs)
(9)
Substituting KAW = CAo/Cw0 and KAM= CA/CWgives VS)/KAMI + (VAO - VS)] [(vW/KAW) + VAOl
-CAO- --[[(VW -k CA
(10)
If most of the chemical is in the water phase, i.e., VW/KAW >> VAo, this reduces to CAO/CA = [(VW + Vd/VWl(KAW/KAM) = (1 + F)WAW/&M)(11)
At low concentrations, vw and U M will be approximately equal, F
