Environ. Sci. Technol. 2001, 35, 3019-3024
Applications of Ultrasound in NAPL Remediation: Sonochemical Degradation of TCE in Aqueous Surfactant Solutions HUGO DESTAILLATS, THOMAS W. ALDERSON II, AND MICHAEL R. HOFFMANN* W. M. Keck Laboratories, California Institute of Technology, 1200 East California Boulevard, Pasadena, California 91125
Surfactant-enhanced pump-and-treat technologies increase the efficiency of nonaqueous-phase liquids (NAPLs) removal from soils. However, high concentrations of surfactants in groundwater impose severe limitations to water treatment. In this paper, we explore the applicability of ultrasonic irradiation as an alternative method for surfactant recovery and contaminant degradation. The combined effects of temperature, initial substrate concentration, and concentration of added surfactant (sodium dodecyl sulfate, SDS) were analyzed for the sonolysis of trichloroethylene (TCE) in batch experiments at an ultrasonic frequency of 500 kHz and 77 W/L applied power density. In the range of 5-30 °C, TCE sonolysis becomes faster at higher temperatures, both in the absence and in the presence of surfactant. This indicates that gas-phase pyrolysis prevails over other chemical reactions in the liquid phase. Inhibition of TCE sonolysis was observed in the presence of surfactant at all SDS concentrations. Changes in the initial TCE concentration (from 250 µM to 1.2 mM) showed no effect on the degradation rates in the presence of SDS. For surfactant levels below its critical micelle concentration (cmc), the inhibition of TCE sonolysis exhibited a highly nonlinear dependence with increasing SDS concentration. A correlation was observed in this range between the relative inhibition of sonolysis and the decreasing surface tension of the solutions. Above the cmc up to an SDS concentration of 5%, the reaction rate decreased less markedly. Micellar sequestration of the contaminant seems to be the main reason for this additional inhibition. Bubble growth prior to collapse may incorporate some of the TCE dissolved in the micelles through their adsorption in the expanding bubble walls, thus partially overcoming the scavenging effect due to micellar entrapment of the contaminant.
Introduction In this paper, we explore the potential applications of ultrasonic irradiation techniques in nonaqueous-phase liquid (NAPL) degradation and surfactant recycling. NAPLs present in the subsurface represent a threat to groundwater supplies and have been the focus of considerable attention over the last 15 years (1-4). NAPLs consist mostly of chlorinated and * Corresponding author phone: (626) 395-4391; fax: (626) 3953170; e-mail:
[email protected]. 10.1021/es0018926 CCC: $20.00 Published on Web 06/09/2001
2001 American Chemical Society
aromatic solvents, which are released in the environment through spillage or leakage from pipelines, storage tanks, or industrial facilities. Although these chemicals are slightly soluble in water, very low concentrations are allowed, thus becoming a long-term source of aquifer contamination. Pump-and-treat technologies can prevent contaminant migration, but have been ineffective in restoring sites to precontamination levels, in part because of NAPL slow dissolution and desorption rates. To remediate NAPLs contaminated sites, it is necessary to achieve a complete mass removal or destruction of the pooled liquid. Several remediation technologies are currently being studied to mitigate this problem through enhanced solubilization in the aqueous phase and/or physical mobilization of NAPLs. Among them, chemical flooding of soils with cosolvents or surfactants(5, 6) and subsurface injection of hot water and steam (7, 8) have proven to be suitable methods that have been widely implemented in the laboratory and the field. Aqueous solutions of surfactant are flushed through the contaminated zone to greatly increase the effective solubility of the NAPL components primarily through micellar solubilization (9-12). To avoid excessive operational costs, surfactant recycling is recommended, which implies a complete elimination of the contaminant from the recycled solution above ground before reinjection (4, 13). Some technologies are considered to be most cost-effective for above ground surfactant regeneration, such as vacuum steam stripping and air stripping-incineration, in combination with micelle ultrafiltration and foam fractionation (4). However, micelle sequestration of contaminants imposes severe limitations to the performance of most treatment methods. Reduced concentrations in the aqueous phase and lower equilibrium vapor pressures result in a dramatic reduction of the contaminant elimination efficiencies when surfactant concentrations higher than the critical micelle concentration (cmc) are employed (14, 15). Microbiological degradation depends strongly on the bioavailability of pollutants, which is also hindered by micelle sequestration in many cases (16, 17), but not always (18). Sonochemical methods have been shown effective for the treatment of a variety of chemical contaminants in aqueous solution (19-26). Ultrasonic irradiation operating at frequencies between 20 and 500 kHz and applied powers between 50 and 100 W is particularly well suited for the degradation of volatile molecules such as chlorinated hydrocarbons (23-25), which partition between the liquid phase and the cavitation bubbles, thus concentrating the target molecule in the sonochemical hot spots. The fast implosion of such bubbles produces a quasi-adiabatic heating of the vapor phase inside the cavity that yields localized high temperatures and pressures, in the range of the thousands of K and hundreds of bar, respectively. Because primary steps in sonolysis take place in the gas phase, the method is a good alternative for eliminating dilute volatile contaminants in the presence of high concentrations of other solutes. The presence of high levels of surfactant may reduce the effectiveness of ultrasonic degradation of chlorinated hydrocarbons, therefore the goal of the present work is to evaluate and quantify the inhibition of sonolysis due to the presence of surfactant. Surface-active molecules interact strongly with cavitation bubbles, both chemically (26) and by influencing the physics of cavitation (27). Here we evaluate the extent of inhibition of the sonolytic degradation of a typical NAPL component, trichloroethylene (TCE), in the presence of varying amounts of the anionic surfactant sodium dodecyl sulfate (SDS). These experiments allow us to assess VOL. 35, NO. 14, 2001 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
3019
SCHEME 1. Phase Equilibria Present in the System for TCE and SDS
the potential of ultrasonic irradiation as an alternative for surfactant recycling technologies in NAPL remediation. They also provide some insight on the nature of sonochemical processes in microheterogeneous media. Scheme 1 illustrates the main phase equilibria involved in this study.
Experimental Procedures Air-saturated aqueous solutions with the appropriate concentrations of TCE (spanning from 40 to 1400 µM) and SDS (in the range 0 to 188 mM) were freshly prepared by diluting a stock solution of TCE (aq) into SDS solutions of varying concentrations (Table 1). Surfactant solutions were prepared gravimetrically, whereas TCE concentrations were determined chromatographically. The instrumentation employed in this study has been described previously (20). The sonochemical batch reactor consists of a 650-mL glass vessel surrounded by a selfcontained water jacket and attached to a piezoelectric transducer (Undatim Ultrasonics). Sonication at 500 kHz was performed operating at 50 W (2 W/cm2). The ultrasonic power input to the reactor was determined calorimetrically (28). The solution was stirred during each experiment and the temperature was kept constant at the values indicated in Table 1 ((0.5 °C) for each run. In all cases the experiments were done under air saturation. Samples (1.0 mL) were collected for analysis at different times through a gastight septum by means of a glass syringe with a stainless steel needle, to avoid possible contamination related to the use of plastic tips in sonicated solutions. The system was tested for possible TCE evaporation losses, which are negligible. The analytical quantification of TCE in each sample was performed using a Hewlett-Packard 1090 series HPLC with a BDS C-18 reverse-phase column at 40 °C. A water/methanol 30:70 elution mixture flowing at 0.7 mL/ min yielded an optimum TCE resolution at a retention time, Tr, of 3.2 min. The critical micelle concentration of SDS, under the conditions of our experiments, was determined conductometrically with a YSI-35 conductance meter, following a procedure described by Shanks and Franses (29). The cmc is determined from the slope break in the conductivitysurfactant concentration (κ - [SDS]) diagram. 3020
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 35, NO. 14, 2001
TABLE 1. Rate Constants (k-TCE), Half-Life Times (t1/2), and Removal Efficiency after 60 min of Sonolysis (θ) for All Experimental Conditions. T, °C
[TCE]0, µM
% SDS
(SDS), mM
k-TCE, min-1
t 1/2, min
θ, %
5
260
0
0
0.0373
18.6
89.3
15
40
0
0
0.0509
13.6
95.3
15
226 216 250 256 279 301
0 0.01 0.1 0.2 1 5
0 0.38 3.78 7.55 37.7 188
0.0456 0.0321 0.0252 0.0203 0.0136 0.0076
15.2 21.6 27.5 34.1 51.0 91.2
93.5 85.4 78.0 70.4 55.8 36.6
15
1410 1213 1165 1486 1264 988
0 0.01 0.1 0.2 1 5
0 0.38 3.78 7.55 37.7 188
0.0337 0.0287 0.024 0.0214 0.0114 0.0069
20.6 24.2 28.9 32.4 60.8 100.5
86.8 82.1 76.3 72.3 49.5 33.9
22
266
0
0
0.0511
13.6
95.3
30
250 250 256 201 237 309
0 0.01 0.1 0.2 1 5
0 0.38 3.78 7.55 37.7 188
0.0602 0.0416 0.0323 0.0313 0.0189 0.0094
11.5 16.7 21.5 22.1 36.7 73.7
97.3 91.8 85.6 84.7 67.8 43.1
Certified 99.5% TCE was obtained from Aldrich. TCE is a common groundwater contaminant with a density of 1.46 g/cm3 and an aqueous solubility of 1100 mg/L (9). High quality SDS (>96%) was obtained from Sigma, and used without further purification. All solutions were prepared using 18.2 MΩ deionized (DI) water.
Results and Discussion The experimental conditions are summarized in Table 1. The temperature dependence of TCE (aq) sonolysis was studied at 5, 15, 22, and 30 °C. The effect of added SDS was evaluated at two different temperatures (15 and 30 °C) and for initial substrate concentrations that can be grouped in
FIGURE 1. Pseudo-first-order kinetic plots for the sonolysis of TCE at 15 °C and [TCE]0 ) 250 µM. b No SDS; 0 0.01% SDS; O 0.1% SDS; 4 0.2% SDS; 3 1% SDS; ) 5% SDS. Straight lines show the best fit in each case. two ranges: [TCE]0 ) (250 ( 50) µM and (1200 ( 200) µM. Experiments were performed in the absence (0.01, 0.1% SDS) and the presence of micelles (1 and 5% SDS), as well as close to cmc (0.2% SDS). Considering that the presence of small amounts of solutes (in the present case, either TCE or impurities present in the SDS) may shift cmc from its literature value (30) of 0.2%, or 8 mM (31, 32), we have experimentally determined the critical micelle concentration in the conditions of our experiments. In the absence of TCE, we have found that cmc ) 7.4 mM, but values measured under [TCE] ) 250 µM and 1.2 mM resulted in lower values, cmc ) 6.5 and 6.3 mM, respectively. The degradation of TCE followed pseudo-first-order kinetics. Figure 1 illustrates the experimental results obtained for reactions at 15 °C and [TCE]0 ) 250 µM, at different SDS concentrations. In each case, the experimental rate constant k-TCE and the half-life time (t1/2) were calculated from the slopes of the depletion curves:
ln
[TCE] ) k-TCE ‚ t; [TCE]0
t1/2 )
ln2 k-TCE
(1)
The experimental errors estimated for k-TCE values in Table 1 are within 2 to 3% of the reported values, according to the standard deviations of the corresponding slopes in Figure 1, evaluated through minimum squares regression. Table 1 also reports the removal efficiency after 1 h of sonolysis, θ:
(
θ ) 100 × 1 -
)
[TCE]60 [TCE]0
(2)
The removal efficiencies varied from > 95% in the absence of surfactant to near 35% for the maximum SDS concentration employed, indicating the magnitude of the inhibition observed in the presence of the surfactant. Sonolysis of TCE (aq). The sonolytical degradation of TCE has been reported recently (33-35). The proposed mechanisms for the sonochemical action are based on gas-phase high-temperature pyrolytic initiation steps involving cleavage of the relatively weak C-Cl bond (338 kJ/mol):
C2HCl3 (g) f •C2HCl2 (g) + Cl• (g)
(3)
The chlorinated organic intermediates generated are unstable and disappear from aqueous solution at faster rates than the substrate molecule (33). This observation is consistent with previous results obtained for the sonolysis of chlorinated methanes (23). The organic radicals undergo further oxidation
FIGURE 2. Temperature dependence of TCE sonolysis in the absence of SDS. [TCE]0 ) 250 µM. both through high-temperature gas-phase reactions and in solution at room temperature, whereas Cl• atoms may recombine to yield Cl2 or reduce to HCl. In the present paper, we have focused on the temperature dependence of the observed TCE (aq) decomposition rates, which increase in the range 5-30 °C, as seen in Figure 2. This positive temperature dependence is opposite to what is commonly reported for sonolysis of nonvolatile compounds (36, 37). The increase in the observed pseudo-first-order reaction rate, k0-TCE, correlates linearly with the temperature dependence of the gas-liquid partition coefficient for TCE, HTCE ) [TCE]V/ [TCE]L (concentrations in mg/L) as measured by Turner et al. (38). Equation 4
k0-TCE (min-1) ) 0.028 + 0.069 HTCE
(4)
corresponds to the best fit for the data in Figure 2. This observation indicates that high-temperature gas-phase chemistry (i.e., pyrolysis reactions inside the cavitation bubbles) most likely determines the observed kinetics of the system. The other possible pathway for sonochemical reactions, the direct oxidation reactions by aqueous ‚OH radicals produced during water molecules sonolyis, appear to be much less significant in this case. The contribution of this latter process to the observed reaction rate can be evaluated as
)
∂[TCE] ∂t
OH
OH • ) kTCE [ OH]ss[TCE]
(5)
9 -1 -1 where kOH is the second-order rate TCE ) 4 × 10 M s constant for the bimolecular reaction (39), and [•OH]ss is the steady-state concentration of •OH radicals in the liquid during sonication. Its value depends strongly on the irradiation conditions, particularly frequency, applied power, temperature, saturation gas, and reactor geometry. It was calculated as [•OH]ss ) 4 × 10-14 M during the sonolysis of nonvolatile and diluted (10 to 20 µM) aromatic substrates at 15 °C in the same experimental conditions (40), and therefore it is possible to use it as a reasonable estimation. Given the higher substrate concentrations employed in the present work and, particularly, the fact that pyrolysis reactions inside the bubbles may consume a significant part of the •OH radicals produced during cavitation, that value should only be considered as an upper limit estimation for [•OH]ss. For an initial concentration of [TCE]0 ) 250 µM, a contribution to the degradation rate of ∂[TCE]/∂t)OH ) 2.4 µM/min can be estimated for the reaction of TCE with aqueous •OH radicals. This accounts for only 20% of the observed initial reaction rate, ∂[TCE]/∂t)0 ) k0-TCE × [TCE]0 ) 11.4 µM/min. Considering that the value employed here for [•OH]ss is an
VOL. 35, NO. 14, 2001 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
3021
FIGURE 3. Effect of added SDS on TCE sonolysis at 15 °C (b) and 30 oC (O). [TCE]0 ) 250 µM. overestimation of the actual amount of •OH radicals in the present case, the contribution of room temperature reactions probably accounts for less than 10% of the observed rates. Sonolysis of TCE in the Presence of SDS. The observed TCE degradation kinetics showed a strong dependence on [SDS]. For [SDS] below the cmc, the observed rate constants, k-TCE, decreased to ∼50% of the value reported in the absence of the surfactant, k0-TCE, with a highly nonlinear dependence on [SDS], as shown in Figure 3. As TCE degradation takes place primarily in the gas phase (vide supra), the lower k-TCE values observed when SDS is added might be related with physical and chemical processes at the bubble-liquid interface. Those processes may include a hindered TCE (aq) mass transfer rate to the bubble as well as chemical reactions of SDS at the bubble interface. Substrate competition has already been mentioned by Drijvers et al. (41) as the reason for the TCE sonolysis inhibition in the presence of a second volatile co-solute (chlorobenzene). Besides, our previous experiments on surfactants sonolysis in aqueous solutions have shown that these molecules are able to accumulate in the surface of imploding bubbles and react through hightemperature pyrolytic reactions (26), thus, the surfactants becoming a potential competitor. A third possible reason for surfactant inhibition of TCE sonolysis can be related to the alterations produced by the reduction of the surface tension in the number and dynamics of the cavitation events. All those inhibitory processes depend strongly on the bubble surface coverage by SDS molecules, i.e., on surfactant partitioning between the bulk liquid and the gas-liquid interface. Figure 4 shows a strong correlation between the inhibition of the observed TCE sonolysis rate constants (k-TCE/ k0-TCE) and the relative reduction in the surface tension of the solution γ/γ0, where γ0 ) 73.5 dynes/cm is the value for pure water (42), as reported by Elworthy and Mysels (43). SDS surface mole fraction data, measured by Bradley et al. (44) employing neutron scattering techniques, correlate closely with the marked reduction of γ/γ0 below cmc: a surface excess of SDS relative to water is observed for increasing bulk SDS concentrations, reaching a complete monolayer close to 4 mM (i.e., 0.5 cmc). Above the cmc, the observed rate constant, k-TCE, drops steadily, but at a lower rate of decrease, as a function of SDS concentrations up to a 5% SDS (190 mM, 25 cmc), as shown in Figures 3 and 4. Surfactant concentrations higher than 5% were not explored, to restrict our study to the domain of spherical SDS micelles in the phase diagram (45). In the presence of micelles, TCE partitions between the aqueous and micellar phases, thus reducing its vapor pressure in the cavitation bubbles. This fact alone should reduce k-TCE considerably, if the micelles were able to sequester TCE from 3022
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 35, NO. 14, 2001
FIGURE 4. Degree of inhibition of TCE sonolysis (k-TCE/k0-TCE). b T ) 15 oC and [TCE]0 ) 250 µM; O T ) 30 oC and [TCE]0 ) 250 µM; 9 T ) 15 oC and [TCE]0 ) 1200 µM; the solid line represents the relative decrease in the surface tension, γ/γ0; the dotted line represents the fraction of the total TCE dissolved in the micellar phase, rMTCE () experimental values). aqueous solution, thus preventing its transfer to the cavitation bubbles. Micelle sequestration of the contaminants is the main reason for reduced efficiencies in currently available remediation technologies, in which the partitioning of the contaminant between the gas and liquid phases (air stripping (14, 15)) or two liquid phases (liquid-liquid extraction (46)) plays a major role. We have determined the extent of micellar partition of TCE at different SDS concentrations by measuring its concentration in solutions of varying surfactant concentrations, equilibrated with excess TCE. Because the aqueous solubility of TCE (8.5 mM (42)) is not affected by the presence of micelles, the excess solubility determined in each case can be attributed to micellar dissolution. Figure 4 shows the determined fraction of the total TCE dissolved in the micellar phase, RMTCE, defined as the ratio
RMTCE )
[TCE]M [TCE]total
(6)
between [TCE]M and [TCE]total, the molar concentration of TCE dissolved in the micelles and in both phases (micelles and water), respectively. To evaluate the effect of micellar partition of TCE in the observed sonolysis rates, the extent of partition should be compared with the degree of sonolysis inhibition observed for [SDS] above cmc. For [TCE]0 ) 250 µM, when [SDS] ) 1% and 5% in weight, nearly 70% and 90% of the total TCE is in the micellar phase, respectively. We determined k-TCE values for those concentrations which correspond to 65% and 37% of the rate constant determined for [SDS] close to cmc (0.2%). This means that most of the substrate, that is dissolved in the micellar phase, is moderately protected from sonochemical activity in the cavitation bubbles. However, a simple micellar sequestration model such as
k-TCE ) k-TCECMC (1 - RMTCE)
(7)
where k-TCECMC is the observed rate constant at the cmc (i.e., immediately before micellization) would predict lower k-TCE values than those reported here. That might be due to direct interaction of micelles (loaded with a substantial proportion of the total TCE) and cavitation bubbles. It is well-known that monomer surfactant molecules are adsorbed into the interface of the cavitation nucleus during bubble growth. Upon acoustic cavitation, the collision of micelles with the expanding bubble wall may contribute to further
FIGURE 5. Effect of TCE initial concentration on TCE sonolysis at 15 °C. b [TCE]0 ) 250 µM; 9 [TCE]0 ) 1200 µM. On the inset, the effect on k0-TCE, including 2 [TCE]0 ) 40 µM. reducing the surface tension (dynamic surface tension), as the micelles are disrupted and surfactant monomers are being incorporated to the interface. This process has been described and quantified by Iliev and Dushkin (47, 48). During transient cavitation induced by ultrasonic irradiation, the bubble radius expands from an initial value of R0 to a critical value of Rmax, just before a resonant coupling with the acoustic field drives the bubble to near adiabatic collapse. Under our experimental conditions, active bubbles have typical R0 values between 0.5 and 2 µm, and can grow to an Rmax of 7 µm. The bubble surface and volume increase on the order of 50 and 350 times, respectively, while incorporating water vapor and volatile solutes through the process of rectified diffusion (24). A fraction of the micelles loaded with TCE can be disrupted and incorporated into the growing interface, operating as carriers that transport the substrate into the gaseous cavity during the expansion phase of the cavitation cycle. This process yields monomer surfactants at the gas-liquid interface and releases TCE into the gaseous cavity. Thus, as the surfactant concentration increases above the cmc, the substrate is increasingly sequestered into the micellar phase but, at the same time, it is less efficiently scavenged from expanding cavitation bubbles as the total concentration of micelles (and their frequency of collision with expanding bubbles) grows. The temperature dependence of TCE sonolysis in SDS solutions follows the previous trend observed for pure aqueous solutions. For each SDS concentration, the rate constant values determined at 30 °C are higher than those at 15 °C, as seen in Figure 3. Moreover, the degree of inhibition of TCE sonolysis at each surfactant concentration, k-TCE/ k0-TCE, is temperature-independent, as observed in Figure 4. For SDS concentrations above cmc our data show a decreasing temperature dependence as free TCE (aq) is sequestered in the micellar phase. Two separate experimental series of k-TCE data are shown for different [TCE]0 in Figure 5. TCE sonolysis appears to be insensitive to its initial concentration, except in the absence of surfactant. The inset of Figure 5 shows the dependence of k0-TCE with [TCE]0 in pure aqueous solutions (i.e., without SDS). Lower sonolysis rates are observed when [TCE]0 increases, probably due to competition reactions with sonolytical intermediates or byproducts. The addition of the surfactant seems to eliminate the difference in reactivity for variable [TCE]0, indicating that the overall reaction rates depend on the relative competition of SDS and TCE molecules, in that case. The lower degree of sonolysis inhibition observed in Figure 4 for the highest [TCE]0 can also be explained with the same argument, as the molar ratio
[TCE]/[SDS] is significantly higher for [TCE]0 ) 1.2 mM only at lower SDS concentrations. As [SDS] increases, the relative inhibition of k-TCE/k0-TCE shows a lower sensitivity to [TCE]0. In summary, we have evaluated the combined effects of temperature, initial substrate concentration, and concentration of surfactant on TCE (aq) sonolysis rates. In the absence and presence of SDS, sonolysis was faster at higher temperatures, suggesting the preeminence of gas phase pyrolytic mechanisms. We have quantified the degree of inhibition of TCE sonolysis due to the presence of SDS at different temperatures and initial substrate concentrations. We observed inhibition not only above cmc but also in the absence of micelles. That is attributable to chemical and physical effects originated in the presence of monomer SDS molecules in the bubble interface. In the least favorable case ([SDS] ) 5%), a concentration of surfactant almost 3 orders of magnitude higher than that of the substrate reduces the observed reaction rate for substrate degradation by only 20% relative to its value in pure water. Therefore, sonolysis of volatile contaminants such as TCE is, in principle, a possible alternative to groundwater treatment in the presence of high concentrations of surfactant.
Acknowledgments Financial support provided by the Department of Energy (DOE 1963472402) and the U.S. Navy (N 47408-99-M-5049) is gratefully acknowledged. The authors also thank Dr. A. J. Colussi for helpful discussions.
Literature Cited (1) Pankow, J. F.; Cherry, J. A. Dense Chlorinated Solvents; Waterloo Press: Portland, OR, 1996. (2) National Research Council. Alternatives for Groundwater Cleanup; National Academy Press: Washington, DC, 1994. (3) Sabatini, D. A.; Knox, R. C. Subsurface Contaminants; American Chemical Society: Washington, DC, 1992. (4) Sabatini, D. A.; Knox, R. C.; Harwell, J. H. Surfactant-Enhanced Subsurface Remediation. Emerging Technologies; American Chemical Society: Washington, DC, 1994. (5) Lowe, D. F.; Oubre, C. L.; Ward, C. H., Eds. Surfactants and Cosolvents for NAPL Remediation. A Technology Practices Manual; CRC-Lewis: Boca Raton, FL, 1999. (6) Fiorenza, S.; Miller, C. A.; Oubre, C. L.; Ward, C. H. NAPL Removal: Surfactants, Foams and Microemulsions; CRCLewis: Boca Raton, FL, 2000. (7) Wickramanayake, G. B.; Gavaskar, A. R., Eds. Physical and Thermal Technologies: Remediation of Chlorinated and Recalcitrant Compounds; Battelle Press: Columbus, OH, 2000. (8) Imhoff, P. T.; Frizzell, A.; Miller, C. T. Environ. Sci. Technol. 1997, 31, 1615-1622. (9) Mayer, A.; Zhong, L.; Pope, G. Environ. Sci. Technol. 1999, 33, 2965-2972. (10) Sahoo, D.; Smith, J. Environ. Sci. Technol. 1997, 31, 1910-1915. (11) Sahoo, D.; Smith, J.; Imbrigiotta, T.; McLellan, H. Environ. Sci. Technol. 1998, 32, 1686-1693. (12) Okuda, I.; McBride, J.; Gleyzer, S.; Miller, C. Environ. Sci. Technol. 1996, 30, 1852-1860. (13) Lowe, D. F.; Oubre, C. L.; Ward, C. H. Reuse of Surfactants and Cosolvents for NAPL Remediation; CRC-Lewis: Boca Raton, FL, 2000. (14) Choori, U.; Scamehorn, J.; O’Haver, J.; Harwell, J. Ground Water Monit. Rem. 1998, 18, 157-165. (15) Lipe, K.; Sabatini, D.; Hasegawa, M.; Harwell, J. Ground Water Monit. Rem. 1996, 16, 85-92. (16) Laha, S.; Luthy, R. Environ. Sci. Technol. 1991, 25, 1920-1930. (17) Laha, S.; Luthy, R. Biotechnol. Bioeng. 1992, 40, 1367-1380. (18) Liu, Z.; Jacobson, A.; Luthy, R. Appl. Environ. Microbiol. 1995, 61, 145-151. (19) Weavers, L.; Malmstadt, N.; Hoffmann, M. Environ. Sci. Technol. 2000, 34, 1280-1285. (20) Joseph, J.; Destaillats, H.; Hung, H.; Hoffmann, M. J. Phys. Chem. A 2000, 104, 301-307. (21) Kang, J.; Hoffmann, M. Environ. Sci. Technol. 1998, 32, 31943199. (22) Weavers, L.; Ling, F.; Hoffmann, M. Environ. Sci. Technol. 1998, 32, 2727-2733. VOL. 35, NO. 14, 2001 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
3023
(23) Hua, I.; Hoffmann, M. Environ. Sci. Technol. 1996, 30, 864-871. (24) Colussi, A.; Hung, H.; Hoffmann, M. J. Phys. Chem. A 1999, 103, 2696-2699. (25) Hung, H.; Hoffmann, M. J. Phys. Chem. A 1999, 103, 27342739. (26) Destaillats, H.; Hung, H.; Hoffmann, M. Environ. Sci. Technol. 2000, 34, 311-317. (27) Ashokkumar, M.; Hall, R.; Mulvaney, P.; Grieser, F. J. Phys. Chem. B 1997, 101, 10845-10850. (28) Mason, T. J. Practical Sonochemistry: User’s guide to applications in chemistry and chemical engineering; Ellis Horwood: London, 1991. (29) Shanks, P. C.; Franses, E. I. J. Phys. Chem. 1992, 96, 1794-1805. (30) Israelachvili, J. Intermolecular & Surface Forces; Academic Press: London, 1992. (31) Chauhan, M. S.; Kumar, G.; Kumar, S.; Chauhan, S. Colloids Surf. A 2000, 166, 51-57. (32) Bakshi, M. S. Bull. Chem. Soc. Jpn. 1996, 69, 2723-2729. (33) Drijvers, D.; DeBaets, R.; DeVisscher, A.; VanLangenhove, H. Ultrason. Sonochem. 1996, 3, S83-S90. (34) Suri, R.; Paraskewich, M.; Zhang, Q. Environ. Eng. Sci. 1999, 16, 345-352. (35) Cyr, P.; Paraskewich, M.; Suri, R. Water Sci. Technol. 1999, 40, 131-136. (36) Thompson, L.; Doraiswamy, L. Ind. Eng. Chem. Res. 1999, 38, 1215-1249. (37) Suslick, K. S.; Mdleleni, M. M.; Ries, J. T. J. Am. Chem. Soc. 1997, 119, 9303-9304.
3024
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 35, NO. 14, 2001
(38) Turner, L.; Chiew, Y.; Ahlert, R.; Kosson, D. AICHE J. 1996, 42, 1772-1788. (39) Buxton, G.; Greenstock, C.; Helman, W.; Ross, A. J. Phys. Chem. Ref. Data 1988, 17, 513-886. (40) Destaillats, H.; Colussi, A. J.; Joseph, J. M.; Hoffmann, M. R. J. Phys. Chem. A 2000, 104, 8930-8935. (41) Drijvers, D.; van Langenhove, H.; Kim, L.; Bray, L. Ultrason. Sonochem. 1999, 6, 115-121. (42) Lide D. R., Ed. Handbook of Chemistry & Physics; CRC: Boca Raton, FL, 1998. (43) Elworthy, P. H.; Mysels, K. J. J. Colloid Interface Sci. 1966, 21, 331-347. (44) Bradley, J.; Lee, E.; Thomas, R.; Wilatt, A.; Penfold, J.; Ward, R.; Gregory, D.; Waschkowski, W. Langmuir 1988, 4, 821-826. (45) Laughlin, R. G. The Aqueous Phase Behavior of Surfactants; Academic Press: London, 1994. (46) Hasegawa, M.; Sabatini, D.; Harwell, J. J. Environ. Eng.-ASCE 1997, 123, 691-697. (47) Iliev, T.; Dushkin, C. Colloid Polym. Sci. 1992, 270, 370-376. (48) Dushkin, C.; Iliev, T.; Radkov, Y. Colloid Polym. Sci. 1995, 273, 370-378.
Received for review November 20, 2000. Revised manuscript received April 3, 2001. Accepted April 19, 2001. ES0018926
