Environ. Sci. Technol.

1997, 31, 670-675

Soil Washing Potential of a Natural Surfactant

D . R O Y , * , R . R . K O M M A L A P A T I , , S . S . M A N D A V A , ,| K . T . V A L S A R A J , A N D W . D . C O N S T A N T , Department of Civil & Environmental Engineering, Polytechnic University, Brooklyn, New York 11201, and Hazardous Waste Research Center, 3418 CEBA Building, Department of Civil & Environmental Engineering, and Department of Chemical Engineering, Louisiana State University, Baton Rouge, Louisiana 70803

Biodegradable natural surfactants obtained from plants can be an attractive alternative to synthetic surfactants in the remediation of contaminated soils. In this research, a plant-based surfactant obtained from the fruit pericarp of Sapindus mukurossi, a tree generally grown in tropical regions of Asia, is tested. A simple and economical method for the preparation of the surfactant is developed. An empirical formula for the surfactant was determined to be (C26H31O10)n. The aqueous solubilities of hexachlorobenzene (HCB) and naphthalene in the natural surfactant solutions were found to vary linearly with the concentration of the surfactant showing trends comparable to that of typical commercial surfactants. Natural surfactant solutions were also employed for flushing HCB from one-dimensional soil columns. HCB recoveries after 12 pore volumes of flushing with 0.5 and 1% natural surfactant solutions were 20 and 100 times more than that recovered by water flooding. These promising results warrant further research to establish the usefulness of plant-based surfactants for soil washing applications.

Introduction
Surfactants have been shown to increase the removal of contaminants from laboratory soil columns by solubilizing significant amounts of the contaminants (1-5). They can also be useful in mobilizing non-aqueous phase liquids from aquifers by decreasing the interfacial tension between the organic and water phases (6, 7). Use of surfactant solutions for soil washing has two advantages: (i) surfactant molecules tend to concentrate at the organic/aqueous interfaces and lower the interfacial tension considerably and (ii) surfactant molecules also form aggregates known as micelles at concentrations beyond the critical micelle concentration (cmc) that can solubilize hydrophobic organic compounds (HOCs). Mostly synthetic surfactants are currently being tested for the remediation of subsurface soils contaminated with complex wastes at hazardous waste sites (3, 8). In our recent search for biodegradable natural surfactants, we have identified a class of surfactants derived from plants
* Corresponding author fax: 718-260-3433; e-mail: droy@duke. poly.edu. Department of Civil & Environmental Engineering, Polytechnic University. Hazardous Waste Research Center, Louisiana State University. Present address: Rust Environment & Infrastructure, Greenville, SC 12345. | Department of Civil & Environmental Engineering, Louisiana State University. Department of Chemical Engineering, Louisiana State University.

belonging to the genus Sapindaceae that appear to have potential for the remediation of contaminated soils. These plants produce saponaceous substances called saponins, which form lather or foam in water (9, 10). Sapindus mukurossi, Sapindus trifoliatus, Sapindus laurifolius, and Sapindus emarginatus plants are widely grown in India and other tropical and subtropical regions of Asia, and their fruits are known by the common name Ritha or soapnut. About 56% of the fruit is pericarp, and the balance is the seed (11). These natural products have been traditionally used as a detergent for fabric washing and bathing and in folk medicine. The recorded external uses of this natural surfactant as a washing soap does not cite any toxic effects on human skin and eyes (12). Several researchers have isolated and identified the saponins from the fruit pericarps of the plants of genus Sapindus (13-17). Saponins are complex substances and are essentially glycosides with their aglycones related to either sterols or triterpenes. Saponin fraction was determined to be 10.1% of the weight of the pericarp and 6.1% based on the weight of the fruit (15). Extraction of the fruit pericarp with water has been the most commonly used method both for scientific and domestic uses such as washing hair and fabric. The aqueous solution thus obtained can be purified by isolating the saponin fraction. Research work on these plant products primarily focused on isolation and identification of the chemical constituents of the saponin fraction of the fruit extract and to a lesser extent on the application of these fruits as folk medicines. However, no scientific research exploring their application for soil remediation has been found in literature. Preliminary experiments conducted in our laboratory using natural surfactant solutions for the remediation of soils contaminated with naphthalene and hexachlorobenzene indicated that these solutions can desorb and solubilize significant amounts of these hydrophobic compounds. Also, these natural surfactant solutions were noted to enhance the growth of soil microorganisms under aerobic and anaerobic conditions by Kommalapati and Roy (18, 19). This paper focuses on evaluating the potential of this plant-based surfactant for soil flushing.

Experimental Section
Materials. Dry fruits of Sapindus mukorossi were procured from India. These fruits are golden brown in color and globular in shape with a diameter between 1 and 3 cm. After removing the seed, the outer pericarps were dried in an oven at 50 C for about 2 days. The pericarps were ground and sieved through U.S. Standard No. 20 sieve (840 m). Hexachlorobenzene (HCB), an aromatic chlorinated hydrocarbon, and naphthalene, a polycyclic aromatic hydrocarbon, were used as test compounds. These chemicals obtained from Aldrich Chemical Company (Milwaukee, WI) were 99% pure and used as supplied. An uncontaminated soil from a Superfund site near Baton Rouge, LA, was air-dried, homogenized, and kept in an oven overnight at 105 C for drying. Soil was ground, and the fraction passing U.S. Standard No. 10 (2 mm) sieve was used for the soil flushing experiments. This soil is classified as a sandy loam and has a very low organic matter content (e0.3%). Preparation of Natural Surfactant Solutions. Based on the available literature, water, methanol, ethanol, and a benzene:methanol (1:3) mixture were examined as solvents for preparing the natural surfactant solutions. The powder was added to solvents and stirred for 3 h at room temperature. The mixture was centrifuged at 10 000 rpm for 45 min, and the supernatant was filtered through a 0.44-mm prefilter followed by a metricel 0.45-m membrane filter (Gelman

670

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 31, NO. 3, 1997

S0013-936X(96)00181-2 CCC: $14.00

1997 American Chemical Society

Scientific, Ann Arbor, MI) in sequence. The filtrate was allowed to evaporate on a water bath at 70 C. The dry paste obtained was re-dissolved in water and used as stock solution. A 10% natural surfactant solution was prepared by extracting 10 g of fruit pericarp powder in 100 mL of deionized water in a similar way except that the last two steps, i.e., evaporating the filtrate and redissolving it in water, were eliminated. Surface Tension, Viscosity, and UV Absorbance Measurements. Surface tension measurements of aqueous solutions of the surfactant were made using a processor tensiometer K14 (Kruss GmbH, Hamburg, Germany). Vis cosity measurements were made using a Bohlin VOR rheometer system (Bohlin Instruments International AB, Lund, Sweden). The UV-visible absorbance spectrum of natural surfactant solutions was obtained using HP 8452A diode array spectrophotometer with HP 89531A-UV-VIS software (Hewlett-Packard Company, Wilmington, DE). Solubility Experiments. Aqueous solubility experiments were performed in Erlenmeyer flasks using several concentrations of surfactant in the presence of excess quantities of HCB or naphthalene crystals. The flasks were equilibrated for about 36 h, and the samples were centrifuged in Teflon tubes for 15 min at 15 000 rpm. The supernatant was extracted and analyzed for HCB or naphthalene. Soil Flushing Experiments. Glass columns (10 cm long and 5.75 cm in diameter) with a stainless steel top and bottom were used for all soil flushing experiments (7). To prevent soil from being washed out of the column and to distribute the flow uniformly, the outlet and inlet ends of the column were fitted with fine wire mesh sandwiched between two coarse wire meshes. Soil was contaminated by dissolving an appropriate quantity of HCB in petroleum ether and mixing it with a known weight of soil. This slurry was mixed thoroughly and the ether was allowed to evaporate slowly. The dry soil was transferred into a bottle and tumbled for about 1 week before the experiments. A soil packing procedure was adopted to achieve a bulk density (1.6 g/cm3) similar to that observed in the field (7). The column packed with contaminated soil was saturated with deionized water. Experiments were conducted in the downflow mode with water and natural surfactant solutions at a flow rate of 2.5 mL/min (pore water velocity ) 0.24 cm/min). The effluent samples were collected and analyzed for HCB. Analysis of Hexachlorobenzene. The traditional liquidliquid extraction for HCB analysis (EPA Method 612) did not work effectively in the presence of surfactant solutions. Commercially available SepPak C18 cartridges (Millipore Corporation, Milford, MA) were used for extracting HCB from aqueous solutions (10). The cartridges were activated by passing 5 mL of deionized (DI) water, 5 mL of methanol, and 5 mL of DI water again. The sample was eluted through the cartridge at a rate of 5 mL/min followed by a wash with 5 mL of DI water. The cartridge was then eluted with 5 mL of hexane, which was collected and analyzed for HCB on a HP 5890 Series II gas chromatograph (Hewlett-Packard Company, Wilmington, DE) fitted with a HP 7673 autosampler and Ni63 electron capture detector (ECD). The GC was equipped with a 30 m PTE-5 capillary column, with a 0.32 mm internal diameter and 1.0 m film thickness (Supelco Inc., Bellefonte, PA.). The chromatographic conditions were as follows: 1 L splitless injection; injection temperature 275 C; temperature program 50 C for 1 min to 270 C at 10 C/min, and hold for 3 min; and ECD temperature 325 C. Minimum detection limit for the method was 1 pg. Recovery of HCB using this method was found to be 93.7% with a standard deviation of 2.2% (10).

FIGURE 1. Variation of HCB solubility with soapnut concentration for extracts of several solvents. mixture at room temperature. Ethanol extracted the least (59.5%) and had a light yellowish golden color, whereas water extracted the most (71.2%) and had a very dark golden color. Methanol and methanol:benzene mixture extracted 68.6% and 63.2%, respectively, and acquired colors between those of water and ethanol extracts. Figure 1 shows the variation of HCB solubility with concentration of surfactant for the four solvents. The lines represent the regression data, and the points indicate the experimental data. The solubility of HCB increased with an increase in surfactant concentration for all extracts. HCB solubility at any soapnut concentration was highest for methanol extract, followed by methanol: benzene mixture, ethanol, and water extract. The differences between the different extracts were larger at higher concentrations (5% and 10%), and the variation decreased with a decrease in concentration. The variation between the HCB solubilities of different extracts is not significant at the 95% confidence level for concentrations up to 2.5% with Tukey multiple comparison method. Natural surfactant concentrations beyond 2.5% may also pose practical limitations (10). Hence, for simplicity in preparation, the use of water to extract fruit pericarp is recommended for field applications. (b) Mixing Time and Temperature. The optimum mixing time and temperature required to prepare the surfactant were determined experimentally by measuring the solubility of naphthalene in 10% water extracts for several mixing times (1-12 h) at room temperature. The highest naphthalene solubility was obtained for mixing times of 1 and 3 h. The solubility of naphthalene progressively decreased as the mixing time was increased beyond 3 h. In the subsequent work, a mixing time of 3 h was adopted. The variation in naphthalene solubility for several mixing temperatures in the range of 20-50 C was studied. The maximum solubility of naphthalene was observed at 28 C, and the solubility at room temperature was only about 8% less than the maximum solubility. Thus, it was concluded that extraction of soapnut into water at room temperature using a mixing time of 3 h with subsequent centrifugation and filtration with 0.45 m metricel filter was satisfactory. By using water as the solvent, two additional steps (evaporation of solvent and redissolution) were avoided. The only byproduct in the preparation of natural surfactant solutions is the un-extracted residue, which is about 30% of the total mass. The simple and safe method of preparation for this natural surfactant will definitely be an added advantage for its application in the real world. Several experiments were conducted to study the effect of refrigeration storage time on the performance of natural surfactant solutions. Refrigeration of extracts up to 6 days did not change significantly the solubilizing capacity of natural

Results and Discussion

Preparation of Natural Surfactant Solutions. (a) Solvent Selection. The fruit pericarp powder was extracted using either water, methanol, ethanol, or a methanol:benzene (3:1)

VOL. 31, NO. 3, 1997 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

671

FIGURE 2. Determination of cmc for natural surfactant solutions using surface tension measurements. surfactant solutions. However, surfactant solutions that were not refrigerated became cloudy within 2 days, indicating substantial microbial growth. Characterization of Natural Surfactant Solutions. (a) Properties. The natural surfactant was characterized by determining pH, chemical oxygen demand (COD), total organic carbon (TOC), total nitrogen, and total phosphorous. All the values were measured for a 10% surfactant solution. The COD and TOC of a 10% surfactant solution were 124 and 41 g/L, respectively. The nitrogen and phosphorous content of natural surfactant solutions were non-detectable using Standard Methods (20). The pH of the surfactant solution varied from 7.2 to 4.5 as the concentration increased from 0.005% to 1% and did not vary beyond 1%. The acidic nature of the solution could be due to the hydrolysis of the glycosides present in the fruit pericarps (16). The empirical formula of a compound can be determined either from the percentage composition of each element or from the oxidation of the compound. The latter method is used in this study due to its simplicity. Knowing the COD, TOC, nitrogen, phosphorous, and total weight of the organic compound, an empirical formula was calculated from the generalized oxidation half-reaction for an organic (10). The empirical formula for natural surfactant solutions was found to be (C26H31O10)n, where n is a constant and needs to be determined either from vapor density or by a cryoscopic method. The total carbon content of natural surfactant solutions expressed as percent of total weight of the compound was 60.6%. Row and Rukmini (16, 17) reported that the carbon content was 57.8% and 61.8% and a molecular formula of C52H84O212H2O and C47H76O17 for the saponin fraction of S. mukurossi and S. emarginatus, respectively. The empirical formula obtained in this study incorporates all the water-soluble constituents of the compound. Natural surfactant solutions are mixtures of complex organic compounds, and mass spectrometry was not helpful since the extract had to be further purified significantly. The empirical formula obtained in this study gives a fair idea of the ratio of carbon, hydrogen, and oxygen in natural surfactant solutions. (b) Critical Micelle Concentration. A very fundamental and important property of surfactants is micelle formation, which affects almost every measurable physical property that depends on size or number of particles in solution such as conductivity, surface tension, osmotic pressure, and viscosity. All these properties will show a break in the curve plotted against the concentration of surfactant in the region of the cmc (21). Figure 2 shows the variation of surface tension with the natural surfactant concentration in the range of

FIGURE 3. Solubility of (a) hexachlorobenzene and (b) naphthalene in natural surfactant solutions. 0.0005-10%. It is apparent from this figure that there is a sharp break in the surface tension curve in the concentration range 0.08-0.1%. As the concentration of natural surfactant increased from 0.0005 to 0.1%, surface tension dropped steadily, and beyond 0.1% the value remained stable at about 35 mN/m. This value is very similar to the range of 30-40 mN/m reported for several commercial surfactants (2, 3). The viscosity of natural surfactant solutions at a low concentration (0.01%) was significantly higher than that of water, and it reached a minimum when the concentration increased from 0.01% to 0.1%. At low surfactant concentration, the molecules are present in the monomeric form, and the hydrophilic moieties of surfactant are surrounded by ordered water molecules resulting in an increase in the viscous resistance. However, the formation of micelles releases the ordered water molecules and causes an abrupt change in viscosity. Viscosity measurements also indicate that the cmc of natural surfactant is 0.1%, thus confirming the cmc value obtained from surface tension measurements. When the natural surfactant concentration was increased beyond cmc (0.1%), the viscosity increased again due to the formation of more micelles and the resultant intermicellar interactions. Measurement of Natural Surfactants Using UV Absorbance. Natural surfactants, a mixture of saponins, glycosides, and other sugars, are fairly difficult to quantify accurately without chromatographic separation of individual compounds. However, several researchers used UV absorption as a surrogate measure for selected organic constituents in freshwater, wastewater, and salt water (22). Natural surfactant solutions of several concentrations were analyzed with a UV/visible spectrophotometer. The absorbance at any wavelength increased with an increase in surfactant concentration. The concentrations of natural surfactant and absorbance at wavelengths 252 and 292 nm were correlated, and the resulting equations are

672

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 31, NO. 3, 1997

A252 ) 0.05 + 2.31C A292 ) 0.002 + 1.675C

(R 2 ) 0.992) (R 2 ) 0.994)

(1) (2)

where A is absorbance and C is concentration of natural surfactant in weight percent. The strong correlation between UV absorbance and natural surfactant concentration (R2 > 0.99) suggests that these equations can be employed to determine the concentration of natural surfactant solutions. The correlations are valid in the range of 0.1-1.2% and 0.11.5% for wavelengths 252 and 292 nm, respectively. Solubility of Hexachlorobenzene and Naphthalene in Natural Surfactant Solutions. Figure 3a shows the variation in HCB solubility in natural surfactant solutions in the concentration range of 0.1-10%. Four measurements at each concentration are indicated by points, and the regression fit is represented by a line. Solubility of HCB in 10% natural surfactant solution increased several hundred times more than that in water. There appeared to be a linear relationship between solubility and natural surfactant concentration up to 10% surfactant concentration. The linearity between HCB solubility and surfactant concentration beyond cmc has been well established for commercial surfactants (2, 4, 21, 23). The slope of the regression line represents the maximum amount of HCB per mass of surfactant in equilibrium with solid phase HCB at standard temperature and pressure. The slope of the solubility curve can be used to calculate the solubilization ratio (SR), defined as

SR )

Co - Co* Cs - Cs*
FIGURE 4. Comparison of natural surfactant solutions with commercial surfactants: (a) hexachlorobenzene and (b) naphthalene. water to solubilize 1 mg of organic. The results of HCB solubility in different surfactants are presented in Figure 4. Brij 30, POE 10-LE, and Tween 85 seem to be very effective in solubilizing HCB as only a fraction of 1 g of surfactant was needed to solubilize 1 mg. About 2-3 g of Brij 35, Tween 20, Tween 80, and Exxal F 5715 was needed to solubilize a comparable quantity of HCB. Triton X-705 and Pluronic P-65 were less effective, requiring approximately 51 and 83 g, respectively, to solubilize 1 mg of HCB. In contrast, about 10.5 g of soapnut and about 5.5 g of SDS were required to solubilize 1 mg of HCB. A similar comparison of naphthalene solubility in different surfactants is presented in Figure 4b. To solubilize 1 mg of naphthalene in 1 L of water required about 215, 52, and 53 mg of soapnut powder, sodium dodecyl sulfate (SDS), and sodium dodecyl benzene sulfonate (SDBS), respectively. Only about 30 mg of non-ionic surfactant Tergitol and 15 mg of cationic surfactant hexadecyl trimethyl ammonium bromide (HTAB) are required to solubilize 1 mg of naphthalene. In all these comparisons, it should be noted that only 70% of the soapnut is dissolved in water, which makes the net weight of soapnut powder requirement for solubilizing 1 mg of HCB and naphthalene as 7.5 g and 150 mg, respectively. Application of Natural Surfactant Solutions to Soil Flushing. Results of soil flushing experiments are summarized in Figure 5a,b for 0.5% and 1% natural surfactant solutions, respectively. The columns were packed with soil contaminated to a level of approximately 2, 70, and 110 mg of HCB/kg of soil. It is clear from these graphs that the removal of HCB during the first pore volume after saturation of the column was practically negligible. HCB concentration in the effluent increased as more surfactant was pumped through the column. During the first pore volume, the saturation water was replaced with surfactant solution. The surfactant molecules pumped during the initial pore volumes are known

where Co and Co* are the concentrations of organic in a surfactant solution at any surfactant concentration and at critical micellar concentration (in g/g), respectively, and Cs and Cs* are the concentration of surfactant and the cmc of the surfactant (in g/g), respectively. One could also use molar concentrations and calculate the molar solubilization ratio (2). Since the surfactant concentration is expressed as percent (w/w), the solubility units are converted from micrograms per liter to grams per gram, and the solubilization ratio is calculated. For natural surfactant and HCB, the solubilization ratio is found to be in the range of 8.1 10-5 to 9.5 10-5. Figure 3b shows solubility of naphthalene in natural surfactant solutions. Naphthalene solubility also was directly proportional to the natural surfactant concentration in the range of 0.1-10%. This confirmed the general trend of linear proportionality of solubilization of a hydrophobic compound by a surfactant above its cmc. The solubilization ratio for natural surfactant and naphthalene was determined to be 4.8 10-3 g/g. It is believed that a partition-like interaction of the hydrophobic solute with the microscopic organic environment of dissolved organic molecules is responsible for the increased solubility of the solute. Natural surfactant, due to its constituents, may behave very similarly to natural organic matter and thus improve the solubility of solutes. This increase in solubility can be in addition to what was provided by saponins, the fraction responsible for the surfactant properties of the fruit extract. Comparison of Natural and Commercial Surfactant Solutions. Solubility of HOCs in natural surfactant solutions exhibited a similar pattern as typically observed for commercial surfactants showing a linear increasing trend of solubility with an surfactant concentration. Roy et al. (24) reported the solubility of naphthalene in four commercial surfactants, and Jafvert et al. (23) used 10 different surfactants to solubilize HCB in aqueous solutions. The solubility parameter (mmol of organic/mol of surfactant) reported in these studies is used for comparison. The comparison is done on the basis of grams of surfactant required in 1 L of

VOL. 31, NO. 3, 1997 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

673

FIGURE 5. Recovery of HCB from one-dimensional soil columns flushed with natural surfactant solutions: (a) 0.5% and (b) 1% natural surfactant. to undergo adsorption onto soil, and the effective surfactant concentration monitored by UV adsorption in the effluent was considerably lower than that in the influent. The adsorption of surfactant continued for about 3-4 pore volumes before the effluent surfactant concentration reached a maximum. HCB concentrations increased steadily after the first pore volume and approached the maximum value by the fifth pore volume. This maximum concentration was about 80% of the HCB solubility in the aqueous natural surfactant solution of the corresponding concentration. HCB concentration in the effluent remained fairly constant for the remainder of the experiment. The removal appeared to be limited by mass transfer from the adsorbed phase to the aqueous phase at low contamination levels (2 mg of HCB/kg of soil), since HCB concentration in the effluent was only a small fraction of its solubility. Natural surfactant solution of 1% concentration recovered a total of 110 and 1000 g of HCB in 12 pore volumes from soil contaminated with 2 and 80 mg of HCB/kg of soil, respectively. As the contamination level increased, HCB recoveries increased, but the recoveries were limited by the HCB solubility in the flushing media at higher contamination levels. Total HCB recovered from highly contaminated soil using natural surfactant solutions for 12 pore volumes was about 200 g for 0.5% and about 1000 g for 1.0% natural surfactant. These amounts are about 20 and 100 times more than that recovered with water (10 g) from soil columns contaminated with 90 mg of HCB/kg of soil. The total HCB removed from the soil column as a percent of initial HCB in the column was about 0.02, 0.4, and 4% for water flood, 0.5%, and 1% natural surfactant, respectively. It should be noted that experiments

FIGURE 6. Effect of natural surfactant concentration on the recovery of HCB from one-dimensional soil columns (a) 2 mg/kg and (b) 90 mg of HCB/kg of soil.

FIGURE 7. Pressure buildup across the soil column during the flushing experiments. were terminated after 12 pore volumes, and higher removal could be achieved by continuing the flushing operations. This significantly higher removal of HCB with natural surfactant solutions is very encouraging because these solutions can possibly be used as a supplement at existing pump-and-treat facilities and to enhance the performance of the process. HCB remaining in the soil column was determined by slicing the column into four approximately equal sections and analyzing each section independently in duplicate. The

674

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 31, NO. 3, 1997

removal of HCB was mainly from the influent end, and the HCB present at the exit did not change appreciably. HCB removal from the soil in one-dimensional soil flushing experiments starts at the influent end and progressively moves toward the effluent end. Figure 6a,b depicts the effect of surfactant concentration on the removal of HCB for natural surfactant solutions at two levels of soil contamination (2 and 90 mg of HCB/kg of soil). It is clear that with an increase in the concentration of natural surfactant there is a significant increase in the removal of HCB from the soil column. The higher removals with higher surfactant concentration can be explained by the increased solubility of HCB in the surfactant micelles. With the increase in surfactant concentration the number of micelles will increase, resulting in an increased solubilization of HCB. The solubilized HCB is easily mobilized and washed from the soil matrix. This suggests that micellar solubilization is the primary mechanism responsible for the mobilization and the subsequent washing of HCB from soil. Similar trends were observed for all levels of contamination. When the natural surfactant concentration was increased from 0.5% to 2.5%, the removal increased significantly. After 12 pore volumes using 0.5% solution, about 200 g of HCB was recovered as compared to 1000 g using a 1% solution and 3300 g in 11 pore volumes using a 2.5% solution. In contrast, the water flood was able to recover only 10 g in 12 pore volumes. Pressure Buildup across the Soil Columns. Pressure buildup is one of the controlling factors that determines the application of surfactants in field applications (25). Figure 7 shows the pressure across the soil columns using different flushing solutions. The variation was about the same for all the runs except for the run with 2.5% natural surfactant. The general trend was that pressure increased with an increase in surfactant concentration. However, the increase was not very significant when the natural surfactant concentration was increased from 0.5 to 1%. The pressure increased gradually and approached 60 psi by the end of 8 pore volumes in case of a 2.5% natural surfactant. When the natural surfactant concentration was increased to 5%, the pressure increased to 60 psi in 3-4 pore volumes (not shown in the plot), and the run had to be terminated. It is believed that the interactions of natural surfactant solutions with the soil are responsible for the increase in pressure. Similar observations were reported by Liu (5), noting that an increase in surfactant concentration changed the hydraulic properties and pore geometry of the soil matrix due to increased interactions of surfactant molecules with soil minerals. However, the specific interactions that are responsible for the increase in pressure buildup are not well established at this time. From our studies, it appears that natural surfactant at a concentration of 1% is a good compromise to achieve higher removal and avoid high pressure buildup across the soil column.

Literature Cited
(1) Abdul, S. A.; Gibson, T. L.; Rai, D. N. Ground Water 1990, 28 (6), 920-926. (2) Edwards, D. A.; Luthy, R. G.; Liu, Z. Environ. Sci. Technol. 1991, 25 (1), 127-133. (3) Liu, Z.; Edwards, D. A.; Luthy, R. G. Water Res. 1992, 26 (10), 1337-1345. (4) Abdul, S. A.; Gibson, T. L.; Ang, C. C.; Smith, J. C.; Sobczynki, R. E. Ground Water 1992, 30 (2), 219-231. (5) Liu, M. Mobilization of a Hydrophobic Organic Compound Using Surfactant for Soil Washing. Ph.D. Dissertation, Louisiana State University, Baton Rouge, LA, 1993. (6) Fountain, J. C.; Klimek, A.; Beikirch, M. G.; Middleton, T. M. J. Hazard. Mater. 1991, 28, 295-311. (7) Kommalapati, R. R. Soil Flushing of Non Aqueous Phase Liquids Using Conventional Surfactant Solutions and Colloidal Gas Aphron Suspensions. M.S. Thesis, Louisiana State University, Baton Rouge, LA, 1994. (8) Gannon, O. K.; Bibring, P.; Raney, K.; Ward, J. A.; Wilson, D. J.; Underwood, J. L.; Debelak, K. A. Sep. Sci. Technol. 1989, 24 (14), 1073-1094. (9) Mandava, S. Application of a Natural Surfactant from Sapindus emarginatus to In-Situ Flushing of Soils Contaminated with Hydrophobic Organic Compounds. M.S. Thesis, Louisiana State University, Baton Rouge, LA, 1994. (10) Kommalapati, R. R. Remediation of Contaminated Soils Using Plant Based Surfactant. Ph.D. Dissertation, Louisiana State University, Baton Rouge, LA, 1995. (11) Bor, N. L. Manual of Indian Forest Botany; Oxford University: Calcutta, India, 1953. (12) Windholz, M., Ed. The Merck Index: An Encyclopedia of Chemicals, Drugs, and Biologicals; Merck & Co. Inc.: Rahway, NJ, 1983. (13) Kasai, R.; Nishi, M.; Mizzutani, K.; Miyahara, I.; Moriya, T.; Miyahara, K.; Tanaka, O. Phytochemistry 1988, 27 (7), 22092211. (14) Uppal, I. S.; Mehta, R. L. J. Sci. Ind. Res. 1951, 10B, 190-195. (15) Gedeon, J. J. Sci. Ind. Res. 1954, 13B, 427-428. (16) Row, L. R.; Rukmini, C. Indian J. Chem. 1966, 4, 36-38. (17) Row, L. R.; Rukmini, C. Indian J. Chem. 1966, 4, 149-150. (18) Kommalapati, R. R.; Roy, D. J. Environ. Sci. Health, Part A 1996, 31 (8). (19) Kommalapati, R. R.; Roy, D. J. Environ. Sci. Health, Part A 1996, 31 (10). (20) Greenberg, A. E.; Clesceri, L. S.; Eaton, A. D. Standard Methods for the Examination of Water and Wastewater, 18th ed.; American Water Works Association: Washington, DC, 1992. (21) Rosen, M. J. Surfactants and Interfacial Phenomena; John Wiley & Sons Ltd.: New York, 1978. (22) Eaton, A. J. Am. Water Works Assoc. 1995, 87 (2), 86-90. (23) Jafvert, C. T.; Van Hoof, P. L.; Heath, J. K. Water Res. 1994, 28 (5), 1009-1017 (24) Roy, D.; Kongara, V.; Valsaraj, K. T. J. Hazard. Mater. 1995, 42, 247-263. (25) Nash, J. H. Field Studies of In-Situ Soil Washing; EPA/600/287/110; United States Environmental Protection Agency: Cincinnati, OH, 1987.

Received for review February 27, 1996. Revised manuscript received October 15, 1996. Accepted October 23, 1996.X ES960181Y

Acknowledgments
This research was funded by the United States District Court, Middle District of Louisiana, through the Louisiana State University Hazardous Waste Research Center.

Abstract published in Advance ACS Abstracts, January 15, 1997.

VOL. 31, NO. 3, 1997 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

675