Influence of Rhamnolipids and Triton X-100 on the Desorption of Pesticides from Soils JUAN C. MATA-SANDOVAL, ² JEFFREY KARNS, ‡ AND ALBA TORRENTS* ,² Environmental Engineering Program, Department of Civil and Environmental Engineering, University of Maryland at College Park, College Park, Maryland 20742, and Soil Microbial Systems Laboratory, Room 140 Building 001 BARC-West, 10300 Baltimore Avenue, Beltsville, Maryland 20705-2350 A rhamnolipid biosurfactant mixture produced by P. aeruginosa UG2 and the surfactant Triton X-100 were tested for their effectiveness of enhancing the desorption of trifluralin, atrazine, and coumaphos from soils. Sorption of both surfactants by the soils was significant and adequately described by the Langmuir-type isotherm. Values of maximum sorption capacity (Q max ) and Langmuir constant (K lang ) did not correlate with the amount of soil organic matter. Our results indicate that clay surfaces play an important role in the sorption of surfactants. When surfactant dosages were high enough to reach soil saturation and maintain an aqueous micellar phase, pesticide desorption was only enhanced. At dosages below soil saturation, surfactants sorbed onto soil, increasing its hydrophobicity and enhancing the sorption of the pesticides by a factor of 2. Similar values of water-soil partition coefficients (K sol * ) for aged and fresh added pesticides to soils indicate that the aging process used did not significantly alter the capability of either surfactant to desorb the pesticides. A model able to estimate equilibrium distributions of organic compounds in soil- aqueous-micellar systems was tested against experimental results. The determined organic carbon partition coefficients, K oc values, indicate that, on a carbon normalized basis, sorbed Rh-mix is a much better sorbent of pesticides than TX-100 or soil organic matter. These results have significant implications on determining the effectiveness of surfactants to aid soil remediation technologies. Introduction The potential for contamination of groundwater supplies by pesticides, as a consequence of agricultural usage, is of growing concern. Pesticides present challenges to almost any technological resource applicable to the clean up of contaminated soils. Unlike oil derived products, such as PAHs, which include a limited variety of compounds, pesticides span a wide range of classes of molecular structures with extremely variable environmental behavior. Even though agricultural usage is probably responsible for most of pesticide contamination, point sources amenable to clean up are associated with rinsate wastes from farming or pesticide spills from formulating retailers (1). For example, contamination by DDT, toxaphene, atrazine, parathion, trifluralin, and linuron have all been found at high concen- trations at an agricultural aviation facility (2). Organic compounds freshly added to soils are bound almost exclusively to the soil particle surface (adsorption) and their desorption is almost complete after a short period of time. On the other hand, pollutants that have been in contact with soil for extended periods of time experience a slow sorption process that leads to a sorbed fraction that becomes resistant to desorption and biodegradation. It is most accepted that aged pollutants in soils are likely partitioned into soil organic matter (SOM) or entrapped in soil micropores that are inaccessible to microbial degraders (3, 4). Conventional pump-and-treat technologies are widely used for the remediation of contaminated soils and ground- water. However, the cleaning efficiency of these methods is limited by the low solubility and high interfacial tension of the hydrophobic organic compounds (HOC) in water (5). Recent research has evaluated the solubilization enhance- ment of HOC by micellar surfactant solutions for applications in soil washing or flushing treatment systems (6). The ability of a surfactant to solubilize an HOC from soil depends on (1) the interaction of the compound with the surfactant monomers in the true aqueous phase, (2) the sorption of the compound onto the soil, (3) the sorption of the surfactant onto the soil and its effect on increasing the overall hydrophobicity of the soil, and (4) the interaction of the compound with the surfactant micellar pseudophase. It is generally accepted that surfactant monomers present at concentrations below the CMC are responsible for most of surfactant interaction with soil. Once the CMC is attained, the sorption of micelles onto surfaces directly from bulk solution has been reported to be insignificant (7, 8). Solu- bilization enhancement of HOCs by aqueous surfactant solutions is mostly a micellar phenomenon. In the presence of soil, surfactant solutions will only be able to solubilize sorbed pollutants if the micellar phase is still attained after monomers have reached sorption equilibrium with the soil. Additionally, some researchers have reported that surfactants sorbed onto soil enhance the soils’s capacity for sorbing organic pollutants (8, 9). According to Laha and Luthy (7), partitioning into soil organic matter is the predominant mechanism of surfactant sorption. It is commonly accepted that only pollutants in soil solution are available for microbial action (10). Since pesticides are generally known for their low aqueous solubility and a tendency to stay sorbed in soil, surfactants in amounts greater than their critical micellar concentration (CMC) can enhance their aqueous solubility and be used in soil washing or soil flushing. Although the reported effects of surfactants on biodegradation enhancement of HOCs have been positive in some cases (11-14), in others either no effect or strong inhibition of microbial activity against the target pollutants has been observed (7, 12, 15-17). Biological surfactants have an advantage over synthetic surfactants in that they are less likely to display cellular toxicity (18). While some work has focused on the capability of biosurfactants to mobilize hydrocarbons, little is known of their ability to mobilize pesticides. They have also not been directly compared to chemical surfactants for the same purpose. This work focuses on comparing the ability of a rhamnolipid mixture produced by P. aeruginosa UG2 (Rh-mix) and the synthetic nonionic surfactant, Triton X-100 (TX-100), to desorb pesticides. The * Corresponding author: phone: (301)405-1979; fax: (301)405- 2585; e-mail: [email protected]. ² University of Maryland at College Park. ‡ Soil Microbial Systems Laboratory. Environ. Sci. Technol. 2002, 36, 4669-4675 10.1021/es011260z CCC: $22.00 2002 American Chemical Society VOL. 36, NO. 21, 2002 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 9 4669 Published on Web 09/25/2002
Transcript
Influence of Rhamnolipids andTriton X-100 on the Desorption ofPesticides from SoilsJ U A N C . M A T A - S A N D O V A L , †
J E F F R E Y K A R N S , ‡ A N D A L B A T O R R E N T S * , †
Environmental Engineering Program, Department of Civiland Environmental Engineering, University of Maryland atCollege Park, College Park, Maryland 20742, and SoilMicrobial Systems Laboratory, Room 140 Building 001BARC-West, 10300 Baltimore Avenue,Beltsville, Maryland 20705-2350
A rhamnolipid biosurfactant mixture produced by P.aeruginosa UG2 and the surfactant Triton X-100 weretested for their effectiveness of enhancing the desorptionof trifluralin, atrazine, and coumaphos from soils. Sorptionof both surfactants by the soils was significant and adequatelydescribed by the Langmuir-type isotherm. Values ofmaximum sorption capacity (Qmax) and Langmuir constant(Klang) did not correlate with the amount of soil organicmatter. Our results indicate that clay surfaces playan important role in the sorption of surfactants. Whensurfactant dosages were high enough to reach soil saturationand maintain an aqueous micellar phase, pesticidedesorption was only enhanced. At dosages below soilsaturation, surfactants sorbed onto soil, increasingits hydrophobicity and enhancing the sorption of thepesticides by a factor of 2. Similar values of water-soilpartition coefficients (Ksol
* ) for aged and fresh addedpesticides to soils indicate that the aging process useddid not significantly alter the capability of either surfactantto desorb the pesticides. A model able to estimateequilibrium distributions of organic compounds in soil-aqueous-micellar systems was tested against experimentalresults. The determined organic carbon partition coefficients,Koc values, indicate that, on a carbon normalized basis,sorbed Rh-mix is a much better sorbent of pesticides thanTX-100 or soil organic matter. These results have significantimplications on determining the effectiveness of surfactantsto aid soil remediation technologies.
IntroductionThe potential for contamination of groundwater supplies bypesticides, as a consequence of agricultural usage, is ofgrowing concern. Pesticides present challenges to almostany technological resource applicable to the clean up ofcontaminated soils. Unlike oil derived products, such as PAHs,which include a limited variety of compounds, pesticidesspan a wide range of classes of molecular structures withextremely variable environmental behavior. Even thoughagricultural usage is probably responsible for most ofpesticide contamination, point sources amenable to clean
up are associated with rinsate wastes from farming orpesticide spills from formulating retailers (1). For example,contamination by DDT, toxaphene, atrazine, parathion,trifluralin, and linuron have all been found at high concen-trations at an agricultural aviation facility (2).
Organic compounds freshly added to soils are boundalmost exclusively to the soil particle surface (adsorption)and their desorption is almost complete after a short periodof time. On the other hand, pollutants that have been incontact with soil for extended periods of time experience aslow sorption process that leads to a sorbed fraction thatbecomes resistant to desorption and biodegradation. It ismost accepted that aged pollutants in soils are likelypartitioned into soil organic matter (SOM) or entrapped insoil micropores that are inaccessible to microbial degraders(3, 4).
Conventional pump-and-treat technologies are widelyused for the remediation of contaminated soils and ground-water. However, the cleaning efficiency of these methods islimited by the low solubility and high interfacial tension ofthe hydrophobic organic compounds (HOC) in water (5).Recent research has evaluated the solubilization enhance-ment of HOC by micellar surfactant solutions for applicationsin soil washing or flushing treatment systems (6). The abilityof a surfactant to solubilize an HOC from soil depends on(1) the interaction of the compound with the surfactantmonomers in the true aqueous phase, (2) the sorption of thecompound onto the soil, (3) the sorption of the surfactantonto the soil and its effect on increasing the overallhydrophobicity of the soil, and (4) the interaction of thecompound with the surfactant micellar pseudophase. It isgenerally accepted that surfactant monomers present atconcentrations below the CMC are responsible for most ofsurfactant interaction with soil. Once the CMC is attained,the sorption of micelles onto surfaces directly from bulksolution has been reported to be insignificant (7, 8). Solu-bilization enhancement of HOCs by aqueous surfactantsolutions is mostly a micellar phenomenon. In the presenceof soil, surfactant solutions will only be able to solubilizesorbed pollutants if the micellar phase is still attained aftermonomers have reached sorption equilibrium with the soil.Additionally, some researchers have reported that surfactantssorbed onto soil enhance the soils’s capacity for sorbingorganic pollutants (8, 9). According to Laha and Luthy (7),partitioning into soil organic matter is the predominantmechanism of surfactant sorption.
It is commonly accepted that only pollutants in soilsolution are available for microbial action (10). Sincepesticides are generally known for their low aqueous solubilityand a tendency to stay sorbed in soil, surfactants in amountsgreater than their critical micellar concentration (CMC) canenhance their aqueous solubility and be used in soil washingor soil flushing. Although the reported effects of surfactantson biodegradation enhancement of HOCs have been positivein some cases (11-14), in others either no effect or stronginhibition of microbial activity against the target pollutantshas been observed (7, 12, 15-17). Biological surfactants havean advantage over synthetic surfactants in that they are lesslikely to display cellular toxicity (18). While some work hasfocused on the capability of biosurfactants to mobilizehydrocarbons, little is known of their ability to mobilizepesticides. They have also not been directly compared tochemical surfactants for the same purpose. This work focuseson comparing the ability of a rhamnolipid mixture producedby P. aeruginosa UG2 (Rh-mix) and the synthetic nonionicsurfactant, Triton X-100 (TX-100), to desorb pesticides. The
† University of Maryland at College Park.‡ Soil Microbial Systems Laboratory.
Environ. Sci. Technol. 2002, 36, 4669-4675
10.1021/es011260z CCC: $22.00 2002 American Chemical Society VOL. 36, NO. 21, 2002 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 9 4669Published on Web 09/25/2002
study also looks at the effect of soil composition, physical-chemical properties of the pesticides, and the aging ofpesticides onto soils on the overall capability of the surfactantto desorb the pesticide.
Materials and MethodsMicroorganism and Rhamnolipid Synthesis. P. aeruginosaUG2 was provided by the Department of EnvironmentalBiology at the University of Guelph, Ontario, Canada.Cultivation conditions for P. aeruginosa UG2, and rham-nolipid extraction/purification from the culture broth, havebeen described previously (19, 20).
Materials. Rh-mix produced by P. aeruginosa UG2 hasbeen previously identified as a mixture of Rh2C10C10 dirham-nolipid (60% w/w), RhC10C10 monorhamnolipid (21%), withthe remaining 19% composed of a mixture of Rh2C10C12
dirhamnolipid and Rh2C10C12-H2 “dehydro-dirhamnolipid”.An average molecular weight of 624 and a CMC of 54 mg/L(86.5 µM) were obtained experimentally for this preparation(19, 21).
TX-100, a heterogeneous nonionic octylphenol ethoxylatesurfactant was obtained from Pierce Chemical Co., Rockford,IL (100 g/L solution, mw ) 625). A CMC value of 125 mg/L(200 µM) was obtained experimentally. Trifluralin (98.7%)was obtained from Eli Lilly & Co. Indianapolis, IN, atrazine(94% w/w) from Ciba Geigy Corp., Greensboro, NC, andcoumaphos (97.1%) from Bayvet Division of Cutter Labo-ratories Inc., Shawnee, KS. All the chemicals were used assupplied. Beltsville A soil (BVA), Hagerstown A (HTA), andHagerstown B (HTB) soils were obtained from the soilcollection at the U.S. Department of Agriculture (USDA-BARC), Beltsville, MD. The same soils were used in ourprevious study (22) and characteristics are summarized inTable 1.
Soil Preparation. Soils were ground and passed througha 1.0 mm stainless steel sieve to obtain the fraction smallerthan 1.0 mm. Soils were amended as described in Mata-Sandoval et al. (22). The resulting freshly amended soils hadfinal concentrations of approximately 200 and 400 mg/kg ofeach pesticide (0.6 and 1.2 mmol/kg soil for trifluralin, 0.93and 1.9 mmol/kg for atrazine, and 0.55 and 1.1 mmol/kg soilfor coumaphos) and were used directly in the desorptionexperiments.
Pesticide soil aging was promoted by using soil dryingand rewetting cycles as suggested by Shelton et al. (23). Soils
freshly amended with pesticide were wetted with deionizedwater (soil:water ) 5:1 w/w) and allowed to dry at roomtemperature for 1 week. The procedure was repeated 5 timesbefore using the aged soils in the desorption experiments.
Maximum Desorbable Concentration of Freshly Addedand Aged Pesticides in Soil. Soil amounts of 1 g containingfresh or aged pesticides were placed in screw cap tubestogether with 4 mL of acetonitrile (for trifluralin and atrazineextraction) or methanol (for coumaphos extraction) andtightly closed. One set of tubes was shaken at 25 °C for 1 hand another one for 5 d at the same temperature. In the thirdextraction experiment, a set of tubes was heated in a waterbath at 90 °C for 24 h. Each tube was centrifuged (3000xg)for 10 min, and the supernatant was removed and filtered(0.45 µm Millipore Millex HV) before analysis by highperformace liquid chromatography (HPLC). Extraction offreshly amended soils gave similar final concentrations ofthe pesticides with the three techniques. In contrast, thesoils with aged pesticides presented up to 25% lowerconcentrations of extracted pesticide after the 1 h shakingmethod when compared to the water bath method or the 5d shaking (results not shown). This difference in concentra-tions indicates that pesticide aging occurred after the wettingand drying cycles. The bath and the 5 d shaking method gavesimilar concentration values of extracted pesticide (70-55,40-7, and 25-20 mg/L for trifluralin, atrazine, and couma-phos, respectively) for both freshly and aged amended soils,and the average of both sets of results was used as themaximum amount of each pesticide potentially available fordesorption.
Rh-mix and TX-100 Sorption Isotherms. Aqueous solu-tions of surfactants were prepared at different concentrationsranging from 0 to 16.8 mM for Rh-mix and from 0 to 11.2 mMfor TX-100. The solutions were buffered at pH ) 8 withNaHCO3 (15 mM), and NaCl (10 mM) was added to keep aconstant ionic strength. Sodium azide (Sigma Chemical Co.,St. Louis, MO) was added to the solutions at 0.01% w/w toprevent microbial activity. A volume of 4 mL of each solutionand 1 g of clean soil were placed in screw cap tubes andtightly closed. The tubes were shaken for 5 d at 25 °C. Thesamples were centrifuged (3000xg) for 10 min, and the liquidsupernatants were filtered (0.45 µm Millipore Millex HV)before analysis by HPLC (Rh-mix) or colorimetric assay (TX-100). For all the experiments, the surfactant concentrationin soil (Cs
surf) at equilibrium was indirectly determined by
TABLE 1. Soil Properties and Langmuir Isotherm Parameters for Rh-mix, TX-100, and for the Individual Rhamnolipid Speciesc
soil
Hagerstown B (HTB) Hagerstown A (HTA) Beltsville A (BVA)
Hagerstown B (HTB) Hagerstown A (HTA) Beltsville A (BVA)
parameters Qmaxb (µmol/g) Klang
b (L/µmol) Qmaxb (µmol/g) Klang
b (L/µmol) Qmaxb (µmol/g) Klang
b (L/µmol)
Rh2C10C10 29.4 0.005 15.9 0.0019 12.5 0.0011RhC10C10 15.4 0.011 6.3 0.0044 5.5 0.0023Rh2C10C12 6.5 0.046 2.4 0.018 2.6 0.0055Rh2C10C12-H2 3.1 0.050 1.8 0.020 1.9 0.0076a USDA-BARC. b This work. c In all cases, Qmax and Klang were determined by duplicate experiments and varied by less than 5%.
calculating the difference between the surfactant concentra-tion in the aqueous phase (Cw
surf) at the beginning and at theend of each experiment.
Analysis of Surfactants. Derivatization of Rh-mix andHPLC analysis of derivatives were performed using themethod of Mata-Sandoval et al. (19). For the analysis of TX-100, a modification of the method for nonionic surfactantanalysis as cobalt thiocyanate active substances was em-ployed (24). Briefly, each sample was air-dried, followed bythe addition of 1.5 mL of cobalt thiocyanate reagent (preparedby the addition of 3 g of Co(NO3)2 (Aldrich Chemical Co.,Inc., Milwaukee, WI) and 20 g of NH4SCN (Aldrich ChemicalCo., Inc., Milwaukee, WI) in 100 mL of water and 3 mL ofCH2Cl2). The mixture was vortexed for 1 min and centrifuged(3000xg) for 10 min. The organic phase containing the cobaltpolyethoxylate complex was separated, and its absorbancewas measured at 620 nm in a Beckman DU-7 spectropho-tometer. Standards of TX-100 were detected on a linear rangeat concentrations as low as 0.04 mM.
Pesticide Desorption Experiments. Aqueous solutionscontaining Rh-mix or TX-100 were prepared at differentconcentrations that ranged from 0 to 25.6 mM (0, 0.8, 1.6,3.2, 4.8, 6.4, 8.8, 11.2, 14.4, 16.0, 19.2, 22.4, and 25.6 mM). Allsolutions were buffered and conditioned identically to thoseused in the surfactants sorption isotherm experiments. Avolume of 4 mL of each solution and 1 g of soil amended withfresh or aged pesticide were placed in screw cap tubes andtightly closed. The procedure was repeated for all combina-tions of soil, pesticide, and surfactant solution. Operationsof shaking, centrifugation, and filtration were identical tothose of the surfactant sorption isotherm experiments. Atequilibrium, the concentrations of pesticides in soil wereindirectly determined by the difference between the maxi-mum desorbable initial concentration and the final aqueousconcentration.
Analysis of Pesticides. All samples were analyzed by HPLC(Waters Model 712 WISP autosampler, two Waters Model510 pumps, and a Waters Model 996 Photodiode ArrayDetector) following the protocol described in our previousstudies (21, 22).
Results and DiscussionModeling the Sorption of Rh-mix and TX-100 by Soils. TheLangmuir model
where Qmax (µmol/g) is the maximum sorption of surfactantin soil, “Klang” equilibrium constant [Cw
surf] (mmol/L) is thesurfactant concentration in the aqueous phase, and [Cs
surf](mmol/g soil) is the concentration of surfactant in the soilphase, best described the sorption of both surfactants (Figure1). Similar sorption behavior has been reported by otherauthors with TX-100 (6, 8, 9). Experimental data and Langmuirisotherm model fit and parameters for Rh-mix and TX-100on the three soils are shown in Figure 1 and Table 1.Hagerstown B, the soil containing the highest clay contentand lowest foc, gave the highest values of Qmax and Klang forboth surfactants. Maximum sorption capacities of HTB claysoil for both surfactants were around twice the values ofthose for the silky soils. Rh-mix tendency to sorb onto soilsincreased as the soil clay content increased (HTB > HTA >BVA), and, coincidentally, the foc decreased (HTB < HTA <BVA) (Table 1). Our results indicate that surfactant sorptionis not a simple partitioning onto SOM. On a molar basis,Rh-mix sorption was higher than that of TX-100 onto HTB(15%) and HTA (26%) soils, while BVA sorbed 27% more TX-100 than biosurfactant. Noordman et al. (25) found no
correlation between soil organic carbon content and theadsorption of a multicomponent rhamnolipid surfactant.Cano et al. (26) observed that the adsorption of an alcoholethoxylate surfactant was better correlated to the percentclay content of a sediment rather than to the percent oforganic content. In light of our results and those of otherauthors (25, 26), a simplified approach of correlating theamount of sorbed surfactant and the soil percentage oforganic carbon is not applicable. Rather than a partitioningprocess, the sorption of both surfactants might be governedby the adsorption of molecules occurring at the soil-waterinterface. In the case of rhamnolipid species, their anioniccharacter might prevent partitioning into negatively chargedhumic matter. Instead, initial interactions between rham-nolipids and SOM or clay surface might be due to ion
[Cssurf] )
[Cwsurf]KlangQmax
1 + [Cwsurf]Klang
(1)
FIGURE 1. Experimental data (symbols ) and Langmuir isothermmodel fit (lines) for the sorption of rhamnolipid mixture by P.aeruginosa UG2 (A) and Triton X-100 (B), onto Beltsville A ((),Hagerstown A (9), and Hagerstown B (2) soils. Error bars representstandard deviation from duplicates.
TABLE 2. Partition Coefficients of Atrazine, Coumaphos, andTrifluralin Used To Model Their Interactions with Soil and theAqueous Phase in the Presence of Rh-mix or TX-100
a Solubility in solutions containing NaCl 10 mM, NaHCO3 15 mM,NaN3 100 mg/L, pH ) 8 (21). b Reference 28. c Reference 29. d This work.e Estimated using [Sw], Ksub, and Ksupra.
exchange reactions involving the carboxylate moiety of thesurfactant, surface complexation, or hydrogen bondinginteractions involving the rhamnose headgroups (25). ForTX-100, interactions between polar groups of SOM or claysurface and the polyethoxylate chain of the surfactant mightbe mostly responsible for its adsorption.
The values of Klang and Qmax for the individual rhamnolipidspecies of the mixture are presented in Table 1. The degreesof adsorptivity of each species are reflected in the Klang valueswhich followed the trend Rh2C10C10 < RhC10C10 < Rh2C10C12
≈ Rh2C10C12-H2. However, the mole fractions of the speciesfollowed the inverse trend, and that explains why Rh2C10C10
still yielded the highest Qmax value. Results indicate thatadsorptivity of rhamnolipid components is related to thedegree of their hydrophobicity. Rhamnolipid species withlonger fatty acid chain sizes and (or) a single rhamnose ringare more hydrophobic than those conformed by shorterlipidic chains and (or) containing double rhamnose ring inthe polar head. A correlation between the degree of pref-erential adsorption of nonionic and anionic surfactantcomponents to soil and to sediments, and their hydropho-bicity has also been observed by other authors (25-27).
Modeling the Interactions among Soils, HOCs, andSurfactants. To model the distribution of an HOC in soilsystems in the presence of surfactants, it is necessary to
consider that surfactant molecules alter the characteristicsof the soil and the aqueous phase, and as a result of this theinteractions of the analyte with both phases are greatlymodified.
Sorbed surfactant molecules can increase soil hydropho-bicity and effectively enhance the capacity of HOCs topartition onto soils. According to Edwards et al. (9), theeffective fractional organic carbon content (foc
*) of a solidafter surfactant sorption can be expressed as
where MWsurf and MWoc-surf are the molecular weights of thesurfactant and the weight of the carbon per molecule ofsurfactant, respectively; fsurf is the fraction of surfactant sorbedto the soil matrix (g sorbed surfactant/g soil), and foc is thefraction of organic carbon in the pure soil (g organic carbon/gsoil). The assumption that, on a carbon-normalized basis,sorbed surfactants can be as effective as humic matter assorbents of HOCs has already been used as a fair ap-proximation by several authors (8, 9).
At equilibrium concentrations of surfactant below theCMC, the distribution coefficient Ksol * (g soil/L) of thepesticide is expressed as
FIGURE 2. Effect of Rh-mix and TX-100 dosage on the partition of freshly added trifluralin (filled symbol) and aged trifluralin (open symbol)onto three different soils. Error bars represent standard deviation from duplicates. Shaded areas represent the model fit using the minimumand the maximum reported values for Koc.
where [Cwsurf-sub] (mmol/L) is the concentration of surfactant
monomers below the CMC, Kmon (L/mmol surf) is the partitioncoefficient of the analyte associated to the surfactantmonomers, Koc (L/g) is the organic carbon normalizedpartition coefficient, Ksub (mmol pest/mmol surf) is themaximum solubilization capacity of surfactant monomersfor the analyte below the CMC, and [Sw] (mmol/L) is thecontaminant maximum aqueous solubility (Table 2).
When the surfactant equilibrium concentration surpassesthe CMC in the aqueous phase, the distribution coefficient
Ksol* (g soil/L) is calculated by the equation
with
where [Cwsurf-supra] (mmol/L) is the total concentration of
surfactant when the concentration is above the CMC, Kmic
(L/mmol surf) is the micelle partition coefficient, Ksupra (mmolpest/mmol surf) is the maximum solubilization capacity ofthe micellar phase for the pesticide, and CMC (mmol/L) isthe critical micelle concentration (Table 2). Concentrationsat equilibrium of surfactant above the CMC [Cw
surf-supra] andbelow the CMC [Cw
surf-sub] and the fraction of surfactantincorporated to the soil fsurf were calculated with the Langmuireq 1.
FIGURE 3. Effect of Rh-mix and TX-100 dosage, on the partition of freshly added atrazine (filled symbol) and aged atrazine (open symbol)onto three different soils. Error bars represent standard deviation from duplicates. Shaded areas represent the model fit using the minimumand the maximum reported values for Koc.
Influence of Surfactant Dosage in the Desorption ofPesticides from Soils. Model results were generated usingthe parameters listed in Tables 1 and 2. All the parameterswere obtained experimentally, except the Koc range of valuesfor each pesticide which were taken from the literature (28,29). In general, the experimental Ksol
* values (calculated fromthe measured aqueous concentration of the pesticide andthe soil concentration determined by difference) for thepesticides followed the same trend and fit within the rangeof values estimated by the model after using the minimumand maximum Koc’s reported in the literature (range il-lustrated by the shaded areas in Figures 2-4). Best model fitcurves and Koc (fit) values are also presented for each soil-aqueous-micellar system in Figures 2-4. As predicted by themodel, the experimental data illustrates that the desorptionof none of the three pesticides was enhanced at equilibriumconcentrations of any surfactant below the CMC; moreover,as a result of the surfactants’ adsorption onto the soil, sorptionof some pesticides were even promoted at these conditions.This trend was more noticeable for atrazine in HTB soil(Figure 2) probably due to the pesticide high Sw value andthe high affinity of the soil to sorbed the surfactant.
At equilibrium concentrations of any surfactant above itsCMC, the pesticides tended to desorb from soils and partitioninto the micellar phase. This behavior was most evident fortrifluralin that has the highest Kmic of the three compounds,followed by coumaphos. Since HTA and BVA soils showedmuch lower capacity to sorb surfactants than HTB, themicellar phase started to form at lower surfactant dosages,
and both pesticides were desorbed from the silt soils afteraddition of either surfactant at amounts as low as 2400 µM(Figure 2). Dosages above 9000 µM of either surfactant werenecessary to observe any onset of trifluralin or coumaphosdesorption from HTB soil. Having very low Kmic values forboth surfactants, atrazine showed a poor desorption en-hancement from HTA and BVA soils at increasing surfactantdosages.
TX-100 was more efficient than the Rh-mix in desorbingthe three pesticides from the three soils. This tendency wasonce again more evident for trifluralin and coumaphos, sincethe Kmic values of both pesticides for TX-100 are almost doubleof those for the biosurfactant. The synthetic surfactant stillproved to be a more effective mobilizer of both pesticides,even in the case of BVA soil where the extent of TX-100sorption to soil is higher than that of Rh-mix.
The Kmic values for atrazine (Table 2) show that thepesticide affinity for Rh-mix micellar phase is much higherthan that for TX-100. Yet, the synthetic surfactant provedslightly more efficient than the biosurfactant in enhancingthe desorption of the pesticide in HTA and HTB soils. Thisis probably due to the higher tendency of rhamnolipidmolecules to sorb onto both soils as compared to TX-100.
Koc (fit) values for all the pesticides were higher in thesystems where Rh-mix was sorbed to the soils than in thosewhere TX-100 was present. Moreover, ratios of 3:1 (3.05 (0.2:1), 2:1 (2.03 ( 0.056:1), and 1.4:1 (1.41 ( 0.15:1) betweenthe Koc (fit) values of trifluralin, coumaphos, and atrazine inRh-mix versus TX-100 were kept constant in all the soils (with
FIGURE 4. Effect of Rh-mix and TX-100 dosage, on the partition of freshly added coumaphos (filled symbol) and aged coumaphos (opensymbol) onto three different soils. Error bars represent standard deviation from duplicates. Shaded areas represent the model fit usingthe minimum and the maximum reported values for Koc.
the exception of trifluralin in HTB clay). The results indicatethat the organic carbon of sorbed Rh-mix has a strongeraffinity for the pesticides than that of sorbed TX-100.
Approximately 66% of the carbon content in the biosur-factant mixture is located in the hydrophobic lipidic tailstructure of the molecules, and only 34% in the hydrophilicsugar and carboxylic group moieties. On the other hand,almost 60% of TX-100’s organic carbon is contained in thepolar poly-ethoxylate sector of the molecule, and 40% islocated in the octyl-phenol hydrophobic tail. Recent studiesof SOM nature in different soil and sediment populationsdemonstrate that the degree of polarity in SOM is a significantfactor in accounting for differences in Koc values for HOCs(30, 31). According to our model results, at surfactant soilsaturation conditions HTB’s organic carbon % increasesabout 10 times (from 0.35% to 2.8% and 3.0% respectivelyfor TX-100 and Rh-mix), HTA’s organic carbon % about 40to 50% (from 3.1% to 4.4% and 4.6%), and BVA’s organiccarbon % about 25 to 35% (from 4.1% to 5.2% and 5.5%). Dueto the important contribution of both surfactants to increaseSOM, and knowing that the hydrophobic organic carboncontent of Rh-mix is more than 1.5 times higher than thatin TX-100, soils coated with the biosurfactant are expectedto become better adsorbents of HOCs than those saturatedwith the synthetic one.
Our model assumed initially, based on the results by otherauthors (8, 9), that the organic carbon of sorbed surfactantmolecules becomes as effective sorbent as the organic carbonof humic matter. Yet, Koc (fit) values of trifluralin andcoumaphos in HTB soil exceeded the maximum valuesreported in the literature during Rh-mix addition. Knowingthat during Rh-mix sorption process in HTB the organiccarbon content increases almost 10 times, we believe thatKoc (fit) values for trifluralin and coumaphos surpass themaximum limits, because sorbed rhamnolipid moleculesbecome even better sorbents of HOCs than SOM. In thepresence of TX-100, Koc (fit) values for trifluralin andcoumaphos in HTB soil were very close to the highest limitsof Koc values reported in the literature for the pesticides. Thisindicates that TX-100 can be as good sorbent of bothpesticides as the natural organic carbon of soil. Similar resultsreported by Edwards et al. (9), show that TX-100 sorbed ontoLincoln fine sand not only enhanced phenanthrene sorptionbut also the sorbed surfactant was as effective sorbent of theanalyte as humic matter.
As shown in Figures 2-4, there were no significantdifferences between the values of Ksol
* for aged and freshpesticides in all soils at different surfactant dosages. Thisindicates that both surfactants can desorb pesticide moleculesthat have partitioned onto SOM or that are entrapped in soilmicropores due to the effect of the aging process, which wasevidenced after the extraction experiments.
In summary, we have evaluated the effect of Rh-mix andTX-100 on the distribution of trifluralin, coumaphos, andatrazine in soil-aqueous-micellar systems of three differentsoils. Our results indicate that the soil clay content contributesto the sorption of Rh-mix and that sorbed surfactants, atconcentrations below their CMCs, may induce an increasesoil hydrophobicity and an increase in the sorption ofhydrophobic pesticides. At concentrations above the CMC,a cosolvent type effect is observed, and pesticides can beeffectively desorbed and are not affected by soil wetting-drying cycles used for these experiments.
AcknowledgmentsThe authors thank the Department of Environmental Biologyat the University of Guelph, for supplying the P. aeruginosaUG2 cultures. Our acknowledgments to the U.N.A.M. (Uni-
versidad Nacional Autonoma de Mexico), CONACYT (ConsejoNacional de Ciencia y Tecnologıa), and The FulbrightComission for their economic support on the developmentof this work. The authors acknowledge the National ScienceFoundation Environmental Engineering Program (Grant BES-9702603) and the U.S. Department of Agriculture for fundingpart of this research. We also thank Sunita Prasad for helpin generating part of the experimental results.
Literature Cited(1) Myrick, C. Proceeding of international Workshop on Research
in Pesticide Treatment/Disposal/Waste Minimization; EPA/600/9-91/047; 1992.
(2) Winterlin, W.; Seiber, J. N.; Craigmill, A.; Baier, T.; Woodrow,J.; Walker, G. Arch. Environ. Contam. Toxicol. 1989, 18, 734-747.
(3) Scribner, S. L.; Benzig, T. R.; Sun, S.; Boyd, S. A. J. Environ. Qual.1992, 21, 115-120.
(4) Hatzinger, P. B.; Alexander, M. Environ. Sci. Technol. 1995, 29,537-545.
(5) Kommalapati, R. R.; Valsaraj, K. T.; Constant, W. D.; Roy, D.Water Res. 1997, 31, 2161-2170.
(6) Adeel, Z.; Luthy, R. Environ. Sci. Technol. 1995, 29, 1032-1042.(7) Laha, S.; Luthy, R. G. Biotechnol. Bioeng. 1992, 40, 1367-1380.(8) Chin, Y. P.; Kimble, K. D.; Swank, C. R. J. Contam. Hydrol. 1996,
22, 83-94.(9) Edwards, D. A.; Adeel, Z.; Luthy, R. G. Environ. Sci. Technol.
1994, 28, 1550-1560.(10) Scow, K. M.; Hutson, J. J. Am. Soil Sci. Soc. 1992, 56, 119-127.(11) Aronstein, B. N.; Alexander, M. Appl. Microbiol. Biotechnol. 1993,
39, 386-397.(12) Rouse, J. D.; Sabatini, D. A.; Suflita, J. M. Crit. Rev. Environ. Sci.
Technol. 1994, 24, 325-370.(13) Zhang, Y.; Miller, R. M. Appl. Environ. Microbiol. 1994, 60, 2101-
2016.(14) Zhang, Y.; Maier, W. J.; Miller, R. M. Environ. Sci. Technol. 1997,
31, 2211-2217.(15) Laha, S.; Luthy, R. G. Environ. Sci. Technol. 1991, 25, 1920-
1929.(16) Falatko, D. M.; Novak, J. T. Water Environ. Res. 1992, 64, 163-
169.(17) Deschenes, L.; Lafrance, P.; Villeneuve, J. P.; Samson, R. Appl.
Microbiol. Biotechnol. 1996, 46, 638-646.(18) Thangamani, S.; Shreve, G. S. Environ. Sci. Technol. 1994, 28,
1993-2000.(19) Mata-Sandoval, J. C.; Karns, J. S.; Torrents, A. J. Chromatogr. A
1999, 864, 211-220.(20) Mata-Sandoval, J. C.; Karns, J. S.; Torrents, A. Microbiol. Res.
2001, 155(4), 249-256.(21) Mata-Sandoval, J. C.; Karns, J. S.; Torrents, A. Environ. Sci.
Technol. 2000, 34(23), 4923-4930.(22) Mata-Sandoval, J. C.; Karns, J. S.; Torrents, A. J. Agric. Food
Chem. 2001, 49(7), 3296-3303.(23) Shelton, D. R.; Sadeghi, A. M.; Karns, J. S.; Hapeman, C. J. Weed
Sci. 1995, 43, 298-305.(24) Crabb, N. T.; Persinger, H. E. J. Am. Oil Chem. Soc. 1964, 41,
752-756.(25) Noordman, W. H.; Brusseau, M. L.; Janssen, D. B. Environ. Sci.
Technol. 2000, 34, 832-838.(26) Cano, M. L.; Dorn, P. B. Environ. Toxicol. Chem. 1996, 15, 684-
690.(27) Kibbey, T. C. G.; Hayes, K. F. Environ. Sci. Technol. 1997, 31,
1171-1177.(28) Montgomery, J. H. Agrochemicals Desk Reference; Lewis Pub-
lishers: Chelsea, MI, 1993; pp 31-33, 414-417.(29) Tomlin, C. The Pesticide Manual Incorporating The Agrochemi-
