enhanced animal waste management through ... - PubAg - USDA

ENHANCED ANIMAL WASTE MANAGEMENT THROUGH APPLICATION OF SURFACTANTS TO SOIL MATERIAL: LABORATORY FEASIBILITY TESTING B. Allred, G. O. Brown, L. A. Brandvold ABSTRACT. Laboratory testing was conducted to determine the feasibility of using surfactants to enhance soil performance with regard to animal waste management at feedlot and dairy sites. Three surfactants, one anionic (sodium dodecyl benzene sulfonate) and two cationic (polyoxypropylene methyl diethyl ammonium chloride and tetradecyl trimethyl ammonium bromide), were tested on a sandy loam. The best surfactants are those capable of substantially reducing the amount of leaching under saturated and unsaturated flow conditions. Maintaining acceptable mechanical compaction characteristics, and if possible, limiting nitrate mobility are also important. The experimental program therefore focused on three areas: (1) Surfactant influence on leaching as indicated by effects on soil hydraulic properties (saturated hydraulic conductivity, soil moisture diffusivity, and unsaturated soil wetting front penetration); (2) Surfactant effects on mechanical compaction; and (3) Nitrate transport in surfactant–modified soil. Saturated hydraulic conductivity was measured using falling–head permeability tests. Transient unsaturated horizontal column experiments provided information concerning surfactant influence on soil moisture diffusivity and wetting front penetration. Standard Proctor methods were applied in the study of surfactant effects on mechanical compaction. Adsorption batch tests gauged nitrate mobility in surfactant–treated soil The anionic surfactant (sodium dodecyl benzene sulfonate) and one of the two cationic surfactants (polyoxypropylene methyl diethyl ammonium chloride) produced considerable reductions in saturated hydraulic conductivity, soil moisture diffusivity, and unsaturated soil wetting front penetration. With regard to mechanical compaction, the anionic surfactant and the other cationic surfactant (tetradecyl trimethyl ammonium bromide) affected decreases in the optimum moisture content corresponding to maximum dry bulk density. The moisture content range between optimum and the wet value corresponding to 90% or 95% maximum dry bulk density is also much narrower for these two surfactants, thereby increasing the difficulty of field compaction. Interestingly, the cationic surfactant having the greatest influence on soil hydraulic properties was not the same one having the most substantial effect on mechanical compaction. Adsorption batch tests indicate nitrate mobility in the sandy loam is not significantly altered by soil treatment with any of the three surfactants. The polyoxypropylene methyl diethyl ammonium chloride test results (no adverse effects on mechanical compaction and reduced hydraulic conductivity, diffusivity, and wetting front penetration) are alone enough to suggest the feasibility of using surfactants to enhance soil performance for animal waste management. However, more investigation is needed with a greater variety of surfactants on a number of different soils, particularly with regard to hydraulic property testing using leachate typically generated at animal feedlot and dairy facilities. Keywords. Surfactants, Hydraulic conductivity, Soil moisture diffusivity, Wetting front penetration, Mechanical compaction, Nitrate mobility.

F

eedlot and dairy facility animal waste is often disposed in storage ponds or treatment lagoons. Compacted soil liners, which are an important component of animal waste storage ponds and

Article was submitted for review in March 2000; approved for publication by the Soil & Water Division of ASAE in November 2000. Presented at the 1998 ASAE Annual Meeting as Paper No. 98–4109. Use of trade names is for informational purposes only and does not imply endorsement by the United States Department of Agriculture, Oklahoma State University, or the New Mexico Bureau of Mines and Mineral Resources. The authors are Barry Allred, ASAE Member Engineer, Agricultural Engineer, USDA–ARS, Soil Drainage Research Unit, Columbus, Ohio; Glenn O. Brown, ASAE Member Engineer, Associate Professor, Biosystems and Agricultural Engineering Dept., Oklahoma State University, Stillwater, Oklahoma; and Lynn A. Brandvold, Senior Chemist, New Mexico Bureau of Mines and Mineral Resources, Socorro, New Mexico. Corresponding author: Barry Allred, USDA–ARS, Soil Drainage Research Unit, 590 Woody Hayes Drive, Columbus, OH 43210; phone: 614–292–9806; fax: 614–292–9448; e–mail: [email protected].

Vol. 44(3): 513–524

treatment lagoons (fig. 1), are responsible for preventing the release of contaminants into the subsurface environment. If leakage through the soil liner becomes excessive, deterioration of groundwater quality beneath the waste containment or treatment impoundment will occur. This can be averted by choosing soil liner material that, after compaction, has sufficiently low saturated hydraulic conductivity. Given a minimum thickness of 30 cm, the recommended hydraulic conductivity for a compacted soil liner used in storage ponds and treatment lagoons is 1.25 Ü 10–6 cm/s (USDA–SCS, 1993). Locally available soils often do not meet this saturated hydraulic conductivity criteria. When this is the case, significant transportation costs are incurred bringing in material from off–site. Surfactant additives may provide the solution by reducing the saturated hydraulic conductivity of local soil material to acceptable levels. Several studies (Watson et al., 1969; Allred and Brown, 1994 and 1995; Renshaw et al., 1997) have shown surfactants

Transactions of the ASAE 2001 American Society of Agricultural Engineers ISSN 0001–2351

513

to significantly reduce saturated hydraulic conductivity in soil. In falling–head permeability tests of a loamy soil (52% sand, 31% silt, and 17% clay) having an initial hydraulic conductivity of 6 Ü 10–5 cm/s, Allred and Brown (1994) found that the injection of 0.1 mole/kg anionic, cationic, and amphoteric surfactant solutions reduced hydraulic conductivity by two orders of magnitude or greater after less than 0.5 applied pore volumes. Besides indicating a potential problem regarding the use of surfactants for contaminant removal by in situ soil flushing, these results, in a positive sense, imply that surfactants can be employed beneficially as additives to reduce the hydraulic conductivity of material used in storage pond or treatment lagoon liners. The soil used as liner material is mechanically compacted to enhance its engineering properties; therefore, the effect of surfactants on this compaction process likewise needs to be considered. Figure 1 shows that leaching, and in turn nitrate (NO3–) release, can also occur beneath runoff interception trenches and confined animal feeding areas (CAFAs). Phillips et al. (1997) found shallow wells adjacent to corrals and CAFAs to exhibit increased groundwater NO3– levels during monitoring over a 40–year period. Leachate movement directly beneath a CAFA most likely occurs under unsaturated flow conditions. The runoff interception trench is occasionally filled with water but only after large rainfall events and for short periods of time. As such, unsaturated flow conditions probably prevail beneath these locations as well. Consequently, reduced infiltration through spray application of surfactants to the ground surface could be a way to limit NO3– release from the CAFA and interception trench components of feedlot and dairy facilities. Recent investigations suggest a second way that soil performance can be enhanced through surfactant application. Studies by Brownawell et al. (1990), Burris and Antworth (1992), and Wagner et al. (1994) have shown soils treated with cationic surfactants to immobilize organic contaminants. Bowman et al. (1995) found that cationic surfactant–modified zeolite was capable of sorbing oyxanions, such as chromate, selenate, and sulfate, from aqueous solution. The contaminant of greatest concern with regard to animal waste storage ponds and treatment lagoons is the oxyanion, NO3–. Surfactant additives could therefore serve two very important functions: reduction of leakage and contaminant immobilization. Additionally, several anionic surfactants along with many cationic surfactants exhibit resistance to environmental biodegradation, particularly when present in high concentrations (Huddleston and Allred, 1967; Swisher, 1987; Karsa and Porter, 1995). Greater stability due to biodegradation resistance increases the likelihood that the surfactant will maintain environmental effectiveness over a long period of time. The overall objective of this laboratory investigation was to determine the feasibility of using surfactants to enhance the animal waste management performance of soil material present at feedlot and dairy operations. This objective was accomplished through testing one anionic and two cationic surfactants on a sandy loam soil. Two surfactant types were excluded from the study: nonionic surfactants because previous research (Allred and Brown, 1994 and 1995) indicates they have less effect on reducing hydraulic conductivity, and amphoteric surfactants because they tend to be expensive, and therefore, economically impractical. The experimental program focused on three areas: 514

1. Surfactant influence on hydraulic properties (saturated hydraulic conductivity, soil moisture diffusivity, and unsaturated soil wetting front penetration). 2. Surfactant effects on mechanical compaction. 3. Nitrate mobility in surfactant–modified soil. Surfactant effectiveness in these three areas will largely determine their practicality for use as performance–enhancing amendments. The best surfactants will be those that substantially reduce saturated hydraulic conductivity, soil moisture diffusivity, and unsaturated soil wetting front penetration, that limit nitrate mobility, and that do not adversely affect mechanical compaction.

MATERIALS SURFACTANTS Surfactants are organic compounds that on the molecular level are comprised of both hydrophobic (tail) and hydrophilic (head) components (fig. 2a). The hydrophobic part is typically a linear or branched hydrocarbon chain, and the hydrophilic portion is either charged or highly polar. Because of this amphipathic structure, surfactant molecules in aqueous solution tend to concentrate at phase boundaries, thereby altering interfacial properties such as surface tension. At aqueous solution concentrations below what is referred to as the critical micelle concentration (CMC), surfactants exist as individual molecules or “monomers.” Upon reaching the CMC, surfactant molecules begin to form aggregates called micelles, and the monomer concentration becomes constant. Accordingly, surface tension decreases with increasing aqueous solution surfactant concentration until the CMC is reached, and it then tends to level off beyond this point. Surfactants are classified according to the charge of the hydrophilic group: anionic, cationic, amphoteric (positive and/or negative charge), or nonionic. One anionic and two cationic surfactants were tested in this study. Table 1 provides a list of these surfactants including some of their characteristics. The abbreviations A1, C1, and C2 given in table 1 will be used to designate the three surfactants throughout the remainder of this article. Common commercial availability was one criterion by which these three were chosen. In addition, the anionic surfactant with a branched alkyl chain and both quaternary ammonium cationic surfactants all have molecular structures resistant to biodegradation, particularly within the high concentration levels tested in this study (Huddleston and Allred, 1967; Swisher, 1987; Karsa and Porter, 1995).

Figure 1. Schematic of a feedlot or dairy facility showing locations where nitrate (NO3–) can be released into the subsurface.

TRANSACTIONS OF THE ASAE

If the amount of surfactant present within the system is large enough, hydrophobic interaction between tail groups will occur in conjunction with electrostatic adsorption, often resulting in partial or complete coverage of soil particle surfaces with surfactant bilayers (figs. 2d and 2e) (Rosen, 1989; West and Harwell, 1992; Tadros, 1995). This is particularly common with cationic surfactants, but it may take place with anionic surfactants as well. For the situation of a charged mineral surface in contact with a surfactant solution substantially above the CMC, direct micelle adsorption is a dominant process (Chen et al., 1992). Once the micelle comes into contact with the surface, it collapses and becomes rearranged into a loosely packed bilayer. These bilayers may exist as isolated patches covering only a part of the surface (McDermott et al., 1994; Rutland and Parker, 1994), and when this is the case, the adsorbed surfactant concentration can be significantly less than the cation exchange capacity (Sullivan et al., 1997). Surfactant monolayer coverage from electrostatic attractions alone (figs. 2b and 2c), or in combination with hydrophobic interactions, produces a soil particle surface that has been changed from hydrophilic to hydrophobic. With bilayer formation, the covered portion of the surface is again hydrophilic, but in the case of cationic surfactants, its charge is reversed from negative to positive. Surface charge reversal through cationic surfactant application is one way to increase electrostatic adsorption of anionic contaminants, such as nitrate, thereby limiting their mobility in the soil environment. Figure 2. (a) Schematic of surfactant molecule, (b) electrostatic cation exchange adsorption of cationic surfactant molecules onto a soil particle surface, (c) electrostatic coadsorption of anionic surfactant molecules onto a soil particle surface, (d and e) surfactant bilayer formation due to hydrophobic interactions between tail groups.

There are several mechanisms by which surfactants are partitioned onto soil surfaces in the presence of an aqueous phase. One surfactant loss mechanism is chemical precipitation. Regarding this process, surfactants combine with resident soil anions or cations to form particle–coating precipitates. Under typical pH conditions, soil surfaces normally have a net negative charge. Through cation exchange mechanisms, positively charged cationic surfactants are electrostatically attracted to these surfaces (fig. 2b). Electrostatic adsorption also occurs with anionic surfactants as a result of coadsorption (Gaudin and Chang, 1952). The process of coadsorption involves resident multivalent cations, such as calcium (Ca2+) and magnesium (Mg2+), which bridge surfactant anions to soil surfaces having a negative charge (fig. 2c).

SOIL Surfactants were tested on the Slaughterville sandy loam (Thermic Udic Haplustoll), which was obtained from a location near Perkins, Oklahoma. In its unaltered state, the Slaughterville sandy loam does not meet the saturated hydraulic conductivity criteria of 1.25 Ü 10–6 cm/s needed for liner material used in animal waste storage ponds or treatment lagoons, and for this reason was specifically chosen for testing surfactant enhancement capability. The soil is comprised of 70% sand, 25% silt, and 5% clay. The pH is 7.5, organic matter is only 0.16%, and the cation exchange capacity (CEC) equals 5.5 meq/100g. With the soil organic matter concentration so low, surface adsorption due to direct hydrophobic interaction between surfactants and the resident organic material was not a dominant process. Exchangeable base concentrations for calcium (Ca2+), magnesium (Mg2+), potassium (K+), and sodium (Na+) were 5.84, 0.22, 0.10, and 0.03 meq/100g, respectively. These values indicate that Ca2+ dominates soil cation exchange sites. The relatively high Ca+2 concentration indicates that

Table 1. Surfactant information.

Chemical Name

Surfactant Abbre– Type viation

Chemical Formula

Molecular Weight (g/mole)

CMC (mole/L) and CMC Surface Tension (dynes/cm) [a]

C12H25C6H4SO3Na

348

5.4 × 10–3 and 35.7

Sodium dodecyl benzene sulfonate[b]

Anionic

A1

Tetradecyl trimethyl ammonium bromide[b]

Cationic

C1

C14H29N(CH3)3Br

336

2.4 × 10–3 and 37.8

C2

[(C2H5)2CH3N(C3H6O)6.3H]Cl

600

1.2 × 10–3 and 40.0

Polyoxypropylene methyl diethyl ammonium chloride[b] [a]

CMC values were determined from surface tension measurements of aqueous surfactant solutions at 22°C using a Fisher Scientific Model 21 Tensiomat tensiometer. At this temperature, the surface tension of water is 72.4 dynes/cm. [b] Surfactants A1 (Cat. No. 28,995–7) and C1 (Cat. No. 86,042–5) were obtained from Aldrich Chemical Co., Milwaukee, Wisc.; surfactant C2 (trade name EMCOL CC–9) was provided by CKWitco, Greenwich, Conn.

Vol. 44(3): 513–524

515

coadsorption (fig. 2c) may be an important process regarding anionic surfactants applied to this soil. Particle size distribution was determined at the New Mexico Bureau of Mines and Mineral Resources Soil Testing Lab in Socorro, N.M. Soil chemical characterization was done by the Soil, Water, and Air Testing Laboratory at New Mexico State University in Las Cruces, N.M. Both labs used standard procedures as outlined in Methods of Soil Analysis, Parts 1 and 2 (ASA and SSSA, 1982 and 1986).

Hydraulic flow, mechanical compaction, and nitrate adsorption batch experiments were conducted on surfactant– modified, Slaughterville sandy loam soil. For comparison purposes, testing was also done on unaltered soil. Surfactant treatment concentrations were typically 1.0% or 0.5%, and in one instance, 0.25% of dry soil weight. Soil preparation was accomplished by adding an appropriate amount of 10% (by weight) surfactant solution to a 2 kg batch of dry Slaughterville sandy loam. The 10% surfactant solution was added in 25 g increments using a spray bottle capable of applying a fine mist to the soil surface. The soil was then thoroughly mixed after each 25 g increment. Once the complete amount of surfactant had been added, the soil was thoroughly saturated with distilled water, stirred, and allowed to air dry over a period of 48 h. The final step involved using a mortar and pestle to break up the dry soil aggregates, so that all material could pass through a 2 mm sieve, followed again by further mixing. FALLING–HEAD PERMEABILITY EXPERIMENTS Saturated hydraulic conductivity is by far the most important hydraulic property for gauging the feasibility of using a particular soil material in a storage pond or treatment lagoon compacted liner (fig. 1). The saturated hydraulic conductivity (K) is itself a function of both fluid and porous media properties and can be expressed: (1)

where k = is the intrinsic permeability = is the fluid kinematic viscosity g = gravitational acceleration constant. Intrinsic permeability is governed by the soil pore structure (total porosity, pore size distribution, and tortuosity). By changing either k or u, surfactant treatment can potentially alter the value of K. Falling–head permeability tests (Wray, 1986) were conducted to determine surfactant effects on saturated hydraulic conductivity. All the soil columns used in these experiments had a length of 7.2 cm and a diameter of 4.1 cm. The columns were packed with air–dried soil in 25 g lifts, and dry bulk density averaged 1.62 g/cm3. After packing, vacuum saturation of the columns with distilled water was followed by a 24 h equilibration period prior to test initiation. Two experiments on unaltered soil were conducted, and results compared to those of 0.5% and 1.0% surfactant–modified soils. The value of the hydraulic gradient along the column

516

TRANSIENT UNSATURATED HORIZONTAL COLUMN TESTS Unsaturated flow conditions are likely to exist beneath some feedlot and dairy facility components, such as runoff interception trenches and confined animal feeding areas. Vertical unsaturated flow occurring within soil can be mathematically expressed:

∂q ∂  ∂q  =  D(q ) − K (q ) ∂t ∂z  ∂z 

EXPERIMENTAL AND DATA ANALYSIS METHODS

kg K=u

at the beginning of each falling–head permeability test was 12.6.

(2)

where t = time z = distance in the vertical direction = volumetric moisture content (% by volume) K(q) = unsaturated hydraulic conductivity D(q) = diffusivity. Unsaturated hydraulic conductivity and diffusivity are both functions of moisture content. The relationship between diffusivity and unsaturated hydraulic conductivity is given by:

D(q ) = K (q )

d Ψ k (q )g d Ψ = dq u dq

(3)

where the intrinsic permeability, k(q), in unsaturated soil is also dependent on moisture content, Y is the soil water pressure potential, and all other terms have been previously defined. The soil water pressure potential results from the combined effects of thin film water adsorption (hA ) and capillarity (hC ) or stated mathematically (Hillel, 1980):

Ψ → f (hA , hC )

(4)

In sandy/silty soils and moderate moisture content conditions, capillarity dominates (Hillel, 1980) and can be expressed:

hC =

2g WA cosa g(rW − r A )r

(5)

where W = water/air interfacial tension = soil/water contact angle ρW – ρA = density difference between water and air within the soil environment (≈ ρW) r = pore radius g = gravitational acceleration constant. The cosine of the contact angle is further defined:

g −g cosa = SA SW g WA

(6)

where S = the soil/air interfacial tension SW = the soil/water interfacial tension WA water/air interfacial tension, as previously given. Consequently, equations 3, 4, 5, and 6 indicate that D(q) can be altered through surfactant–influenced changes in several variables, including k(q), u, gWA , S, and gSW.


As suggested by equation 2, decreased soil leaching rates can potentially result from surfactant–affected reductions in diffusivity. The functional relationship between diffusivity and moisture content is determined from transient unsaturated horizontal soil column experiments using a computer–controlled syringe pump apparatus (fig. 3), described by Brown and Allred (1992), to maintain hydraulic boundary conditions defined by Bruce and Klute (1956). These boundary conditions include a column inlet moisture content maintained at a constant value during testing, and for column positions significantly beyond the advancing wetting front edge, a moisture content equal to that uniformly present in the soil before test initiation. Diffusivity values are calculated from soil column moisture profile test data by: *

1  dl  q D(q * )= −   ∫ ldq 2  dq q * q i

(7)

where = x/ t (This is known as the Boltzmann transform, in which x is the horizontal distance from the column inlet, and t is the time duration of the test.) qi = is the initial moisture content

q * = is the moisture content at distance x. The value of q *

is between that of i and inlet moisture content ( ). Testing procedures involved using the computer–controlled syringe pump to inject distilled water at the inlet of a horizontally mounted soil column. The column was comprised of individual 1 or 2 cm acrylic rings that had been taped together. Columns had a diameter of 3.5 cm and lengths ranging from 12.0 to 28.0 cm. Within the column, air–dried unaltered or surfactant–modified soil was packed in 1 cm lifts. Packed dry bulk density averaged 1.60 g/cm3. At the completion of a test, the column was broken down, the soil was extracted from within each ring, and all samples were then oven dried at 105³C for 24 h to determine moisture content. Next, the measured moisture contents along the length of the column were plotted with respect to the Boltzmann transform (). Four experiments, having similar inlet and initial moisture content boundary conditions but different time durations and injection volumes (6 h – 10 mL, 12 h – 14.14 mL, 24 h – 20 mL, and 48 h – 28.28 mL) were conducted for each of the unaltered or surfactant–modified soils. To provide a better

basis for comparison, two of these 4 experiment sets were conducted with the unaltered soil. Similarity of moisture profiles plotted versus the Boltzmann transform for tests having similar boundary conditions but different time durations is an indication that no time–dependent chemical or physical effects are influencing unsaturated flow. With this the case, equation 7 is valid for calculating the relationship between diffusivity and moisture content. A1, C1, and C2 treatment concentrations for the unsaturated tests were 0.25%, 1.0%, and 0.5%, respectively. Concentrations of 0.25% and 0.5% were used for A1 and C2 because higher concentrations produced invalid test results due to leakage at the column inlet. Wetting front penetration shown in the column moisture profiles along with the calculated relationship between diffusivity and moisture content allowed for evaluation of surfactant effects on unsaturated flow. MECHANICAL COMPACTION EXPERIMENTS Mechanical compaction is the most important engineering feature of soil material used in storage pond and treatment lagoon liners. Soil can be reworked through various compaction procedures to produce a denser, and therefore better, liner material. The benefits of compaction are increased shear strength, less settlement, and decreased hydraulic conductivity. Surfactants that limit soil compactability are of little practical use. Surfactant–affected changes in mechanical compaction were established through comparison of standard Proctor test results between unaltered and surfactant–modified (0.5% and 1.0%) soil. The standard Proctor compaction test (Wray, 1986) involves dropping a 2.5 kg weight, from a set 30.5 cm height, 25 times onto an amount of soil placed within a metal cylinder. The metal cylinder has a volume of 944 cm3 and is filled in three lifts using this procedure. Based on cylinder volume, soil weight, and moisture content, a dry bulk density value for the soil can be calculated. For a particular soil treatment, these tests are conducted at different moisture content values (% by weight) followed by construction of a dry bulk density versus moisture content curve. The plotted curve (fig. 4) has a peak corresponding to the optimum moisture content (MCOpt) needed to produce the maximum dry bulk density (DMax). During construction of the liner, keeping the moisture content of soil material at the exact optimum value is impractical, if not impossible. The

Figure 3. Syringe pump apparatus for testing unsaturated flow.

Vol. 44(3): 513–524

517

best that can be hoped for is to maintain the moisture content within a range at which 90% or 95% maximum dry bulk density (D90% or 95%) can be achieved. In addition, the best combination of high shear strength and low hydraulic conductivity is found at moisture content values just wet of optimum (Hausmann, 1990). Hydraulic conductivity is less on the wet side of optimum because a dispersed soil microstructure having parallel particle alignment inhibits flow perpendicular to the particle orientation, while on the dry side of optimum, a more permeable flocculated microstructure is present. Therefore, while the liner is being placed, the moisture content of the soil material should be kept within the range shown in figure 4: between MCOpt and MCWet – 90% or 95%. The compaction curve constructed from standard or modified Proctor testing provides valuable information concerning this crucial moisture content range. The most attractive surfactant treatments are those resulting in large dry bulk density values at the peak, along with an extended moisture content range between MCOpt and MCWet – 90% or 95%. NITRATE ADSORPTION BATCH TESTS One of the greatest environmental concerns at feedlot and dairy facilities is degradation of groundwater quality resulting from nitrate (NO3–) release into the subsurface. Due to lack of solubility constraints and its negative charge, NO3– tends to be one of the most mobile solutes found in soil. Surface charge reversal from negative to positive by cationic surfactant treatment is one way to increase electrostatic adsorption of NO3– anions, in turn limiting their mobility within soil. However, for surface charge reversal to occur, the treatment process needs to at least partially cover soil particles with a bilayer of cationic surfactant molecules.

For this portion of the study, 1% surfactant–modified soils were tested. Due to the typical cost of surfactant products (1 to 4 $/kg), it was felt that surfactant application at concentrations beyond this level would be economically unrealistic. Assuming total adsorption, a 1% soil surfactant concentration corresponds to 54% and 30% of the Slaughterville sandy loam cation exchange capacity for C1 and C2, respectively. This would suggest that not enough of either cationic surfactant was provided to produce complete monolayer coverage, let alone complete bilayer coverage of the surface. However, as stated previously, exposure to surfactant solutions substantially above the CMC often results in the formation of bilayer patches on soil mineral surfaces, and the total adsorbed concentration can be substantially less than the soil cation exchange capacity (Chen et al., 1992; McDermott et al., 1994; Rutland and Parker, 1994; Sullivan et al., 1997). Therefore, since the 10% surfactant solutions used to treat the soil were over two orders of magnitude above the CMC, it was hoped that if C1 and C2 bilayer patches did form, they would prove capable of reducing nitrate mobility. Surfactant effects on NO3– mobility in soil was gauged with batch experiments. Testing involved adding 50 mL of 250 mg/L NO3– (343 mg/L NaNO3) solution to a 125 mL Erlenmeyer flask containing 50 g unaltered or 1% surfactant–modified soil, mixing for 1 h using a mechanical wrist–action shaker, followed by standing periods of 20, 67, or 237 h (~1, ~3, or ~10 d), and finally, analysis of solution nitrate concentration using a Dionex 4000i ion chromatograph with AG14 and AS14 columns and UV/VIS spectrophotometric detection. Surfactant batch test NO3– solution concentrations over time (20, 67, and 237 h) were compared to those of the unaltered soil batch tests. For a particular standing period, a surfactant batch test NO3– solution concentration significantly lower than that observed for the unaltered soil batch test indicates surfactant ability to limit soil nitrate mobility through adsorption. Although the anionic surfactant was included in these experiments, the focus of this part of the study was on the cationic surfactants. Nitrate concentrations during the same time period were also monitored in a refrigerated sample of the original 250 mg/L stock solution. By observing concentration changes over time in both the stock solution and unaltered soil batch tests, the effect of chemical and/or microbiological degradation processes on NO3– levels can be assessed. Nitrate decrease in the refrigerated stock solution sample could have resulted from chemical degradation, while measured concentration reductions in the unaltered soil batch tests could be due to chemical and/or microbiological processes. If the stock solution NO3– level remains constant, but there is an observed concentration decrease in the unaltered soil batch tests, then these reductions result solely from microbiological processes.

EXPERIMENTAL RESULTS AND DISCUSSION

Figure 4. Relationship of dry bulk density versus moisture content determined from standard or modified Proctor testing.

518

FALLING–HEAD PERMEABILITY EXPERIMENTS Saturated hydraulic conductivity (K) values measured by falling–head permeability tests are provided in table 2. Compared with the two values obtained for unaltered soil, the C1 treatment concentrations of 0.5% and 1.0% caused only minor reductions in hydraulic conductivity. Rakhshandeh


Table 2. Falling–head permeability test results. Saturated Hydraulic Conductivity (cm/s) Treatment 1.52 × 10–3 1.17 × 10–3 7.44 × 10–7 2.21 × 10–6 8.19 × 10–4 1.13 × 10–3 3.39 × 10–6 2.12 × 10–7

Unaltered no. 1 Unaltered no. 2 0.5% A1 1.0% A1 0.5% C1 1.0% C1 0.5% C2 1.0% C2

roo et al. (1998) actually observed increases in K for soils treated with a surfactant that structurally was almost the same as C1. However, K decreases were substantial as a result of A1 and C2 soil treatment at 0.5% and 1.0% concentrations. The 1.0% C2 treatment affected the largest K decrease, approaching almost four orders of magnitude. In fact, the 0.5% A1, 1.0% A1, 0.5% C2, and 1.0% C2 treatments all reduced K to values near or below 1.25 Ü 10–6 cm/s, which is the level recommended for soil liners by the USDA Soil Conservation Service (1993). Based on saturated hydraulic conductivity decrease alone, both the A1 and C2 surfactants appear capable of dramatically improving the performance of soil used in storage pond and treatment lagoon liners. Equation 1 indicates that reduced K results from a decrease in porous media intrinsic permeability (k) and/or an increase in solution kinematic viscosity (u). Increased u within the solution moving through the soil column could result from desorption of surfactant from the soil particle surface into the aqueous phase. As shown in table 3, surfactant presence will indeed increase u, but not to the extent necessary to explain the large K decreases obtained with the A1 or C2 surfactants. For example, even at a concentration of 1 Ü 10–1 mole/kg (~6% by weight), which is extremely high and unlikely for the solution moving through the soil column, increased , as compared with distilled water (1.21 mm2/s to 0.96 mm2/s), accounts for a K reduction of only 26% and not the K decrease of three or four orders of magnitude actually measured for the C2 surfactant. Consequently, A1– and C2–affected saturated hydraulic conductivity reductions are due almost totally to a decrease in the porous media intrinsic permeability. There are several possible mechanisms by which surfactants can influence intrinsic permeability (k). Allred and Brown (1994 and 1995) found surfactants capable of mobilizing soil colloids that can clog pore constrictions. Surfactant salt precipitates could likewise plug pore constrictions, as could the formation of surfactant lyotropic liquid crystals. Mustafa and Letey (1969) observed surfactant–affected decreases in soil Table 3. Surfactant solution kinematic viscosity[a] (mm2/s). Solution Concentration (mole/kg) Surfactant

1 × 10–2

2.5 × 10–2

5 × 10–2

1 × 10–1

A1 0.99 – 1.11 – C1 – 0.99 – – C2 1.04 – 1.10 1.21 [a] Surfactant solution viscosities were measured at a temperature of 22°C with a Cannon Instrument Co. Size 50 Viscometer. At this temperature, distilled water had a kinematic viscosity of 0.96 mm2/s.

Vol. 44(3): 513–524

aggregate stability. Surfactants may even facilitate pore size reduction through increased clay mineral swelling. At present, it is unclear which, if any, of these surfactant–related mechanisms is dominant in reduction of k, and in turn, K. TRANSIENT UNSATURATED HORIZONTAL COLUMN TESTS Unsaturated test moisture content profiles plotted versus the Boltzmann transform () are depicted in figure 5. As stated previously, sets of four experiments having similar moisture content boundary conditions but different time durations and injection volumes (6 h – 10 mL, 12 h – 14.14 mL, 24 h – 20 mL, and 48 h – 28.28 mL) were conducted for each of the unaltered or surfactant–modified soils. Two of these 4 experiment sets were conducted with the unaltered soil, thereby allowing better comparison against soils containing surfactant. For the 20 unsaturated tests conducted in this part of the study, 97% of the water injected was accounted for by soil sample weight differences before and after oven drying. Separate symbols are used in figure 5 to denote data from tests of different time duration. In all cases, good similarity is depicted between the 6, 12, 24, and 48 h tests, indicating the appropriateness of using equation 7 to calculate diffusivity values. More importantly, this similarity indicates that no time–varying chemical or physical processes affected unsaturated flow. Stated differently, similarity implies that chemical and physical equilibrium conditions in the soil environment were quickly achieved regarding processes that influence unsaturated flow. Composite moisture curves constructed from the raw test data are also provided in figure 5 for each of the unaltered or surfactant–treated soils. Equation 7 was applied to each these composite curves in order to obtain relationships for diffusivity versus moisture content. For comparison purposes, figure 6 combines the composite moisture profiles for each of the unaltered and surfactant–modified soils into a single graph, which dramatically illustrates the greater wetting penetration in unaltered or 1% C1 soil than in 0.25% A1 or 0.5% C2 soils. This is shown quantitatively by the computed values at the centroid of the composite moisture profiles, which are 0.021, 0.022, 0.015, and 0.015 cm/s0.5, respectively, for unaltered, C1, C2, and A1 soils. The relationships of diffusivity and moisture content are exhibited in figure 7. Both the 0.25% A1 and 0.5% C2 treatments, in comparison to unaltered soil, reduced diffusivity at moisture contents above 0.1. On the other hand, within a moisture content range from 0.025 to 0.17, the C1 surfactant treatment increased diffusivity. Diffusivity values in the unaltered soil continuously increase with greater moisture content (). For the 1.0% C1–treated soil, as becomes greater, diffusivity first increases and then decreases, having a peak value of 0.003 cm2/s at = 0.16. When q reaches 0.2, diffusivity values for 0.25% A1– and 0.5% C2–treated soils tend to level off between 0.0006 and 0.0009 cm2/s. Wetting penetration and diffusivity results depicted in figures 6 and 7 have some interesting implications. The A1 and C2 surfactant treatments reduce unsaturated water movement and therefore potentially enhance soil perfomance by minimizing leachate release into the subsurface.

519

Figure 5. Volumetric moisture content profiles plotted versus the Boltzmann transform (l): (a) unaltered soil, (b) 0.25% A1–modified soil, (c) 1.0% C1–modified soil, (d) 0.5% C2–modified soil.

Figure 6. Composite moisture content profiles for unaltered and surfactant–modified soil.

520

Figure 7. Diffusivity versus moisture content relationships for unaltered and surfactant–modified soils.


Conversely, treatment with the C1 surfactant increases unsaturated flow and thus would be inadvisable as an amendment to soil material used in runoff interception trenches or confined animal feeding areas. Depending on the moisture content and which surfactant was used for soil treatment, diffusivity was increased, decreased, or unaffected. Equation 3 implies that diffusivity, D(q), can be influenced by changes in unsaturated hydraulic conductivity, K(q), and/or the soil water pressure potential – moisture content slope (dY/dq). In turn, K(q) is governed by intrinsic permeability, k(q). Possible surfactant–related processes that lessen intrinsic permeability were discussed in the previous section on falling–head permeability test results and include chemical precipitation, dispersal and mobilization of pore–clogging colloids, lyotropic liquid crystal formation, reduced aggregate stability, and clay mineral swelling. Given the fairly moderate moisture contents and sandy/silty nature of the soil, capillary processes certainly dominated within the soil column portion between the inlet and wetting front edge in which data were obtained for making diffusivity calculations. Consequently, as suggested by equation 5, surfactant–affected changes in the water/air interfacial tension (WA) and/or the soil/water contact angle () could alter and in turn dY/dq. Desorption of surfactant molecules from particle surfaces into the aqueous phase will reduce gWA of water moving through an unsaturated soil treated with A1, C1, or C2. Hydrophobic surfaces have values that are significantly greater than surfaces that are hydrophilic. Natural soils, especially those with low organic matter content such as the one tested in this study, typically have particle surfaces that are hydrophilic. For these soils, particle surfaces become hydrophobic if covered by a surfactant monolayer (figs. 2b and 2c) and return to hydrophilic if covered by a surfactant bilayer (figs. 2d and 2e). The value of is itself a function of the soil/air interfacial tension (SA), the soil/water interfacial tension (SW), and the water/air interfacial tension (WA) (eq. 6). Surfactant

adsorption onto a soil particle surface affects SA, and especially SW, while dissolved surfactants, as stated before, always reduce WA. The same processes by which A1 and C2 reduced saturated hydraulic conductivity (K) probably affected K(q) in the same way, thereby being the most reasonable explanation for the decrease in D(q) caused by these two surfactants at moisture contents above 0.1 (fig. 7). It is uncertain why A1 and C2 had almost no effect on diffusivity at q values below 0.1. Since C1 had little effect on K, it seems unlikely that changes in K(q) caused the increased diffusivity observed with this surfactant between q values of 0.025 to 0.17 (fig. 7). Higher dY/dq must therefore be responsible and is attributed to the balance between surfactant–induced reduction in gWA and changes in cos (eq. 5). For concentrations above the CMC and at the room–temperature conditions (~22³C) at which the unsaturated tests were conducted, C1 is capable of decreasing WA from 72.4 to 37.8 dyne/cm (table 1). The D(q) peak shown by C1 at q = 0.16 is presently unexplained. MECHANICAL COMPACTION EXPERIMENTS Data from the standard proctor tests are provided in figure 8. Figure 8a compares the density – moisture curve of the unaltered soil to those of 0.5% surfactant–modified soils. Figure 8b compares unaltered soil and 1.0% surfactant–modified soils. Figures 8a and 8b show that surfactant presence has little effect on the maximum dry bulk density obtained during mechanical compaction. In fact, maximum dry bulk density fell within narrow bounds of 1.77 to 1.82 g/cm3 for all soils tested. However, the optimum moisture content corresponding to maximum dry bulk density was affected. The optimum moisture contents (% weight) for unaltered, 0.5% C2, and 1.0% C2 soils were 11.5, 11.4, and 11.4, respectively. For 0.5% A1, 1.0% A1, 0.5% C1, and 1.0% C1 soils, the respective optimum moisture contents were 9.1, 9.1, 8.4, and 9.0.

Figure 8. Mechanical compaction test results: (a) unaltered soil compared to 0.5% surfactant–modified soils, (b) unaltered soil compared to 1.0% surfactant–modified soils.

Vol. 44(3): 513–524

521

Table 4. Compaction test results. 95% Compaction Soil Treatment

DMax (g/cm3)

D95% (g/cm3)

MCOpt– MCWet – 95%

Unaltered 0.5% A1 1.0% A1 0.5% C1 1.0% C1 0.5% C2 1.0% C2

1.77 1.81 1.79 1.80 1.82 1.81 1.81

1.68 1.72 1.70 1.71 1.72 1.71 1.72

11.5 – 15.4 9.1 – 9.4 9.1 – 9.9 8.4 – 9.5 9.0 – 10.2 11.4 – 15.0 11.4 – 15.1

Along with optimum moisture content reduction, figure 8 shows that A1 and C1 also produced a steep density drop–off once optimum moisture content was exceeded. In a practical sense, this steep drop–off presents a problem regarding field compaction of soil liner material. With a typical target of 90% or 95% maximum dry bulk density for increased shear strength, the moisture contents corresponding to lowest saturated hydraulic conductivity are found in a range between MCOpt and MCWet – 90% or 95% (fig. 4). For example, tests on a silty clay show that compaction at a wet side moisture content 2% greater than optimum decreased hydraulic conductivity by a factor of 100 compared to a dry side moisture content 2% less than optimum (Hausmann, 1990). Compaction procedures for achieving shear strength and hydraulic conductivity goals are greatly simplified when there is a wide moisture content range between MCOpt and MCWet – 90% or 95%. Conversely, where the density drop–off beyond optimum moisture content is steep, maintaining soil liner material within narrowly spaced MCOpt and MCWet – 90% or 95% limits is difficult and often impractical under field conditions. Compaction results are further quantified in table 4. MCDiff–90% was calculated by subtracting MCOpt from MCWet–90%, while MCDiff–95% reflects the difference between MCOpt and MCWet–95%. Compared to unaltered soil, MCDiff–90% and MCDiff–95% values are substantially smaller as a result of A1 or C1 treatment. This is not the case for C2, which had MCDiff–90% and MCDiff–95% values similar to the unaltered soil. For A1 and C1, the drastically reduced range between MCOpt and MCWet – 90% or 95% suggests that application of these surfactants could make soil liner compaction difficult and should not be recommended. This does not appear to be a problem for C2. NITRATE ADSORPTION BATCH TESTS Results of nitrate adsorption batch tests are provided in table 5. Nitrate concentrations in the stock solution and unaltered soil batch tests remained at original levels (255–257 ppm) up to 67 h after test initiation. At 237 h, the stock solution NO3– concentration had dropped slightly to 242 ppm, probably due to some form of chemical degradation. A larger concentration drop to 228 ppm occurred for the unaltered soil batch test measured at 237 h, suggesting minor involvement from microbial degradation processes. After 20 h, nitrate concentrations for the surfactant–modified soil batch tests had not changed appreciably compared to that of the unaltered soil batch test, indicating no adsorption had taken place up to this time. After 67 and 237 h, only the 1.0% C2–treated soil consistently

522

90% Compaction

MCDiff–95%

D90% (g/cm3)

MCOpt– MCWet – 90%

MCDiff–90%

3.9 0.3 0.8 1.1 1.2 3.6 3.7

1.59 1.63 1.61 1.62 1.63 1.62 1.63

11.5 – 18.0 9.1 – 9.6 9.1 – 10.6 8.4 – 10.7 9.0 – 11.7 11.4 – 17.2 11.4 – 16.5

6.5 0.5 1.5 2.3 2.7 5.8 5.1

Table 5. Nitrate adsorption batch test results. Nitrate Concentration (ppm) Experiment Type

20 h

67 h

237 h

Stock solution Unaltered soil 1.0% A1 soil 1.0% C1 soil 1.0% C2 soil

256 255 252 253 258

256 257 229 241 222

242 228 222 229 207

showed nitrate concentrations even modestly below those corresponding to the unaltered soil. These results are not surprising for A1 because negatively charged nitrate ions are not electrostatically attracted to anionic surfactants. The lack of nitrate adsorption on C1– or C2–treated soil may indicate that cationic surfactant bilayer patches did not form, and therefore, localized reversal of surface charge on particles from negative to positive was not achieved. Without surface charge reversal, electrostatic adsorption of nitrate could not occur. Even if patches of cationic surfactant bilayers did form, they obviously were not capable of reducing nitrate transport. Consequently, from the standpoint of limiting nitrate mobility through adsorption processes alone, surfactant application at the 1% level was not effective with regard to the sandy loam soil tested.

CONCLUSIONS AND RECOMMENDATIONS Overall experimental results support the feasibility of using surfactants to enhance soil performance with regard to animal waste management at feedlot and dairy facilities. Of the three surfactants tested, C2 (polyoxypropylene methyl diethyl ammonium chloride) appears to be the best candidate for this purpose. C2 affected substantial reductions in saturated hydraulic conductivity, soil moisture diffusivity, and unsaturated soil wetting front penetration. In fact, at a treatment concentration of 1% dry soil weight, C2 decreased saturated hydraulic conductivity from 1.35 Ü 10–3 to 2.12 Ü 10–7 cm/s, which is below the recommended level of 1.25 Ü 10–6 cm/s for compacted soil liner material in animal waste storage ponds and treatment lagoons (USDA–SCS, 1993). Although quite impressive, better experimental results could have been achieved with this surfactant. To simplify procedures by not having to maintain soil in a moist state, columns for falling–head permeability experiments were packed with dry material at an average bulk density of 1.62 g/cm 3. Had the C2 soil been packed at the same dry bulk density, but wet of the optimum moisture content, lower saturated hydraulic conductivity values would have been obtained. Packed dry, the 0.5% C2 treatment had a hydraulic TRANSACTIONS OF THE ASAE

conductivity of 3.39 Ü 10–6 cm/s, which was barely above 1.25 Ü 10–6 cm/s liner recommendation. A value below this limit would undoubtedly have been achieved if the 0.5% C2 soil had been packed at the same density but wet of optimum. By packing soil at greater dry bulk density than 1.62 g/cm3, saturated hydraulic conductivity could have been lowered even further. As an added advantage with regard to engineering preparation of soil liner material, C2 showed no adverse effects on mechanical compaction. The optimum moisture content and maximum dry bulk density were relatively unaffected by C2 treatment. Likewise, the range of values between MCOpt and MCWet – 90% or 95% were similar for both unaltered and C2–treated soil. With regard to hydraulic properties, the A2 surfactant (sodium dodecyl benzene sulfonate) also exhibited a clear–cut ability to reduce saturated hydraulic conductivity, soil moisture diffusivity, and unsaturated soil wetting front penetration. However, standard Proctor testing showed that A2 affected a steep density drop–off once optimum moisture content was exceeded. This could potentially make mechanical compaction of A2–treated soil difficult, thereby reducing its usefulness for storage pond or treatment lagoon liners. As with C2, the A1 surfactant could potentially be spray applied on the ground surface to limit the amount of infiltration beneath runoff interception trenches or confined animal feeding areas. However, if this were to be attempted, care would have to be taken in order to insure that increased soil erosion does not take place. This is particularly a concern for A1, since it has the capability to mobilize and disperse clay colloids (Allred and Brown, 1995). C2 may be a better choice, since it does not appear to exhibit these clay colloid mobilization and dispersion characteristics (Allred and Brown, 1994). If surfactant–induced soil erosion is a problem, a layer of soil along the perimeter of the trench could be scraped off and surfactant solution then sprayed onto the new bare surface. The layer of soil initially removed would then re–applied and compacted. Designing the interception trench in this manner will minimize both erosion and infiltration. The C1 surfactant (tetradecyl trimethyl ammonium bromide) reduced saturated hydraulic conductivity only a minor amount but actually increased soil moisture diffusivity and unsaturated soil wetting front penetration. Therefore, it is not recommended for use as an amendment to soil for animal waste management purposes, although it may have environmental application in permeable barriers used to remove contaminants from groundwater. Cationic surfactant–treated soils are well documented for their ability to adsorb organic contaminants (Brownawell et al., 1990; Burris and Antworth, 1992; Wagner et al., 1994). Since it does not influence hydraulic properties, C1 may have an advantage over other cationic surfactants because it would not significantly reduce flow rate through the permeable groundwater barrier. None of the three surfactants substantially increased nitrate adsorption onto soil particle surfaces. However, the A1– and C2–affected reductions in hydraulic conductivity, soil moisture diffusivity, and unsaturated soil wetting front penetration would alone decrease movement of leachate through soil material, thereby limiting nitrate release beneath

Vol. 44(3): 513–524

feedlot and dairy facilities. For example, given saturated flow conditions within a storage pond or treatment lagoon soil liner, surfactants that are capable of reducing hydraulic conductivity by three orders of magnitude, such as A1 and C2, would be just as effective in terms of protecting the subsurface environment as ones that produce 99.9% nitrate adsorption. Modification of hydraulic properties is by far the most effectual long–term strategy with regard to minimizing nitrate release, since the mechanism of adsorption ceases to work once the anion exchange capacity of a soil is reached. Although results from this study indicate the feasibility of surfactants to enhance animal waste management soil performance, more work needs to be done. A wide variety of surfactants should be tested on a number of different soils. From the standpoint of compacted liners, an important focus of any future program should be long–term saturated hydraulic conductivity testing using leachate typical of animal waste storage pond and treatment lagoon facilities. Likewise, surfactant adsorption isotherm experiments need to be conducted. These tests will gauge the ability of surfactants to remain in place over time within soil material. Finally, a better understanding of surfactant effects on unsaturated flow could be obtained from measurement of

versus curves for unaltered and surfactant–treated soil.

REFERENCES Allred, B., and G. O. Brown. 1994. Surfactant induced reductions in soil hydraulic conductivity. Ground Water Monitoring and Remediation 14(2): 174–184. _____. 1995. Surfactant–induced reductions of saturated hydraulic conductivity and unsaturated diffusivity. In ACS Symposium Series 594: 216–230. D. A. Sabatini, R. C. Knox, and J. H. Harwell, eds. Washington, D.C.: American Chemical Society. ASA and SSSA. 1982. Methods of Soil Analysis: Part 2. Chemical and Microbiological Properties. 2nd ed. Madison, Wisconsin: American Society of Agronomy and Soil Science Society of America. _____. 1986. Methods of Soil Analysis: Part 1. Physical and Mineralogical Methods. 2nd ed. Madison, Wisconsin: American Society of Agronomy and Soil Science Society of America. Bowman, R. S., G. M. Haggerty, R. G. Huddleston, D. Neel, and M. M. Flynn. 1995. Sorption of nonpolar organic compounds, inorganic cations, and inorganic oxyanions by surfactant–modified zeolites. In ACS Symposium Series 594: 54–64. D. A. Sabatini, R. C. Knox, and J. H. Harwell, eds. Washington, D.C.: American Chemical Society. Brown, G. O., and B. Allred. 1992. The performance of syringe pumps in unsaturated horizontal column experiments. Soil Sci. 154(3): 243–249. Brownawell, B. J., H. Chen, J. M. Collier, and J. C. Westall. 1990. Adsorption of organic cations to natural materials. Envir. Sci. Tech. 24(8): 1234–1241. Bruce, R. R., and A. Klute. 1956. The measurement of soil moisture diffusivity. Soil Sci. Soc. Am. Proc. 20: 458–462. Burris, D. R., and C. P. Antworth. 1992. In situ modification of an aquifer material by a cationic surfactant to enhance retardation of organic contaminants. J. Contaminant Hydrol. 10: 325–337. Chen, Y. L., S. Chen, C. Frank, and J. Israelachvili. 1992. Molecular mechanisms and kinetics during the self–assembly of surfactant layers. J. Colloid and Interfacial Sci. 153(1): 244–265. Gaudin, A. M., and C. S. Chang. 1952. Adsorption on quartz, from an aqueous solution, of barium and laurate ions. Trans. AIME 193: 193–201. Hausmann, M. R. 1990. Engineering Principles of Ground Modification. New York, N.Y.: McGraw–Hill.

523

Hillel, D. 1980. Fundamentals of Soil Physics. San Diego, California: Academic Press, Inc. Huddleston, R. L., and R. C. Allred. 1967. Surface–active agents: Biodegradability of detergents. In Soil Biochemistry. A. D. McLaren and G. H. Peterson, eds. New York, N.Y.: Marcel Dekker, Inc. Karsa, D. R., and M. R. Porter. 1995. Biodegradability of Surfactants. London, U.K.: Blackie Academic and Professional. McDermott, D. C., J. McCarney, R. K. Thomas, and A. R. Renne. 1994. Study of an adsorbed layer of hexadecyltrimethylammonium bromide using the technique of neutron reflection. J. Colloid and Interface Sci. 162: 304–310. Mustafa, M. A., and J. Letey. 1969. The effect of two nonionic surfactants on aggregate stability of soils. Soil Sci. 107(5): 343–347. Phillips, S. B., W. R. Raun, and G. V. Johnson. 1997. Seasonal and long–term changes in nitrate– nitrogen content of well water in Oklahoma. J. Envi. Qual. 26: 1662–1637. Rakhshandehroo, G. R., R. B. Wallace, S. A. Boyd, and T. C. Voice. 1998. Hydraulic characteristics of organomodified soils for use in sorptive zone applications. Soil Sci. Soc. Am. J. 62(1): 5–12. Renshaw, C. E., G. D. Zynda, and J. C. Fountain. 1997. Permeability reductions induced by sorption of surfactant. Water Resources Res. 33(3): 371–378. Rosen, M. J. 1989. Surfactants and Interfacial Phenomena. 2nd ed. New York, N.Y.: John Wiley and Sons. Rutland, M. W., and J. L. Parker. 1994. Surface forces between silica surfaces in cationic surfactant solutions: Adsorption and bilayer formation at normal and high pH. Langmuir 10: 1110–1121.

524

Sullivan, E. J., D. B. Hunter, and R. S. Bowman. 1997. Topological and thermal properties of surfactant–modified clinoptilolite studied by tapping–mode atomic force microscopy and high–resolution thermogravimetric analysis. Clays and Clay Minerals 45(1): 42–53. Swisher, R. D. 1987. Surfactant Biodegradation. 2nd ed. New York, N.Y.: Marcel Dekker. Tadros, T. F. 1995. Surfactants in Agrochemicals. New York, N.Y.: Marcel Dekker. USDA–SCS. 1993. Design and Construction Guidelines for Considering Seepage from Agricultural Waste Storage Ponds and Treatment Lagoons. Technical Note Series 716. USDA Soil Conservation Service. Wagner, J., H. Chen, B. J. Brownawell, and J. C. Westall. 1994. Use of cationic surfactants to modify soil surfaces to promote sorption and retard migration of hydrophobic organic compounds. Envir. Sci. Tech. 28(2): 231–237. Watson, C. L., B. L. McNeal, and J. Letey. 1969. The effect of surfactants on the hydraulic conductivity of salt–affected soils. Soil Sci. 108(1): 58–63. West, C. W., and J. H. Harwell. 1992. Surfactants and subsurface remediation. Envir. Sci. Tech. 26(12): 2324–2329. Wray, W. K. 1986. Measuring Engineering Properties of Soil. Englewood Cliffs, New Jersey: Prentice–Hall, Inc.
