Environ. Sci. Technol. 1989, 23, 1365-1372
Enhanced Retention of Organic Contaminants by Soils Exchanged with Organic Cationst Jlunn-Fwu Lee, James R. Crum, and Stephen A. Boyd'
Department of Crop and Soil Sciences, Michigan State University, East Lansing, Michigan 48824-1325 The ability of soils to remove nonionic organic contaminants (NOCs) from water can be greatly enhanced by ion-exchange reactions of organic cations for natural inorganic exchange ions. Soils exchanged with organic cations of the form [(CH,),NR]+, where R is a Cg-C16 hydrocarbon, have significantly higher organic matter (OM) contents and display high sorptive uptake of common groundwater contaminants. Sorption coefficients (K values) of NOCs on hexadecyltrimethylammonium(HDTMA-) treated subsurface soil increased by -200 times. The K values for HDTMA-treated surface soils also increased significantly. When smaller organic cations, i.e., dodecyltrimethylammonium (DDTMA) and nonyltrimethylammonium (NTMA), were used, the exchange reactions were less complete, and the increases in OM contents and K values were less. Organic matter derived from the exchanged organic cations appeared to form an effective partition medium for NOCs. The organic phase derived from exchanged HDTMA was at least 10 times more effective than natural organic matter for removing NOCs from water; exchanged DDTMA was nearly as effective as HDTMA; however, NTMA was less effective. HDTMA was equally effective as a partition medium for NOCs when exchanged on soils of widely different textural classes.
Introduction Two mechanisms, in general, have been proposed to account for the sorption of organic species in soil-water systems. First, the sorption of organic chemicals may include ion exchange, protonation, hydrogen bonding, coordination/ion-dipole reactions with clays (1,2),and nucleophilic addition reactions with soil organic matter (3). These sorptive mechanisms account for the uptake of organic cations and some polar organic molecules (e.g., amines) on soil. The second major mechanism of organic chemical-soil interaction involves nonionic organic compounds (NOCs), which are the largest and most problematic group of organic contaminants. This mechanism has been described by Chiou et al. ( 4 , 5 )as a partition process involving nonpolar interactions between soil organic matter and NOCs. Sorption of this type depends almost exclusively on the solute water solubility (4, 5 ) and the soil organic matter content (6-9). The importance of soil organic matter in the sorption of NOCs by wet soils is demonstrated by the relative invariance of sorption coefficients of a given organic compound among soils, when normalized on the basis of the organic matter content of the soil (KOm).The established relationship between soil organic matter content and the uptake of NOCs has greatly simplified the predictions of sorption of organic contaminants in soil-water systems (5, 10-14). However, until recently there was no clear mechanistic understanding of the interactions of NOCs with soil organic matter. Lambert et al. (15) and Lambert (16,17) recognized that, in the sorption of neutral organic 'Journal article no. 13131 from the Michigan Agricultural Experiment Station. 0013-936X/89/0923-1365$01.50/0
pesticides on soils, the role of soil organic matter was similar to that of an organic solvent in liquid/liquid extraction. Recently, Chiou et al. (4,5) in studies of sorption of NOCs from water on soil showed that (1)the sorption isotherms of NOCs were linear over a wide range of aqueous concentrations, (2) soil uptake of NOCs exhibited a small heat effect, (3) no apparent solute competition occurred in multisolute systems, and (4) the sorption coefficient was dependent on the water solubility of the solute. On the basis of these observations of sorptive characteristics in soil-water equilibria, Chiou et al. ( 4 , 5 ) suggested that sorption of NOCs from water by soil is primarily due to the partitioning of NOCs into the soil organic (humus) phase. This is a process of solubilization where soil organic matter functions essentially as a bulk organic solvent phase and organic solutes distribute themselves between water and this organic phase. Adsorption of NOCs by soil minerals is suppressed in the presence of water because the poorly water soluble organic species cannot compete with strongly held water for the mineral adsorption sites (18). It is now apparent that low organic matter soils, clays, and aquifer materials have little sorptive capacity for removing of NOCs from water and are therefore ineffective in attenuating organic contaminant mobility. However, this limited sorptive capacity of low organic matter clays and soils can be greatly improved by replacing natural metal-exchange cations with large organic cations through ion-exchange reactions. This simple modification for improvement of the retardation capabilities of low organic matter soils and clays has been examined by Boyd and his co-workers (19-21). These authors used hexadecyltrimethylammonium (HDTMA) ions to replace naturally occurring metal ions, resulting in significantly higher organic matter contents and greatly enhanced sorptive capabilities for NOCs. The organic phase derived from HDTMA exchanged on soils (19) and on smectite clays (20, 21) was shown to be a powerful partition medium for the removal of NOCs from water. It was suggested that such modified soils and clays may be used to improve the retardation capabilities of low organic matter soils and aquifer materials and to enhance the containment capabilities of clay landfill liners and slurry walls. In the present article, the studies of soil modification for attenuating organic contaminant mobility have been expanded to include seven NOCs with water solubilities ranging over approximately 2 orders of magnitude. Many of the NOCs studied here have relatively high water solubilities and are common groundwater contaminants. In addition, the structural effects of different exchanged organic cations of the general form [(CH,),NR]+ in relation to the sorption of NOCs from water are established. Three organic cations, HDTMA (R = Cl6HS3), dodecyltrimethylammonium (DDTMA) (R = C12H2&and nonyltrimethylammonium (NTMA) (R = CSHlg),have been used for the preparation of soil-organic complexes from which different sorptive behaviors derived from different organic cations can be elucidated. Finally, soils having different textural compositions are used to assess the influence of clay content and cation-exchangecapacity (CEC)
0 1989 American Chemical Society
Environ. Sci. Technol., Vol. 23, No. 11, 1989
1365
on the sorptive properties of HDTMA-exchanged soils. Experimental Section
Soils. The following soils were used in this study: Marlette soil A and B, horizons (Fine-loamy,mixed, mesic Glossoboric Hapludalfs), Oshtemo soil B, horizon (Coarse-loamy, mixed, mesic Typic Hapludalfs), and St. Clair soil B, horizon (Fine, illitic, mesic Typic Hapludalfs). Organochemical-soil complexes were prepared by reacting 200 g of soil with HDTMA, DDTMA, or NTMA chloride salts. The organic cation salts were dissolved in water (1800 mL) and added in an amount equal to the CEC of the soil. After thorough mixing, the organochemical-soil complexes were dialyzed against distilled water until a negative chloride test was obtained. Then, the exchanged soils were freeze-dried and stored at room temperature. Chemicals. The organic chemicals used in this work were as follows: benzene, toluene, ethylbenzene, tetrachloroethylene (perchloroethene, PCE), trichloroethene (TCE), o-dichlorobenzene (o-DCB), and 1,2,4-trichlorobenzene (1,2,4-TCB). All reagents used in this study were analytical grade or better and were purchased from Aldrich Chemical Co., Milwaukee, WI. [Methyl-14C]HDTMAand -DDTMA were obtained from American Radiolabeled Chemicals, Inc. Specific activity was 52 mCi/mmol for HDTMA and 11mCi/mmol for DDTMA. Radiochemical purity was greater than 97%. Soil Properties. The soil CEC was taken to be the sum of extractable cations: A1 + Ca + Mg K + Na (nonbuffered), or H + Ca + Mg + K + Na (buffered). Exchangeable Ca, Mg, K, and Na were extracted with 1 N ammonium acetate at pH 7 and determined by atomic absorption spectrophotometry. Exchangeable Al from each soil was determined by displacement with 1 N KC1. The quantity of Al in the extract was analyzed by titrating with standardized base. Exchangeable H was extracted with BaC1,-triethanolamine at pH 8. Organic carbon analysis was performed by Huffman Laboratories, Inc., Golden, CO. Factors used to convert soil, HDTMA, DDTMA, and NTMA organic carbon to organic matter were 2,1.24, 1.27, and 1.29, respectively. Batch Equilibrium Experiments: (1) Adsorption Isotherms of Organic Cations on Soils. One gram of St. Clair, Marlette, or Oshtemo soils B, horizon and 10 mL of water were placed in Corex glass tubes. Portions of an ethanol solution containing 1:lOOO ['4CIHDTMA:HDTMA or [14C]DDTMA:DDTMAwere added in an amount equal to 0.25,0.5,0.75, 1, 1.5, and 2 times the CEC of soils. The mixed suspension were shaken for 24 h at 20 "C in a reciprocating shaker. The aqueous phase was separated by centrifugation at 8000 rpm for 30 min in a Sorvall RC-5C centrifuge equipped with an SS-34 rotor, and 1mL was assayed for 14Cby liquid scintillation counting. (2) Adsorption Isotherms of Organic Chemicals on Soil-Organic Complexes. Adsorption isotherms were obtained by adding varying quantities of test organic chemicals (22) to Corex glass tubes, which contained either 25 mL of water and 3 g of organochemical-soil complexes or 20 mL of water and 15 g of untreated soils. The tubes were closed with foil-lined screw caps. Benzene and toluene were added to the soils as the neat liquids, whereas the other organic compounds were added as methanol solutions with a Hamilton microliter syringe. To obtain equilibrium, mixed solutions were placed on a reciprocating shaker for 24 h at 20 "C. Preliminary kinetic investigations indicated that the sorption equilibrium was reached in 20 h. Phases were separated by centrifugation as above. Aliquots of the equilibrium supernatant (1 mL) were then transferred into glass vials containing 10 d of hexane (for
+
1366
Environ. Sci. Technol., Vol. 23, No. 11, 1989
o-DCB, 1,2,4TCB, TCE, and PCE) or carbon disulfide (for benzene, toluene, and ethylbenzene). The vials were closed with foil-lined screw caps and shaken vigorously for 1 h on a reciprocating shaker, and the concentrations of the test chemical were then determined by gas chromatography (GC). The amount sorbed by soil waa calculated by difference. The sorption coefficients (K values) of the linear isotherms correspond to the ratio of the concentration of the sorbed chemical in the soil to its equilibrium solution concentration in water. These values are obtained by dividing the solute uptake, Q (mg/kg), by the equilibrium concentration, C (mg/L), or from the slopes of the linear isotherms. Analytical Procedures. The hexane and carbon disulfide extracts were analyzed for the organic compounds by GC using a Hewlett-Packard Model 5890A equipped with both electron capture (for o-DCB, 1,2,4-TCB, TCE, and PCE) and flame ionization (for benzene, toluene, and ethylbenzene) detectors. The mass of organic compounds in the extracts was determined by measurement of peak areas with a Hewlett-Packard 3392A integrator. For 1,2,4-TCB and o-DCB, a 1.5% SP-2250/1.95% SP-2401 on 100/120 Supelcoport packed glass column (2.4 m X 6.4 mm id.) was used. For TCE and PCE, the separation was performed on 1% SP-lo00 coated on 60/80 Carbopack B packed steel column (1.8 m X 3.2 mm i.d.). Finally, a column packed with 5% SP-1200/1.75% Bentonite 34 (1.8 m X 3.2 mm i.d.) coated on 100-120-mesh Supelcoport support was used for separation of benzene, toluene, and ethylbenzene. Carrier gas was high-purity Nz at flow of 40 mL/min. Operating temperatures for injection port and detector were 270 and 300 "C, respectively. Each run was isothermal: oven temperatures were 80 "C for benzene and toluene, 100 "C for ethylbenzene, 150 "C for TCE, o-DCB, and 1,2,4-TCB,and 180 "C for PCE. All samples were run in duplicates and the recovery of these test chemicals ranged from 90 to 98%. The equilibrium concentrations measured were not adjusted for the recovery. Results
The physical and chemical properties of untreated Marlette soil and the Marlette soil treated with three different organic cations are shown in Table I. The three organic cations used were all of the form [(CH,)&VFi]+,and R varied in size from a CIS(HDTMA) to a Clz (DDTMA) to a C9 (NTMA) unbranched hydrocaibon. When the Marlette A and Bt horizon soils were treated with HDTMA, DDTMA, and NTMA ions, the organic carbon (OC) contents increased significantly. The organic matter (OM) contents of the treated soils are the sums of indigenous soil OM plus OM derived from the exchanged organic cations. For example, the total OM content of HDTMA-treated Marlette soil B, horizon (4.85%) was obtained by adding the indigenous soil OM (0.6%) plus OM derived from HDTMA ions (4.25%). As expected, the larger organic cation (HDTMA) was more effective for the contribution of OM than the smaller organic cations (DDTMA and NTMA). Based on a CEC of 14.6 mequiv/ 100 g, and assuming complete exchange of the added organic cation, the amounts of synthetic OM derived from the organic cations are ca. 4.0,3.2, and 2.6% for HDTMA, DDTMA and NTMA, respectively. The actual observed synthetic OM contents were 4.25, 2.13, and 1.74% for HDTMA, DDTMA and NTMA, showing incomplete exchange of added DDTMA and NTMA. In the treated Bt soils most of the total organic matter is derived from the exchanged organic cations because of the low indigenous soil OM levels. In contrast, in the treated A horizon soils, natural soil OM exceeds the added synthetic OM.
Table I. Properties of the Untreated Marlette Soil, and the Marlette Soil Treated with Three Different Organic Cations
B, horizon property
untreated
CEC, mequiv/100 g particle size, % sand silt clay PH organic carbon, % organic matter, % soil organic matter organic matter derived from organic cations total organic matter
225r/
HDTMA
treated DDTMA
NTMA
untreated
A horizon treated HDTMA DDTMA
14.6
16.4
38.8 31.6 29.6 5.4 0.30
3.71
1.98
1.18
56.6 22.0 21.4 6.4 2.59
6.48
4.31
0.60
0.60
0
4.25
0.60
4.85
0.60 2.13 2.73
0.60 1.14 1.14
5.18 0 5.18
5.18 4.85 10.03
5.18 2.25 1.43
n
Marlette
n Oshtemo
45J OO
2Aqueous 4Phase C6o n r , C (,umole/ml) 8 10
12
0
0
2.5
5 7.5 10 12.5 Aqueous Phase C o n r , C (pmole/ml)
5
Flgure 1. Adsorption isotherms of hexadecyltrlmethylammonlum (HDTMA) on Bt horizons of three solls.
Flgure 2. Adsorption isotherms of dodecyltrimethylammonlum(DDTMA) on 6,horlzons of three soils.
The properties of three different soils either untreated or treated with HDTMA are shown in Table 11. The soils were selected to give a range of textural classes, CECs, and clay contents. Of these, the Oshtemo (sandy loam) has the lowest clay content and the lowest CEC. The St. Clair and the Marlette soils have similar CEC values, which are significantly higher than the Oshtemo soil. Table 11shows that the OM contents of the treated soils are dependent on the CEC of the so& the treated Oshtemo soil, with the lowest CEC, had the lowest total OM content. As mentioned above, the soil-organic complexes were prepared by adding organic cations in an amount equal to the (nonbuffered) CEC of soils. The adsorption of HDTMA ions by the three soils (B, horizons) is shown in Figure 1. HDTMA was added in varying amounts up to twice the CEC of the soil. Adsorption isotherms of HDTMA ions on soils were typical
for high-affinity ion-exchange reactions. The isotherm shape is type I in Brunauer's classification (23),showing a high degree of preference of the soil for the added HDTMA exchange ions. As illustrated in Figure 1, the HDTMA ions are very strongly adsorbed from solution by the soils. The isotherms show very little HDTMA in solution when added in an amount less than or equal to the CEC of soils, and some continued sorption of HDTMA ions in excess of the (nonbuffered) CEC. However, when HDTMA ions were added at 1.5 and 2 times the CEC of the soil, the solution equilibrium concentrations increased significantly. The adsorption of [I4C]DDTMA on the three soils is shown in Figure 2. These isotherms clearly show that, compared to HDTMA, DDTMA is a less competitive exchange ion. As a result, the exchange sites were not completely saturated with DDTMA even when the organic Environ. Sci. Technol., Vol. 23, No. 11, 1989
1367
Table 11. Properties of Three Different Soils (B, Horizons) either Untreated or Treated with HDTMA property CEC, mequiv/100 g buffered nonbuffered particle size, % sand silt clay illite' vermiculite' kaolinite" chlorite" quartza PH organic carbon, % organic matter, % soil organic matter HDTMA organic matter total organic matter
St. Clair
untreated Marlette
Oshtemo
22.6 18.3
18.6 14.6
4.5 3.5
21.0
38.8 31.6 29.6
89.3 4.4 6.3
xxx xxx
xxx
34.9 44.1 xxxx xxx xxx
St. Clair
HDTMA-treated Marlette
Oshtemo
3.25
3.71
0.83
0.88 3.50 4.38
0.60 4.25 4.85
0.22 0.90
xxx xxx xx
xxx xx
xx
xx
xx
6.72 0.44
5.40 0.30
5.84 0.11
0.88
0.60
0.22
0
0
0
0.88
0.60
0.22
1.12
'xx, small (5-15%); xxx, moderate (1530%);xxxx, abundant (3040%).
1375 -
. 1100
toluene
-
2oo 100
'0
25 50
75
100 125 150
Aqueous Phase Concentration, C [me/lJ Figure 3. Sorption of benzene, toluene, and ethylbenzene on untreated (triangles) and HDTMA-treated (circles) St. Clair (a-c) and OsMemo (d-f) B, horizon soils.
cation was added at twice the CEC of the soils. This result corresponds to the measured OM contents, which show complete exchange of added (1 equiv) HDTMA and incomplete exchange of added DDTMA (Table I). The release of exchanged [I4C]HDTMAfrom the Marlette B,horizon soil after mixing with aqueous solutions of CaC12 was also measured (data not shown). Calcium chloride was added as an aqueous solution in amount equal 1368 Environ. Sci. Technol., Vol. 23, No. 11, 1989
to 10, 50, and 100 times the amount of exchanged HDTMA; however, no [I4C]HDTMA was released into solution. Figure 3 shows the dramatic effect of HDTMA treatment on the sorptive properties of two subsurface soils. Sorption isotherms of benzene, toluene, and ethylbenzene on the untreated and HDTMA-treated St. Clair and Oshtemo soil B, horizons are shown in Figure 3. As ex-
Table 111. Sorption Coefficients of Organic Chemicals on the Untreated Marlette Soil and on the Marlette Soil Treated with Three Different Organic Cations
sorption coeff
B, horizon treated HDTMA DDTMA
untreated
A horizon treated HDTMA DDTMA water sol., ppm log Kow
NTMA
untreated
2.08
0.57 0.88 1.29 2.60 2.79 7.25 40.25
7.80 26.70 20.38 37.03 43.53 119.50 293.75
4.75 5.82 10.76 14.93 17.94 47.37 109.90
1.04 1.23 1.39 1.70 1.73 2.14 2.89
1.89 2.42 2.30 2.56 2.63 3.07 3.47
1.81 1.89 2.16 2.30 2.37 2.80 3.17
K benzene TCE toluene PCE ethylbenzene 0-DCB 1,2,4-TCB log Kom benzene TCE toluene PCE ethylbenzene 0-DCB 1,2,4-TCB
0 0 0
0.17 0.41 0.80 5.60
,
16.50 29.20 26.90 27.30 62.51 162.59 302.73
7.78 6.32 14.36 11.93 25.94 65.52 86.33
2.53 2.78 2.74 2.75 3.11 3.52 3.97
2.46 2.36 2.72 2.64 2.98 3.38 3.50
1.64 1.83 2.14 2.97
3.32 4.53 12.93 2.08 2.28 2.42 2.87
1780 1100 520 200 155 150 20 2.13 2.61 2.69 3.38 3.15 3.38 4.02
Table IV. Sorption Coefficients of Three Different Soils (B, Horizons) either Untreated or Treated with HDTMA sorption coeff
St. Clair
untreated Marlette
St. Clair
HDTMA-treated Marlette
Oshtemo
0.21
15.96 34.47 75.54
16.50 26.90 62.51
3.83 6.59 12.84
1.97
2.56 2.89 3.23
2.53 2.74 3.11
2.53 2.76 3.06
Oshtemo
K benzene toluene ethylbenzene log Kom benzene toluene ethylbenzene
0
0.17 0.81 1.29 1.96
0 0 0.41
1.83
p e d , the untreated soils display very low uptake of these organic chemicals from water, presumably due their low OM contents; the isotherms of benzene and toluene could not be distinguished from the blank (no soil) samples. In contrast to the lack of sorption by the untreated soils, very high uptake of benzene, toluene, and ethylbenzene was found on both HDTMA-treated soils. This is reflected in the dramatic increase in the slopes of isotherms shown in Figure 3. All isotherms were linear over a wide range of aqueous-phase concentrations. Figure 4 compares the effect of HDTMA treatment on the sorptive properties of a surface (A) and subsurface (BJ horizon. Sorption isotherms for benzene, toluene, and ethylbenzene on the untreated and HDTMA-treated Marlette soil A and B, horizons are presented. Again, uptake by the B,horizon of all chemicals tested was greatly enhanced in the HDTMA-treated soil. For the treated A horizon, uptake was also enhanced significantly, but the relative increase was lower than for the B, horizon soil because of the higher idigeneous soil OM content in the A horizon. Although the HDTMA-treated Marlette A horizon soil is -50% natural soil OM and 50% HDTMA OM (Table I), sorption was much more than doubled in the treated soil. The sorptive behaviors of three different HDTMAtreated soils (St. Clair, Marlette, and Oshtemo) for benzene, toluene, and ethylbenzene can be compared from the isotherms shown in Figures 3 and 4. Soils with higher CEC values (Marlette N St. Clair > Oshtemo) resulted in steeper isotherms (higher slopes). The data show that in overall sorptive behavior the HDTMA-treated St. Clair and Marlette soils were similar and were higher than HDTMA-treated Oshtemo soil. This follows the same trend in total OM contents of the HDTMA-treated soils
+I
0 0
that decrease in the order Marlette N St. Clair > Oshtemo. Figure 5 shows the isotherms of benzene, toluene, and ethylbenzene uptake by Marlette soil B, horizon having HDTMA, DDTMA, and NTMA as the exchanged cations. A positive relation is apparent between the amount of sorbed benzene, toluene, and ethylbenzene and alkyl chain length of exchanged organic cations. As the chain length decreases in going from HDTMA (CI6)to DDTMA (C12) to NTMA (C&,the degree of uptake of the organic species from water decreased. The decrease in sorption follows the total OM contents, which decrease in the same order: HDTMA > DDTMA > NTMA. Again, sorption isotherms were linear over a wide range of aqueous-phase concentration. Except for the uptake of benzene by NTMAtreated Oshtemo soil, the linear correlation coefficients (r) for all isotherms are greater than 0.95. Tables I11 and IV show the water solubilities, octanolwater partition coefficients (Kow),sorption coefficients ( K ) , and the OM normalized KO, values (Klfraction OM) of the seven organic chemicals studied. Table I11 shows the effect of three different exchange cations on the sorption coefficients of the Marlette soil. Table IV shows the effect of HDTMA treatment on the sorption coefficients of three different soils. Examination of Tables I11 and IV reveals that the K values increased with the OM content of soil and were higher for compounds with lower water solubilities. The relation between sorption and OM content can be seen by comparing the K values from the treated and untreated soils, and from soils that have been treated with different organic cations. For example, the K values obtained from the uptake of PCE on untreated (0.6% OM) and HDTMA-treated (4.85% OM) Marlette soil Bt horizon are 0.17 and 27.3, respectively (Table 111). Similarly, the K Environ. Sci. Technol., Vol. 23,
No. 11, 1989 1369
1800
1350
750
11
2400
c 0
L
0
t
0
200 400 600 800 1000 1200
'0
t i
ethylbenzene
toluene 1200
u
1 4
40 80 120 160 200 240
L
1
250 10
OO
20 30
40
50
60
c
. )
u
C 0
u
-
-
'
-
750
-B
-
500
-
-
1000-
-
B
ID
0
0
100 200 300 400 500 600
0
ethylbenzene
250
60 120 180 240 300 360
O
w
5
0
Aqueous Phase Concentratlon, C (mW4 Figure 4. Sorption of benzene, toluene, and ethylbenzene on untreated (triangles) and HDTMA-treated (ckcles) Marlette A and 6,horizon soils.
Aqueous Pham Concentration, C (mu4 Flgure 5. Sorption of benzene (a), toluene (b), and ethylbenzene (c) on treated Martette B, horizon soil: NTMA-treated (circles), DDTMA-treated (triangles), HDTMA-treated (squares).
values obtained from treated soils decrease in the same order as their OM contents: HDTMA-treated > DDTMtreated > NTMA-treated soils. For example, ethylbenzene 1370 Environ. Sci. Technol., Vol. 23, No. 11, 1989
on the treated Marlette Bthorizon soils had K values of 62.51, 25.94, and 4.53 for HDTMA-, DDTMA-, and NTMA-treated soils, respectively (Table 111).
The K values of organic chemicals were also affected profoundly by their water solubilities. As the water solubilities decreased in going from benzene (1780 ppm) to 1,2,4-TCB (20 ppm), the K values for sorption by the untreated Marlette A horizon soil increased from 0.57 to 40.25 (Table 111). Similarly K values for the treated soils also increased with decreasing water solubilities of the organic solutes (Tables I11 and IV). The sorption coefficients K normalized for the OM content (KO,)are also given in Tables I11 and IV. Some KO, values could not be obtained from untreated soil B, horizons because the sorption coefficients were very small and K values could not be determined with sufficient accuracy. It can be seen from Table I11 that, for a given compound, the log KO, values decreased in the order HDTMA N DDTMA > NTMA > untreated soil. Table IV gives the calculated log K, values of benzene, toluene, and ethylbenzene for three untreated and HDTMA-treated soils. For all three soils, the log KO, values tended to converge on a single value for each organic chemical. For example, the three HDTMA-treated soils gave log KO, values of 3.1-3.2 for ethylbenzene, and the three untreated soils gave log Komvalues of approximately 1.9. In all cases, KO, values for treated soils were greater than soil KO, values for untreated soils.
Discussion The results presented here clearly demonstrate that the sorptive properties of surface and subsurface soils for NOCs can be greatly enhanced by simple ion-exchange reactions of large organic cations for naturally occurring inorganic exchange ions. The ion-exchange reaction of large organic cations (e.g., HDTMA) results in the nearly stoichiometric displacement of the small naturally occurring inorganic exchange ions. This can be seen clearly in the adsorption isotherms of [14C]HDTMA by the three subsurface soils. It is apparent from the very sharp initial rise of these isotherms that HDTMA is highly preferred as an exchange ion and will easily displace inorganic exchange ions. For the smaller organic cations studied, where the hydrophobic tail was C12or Cg hydrocarbon, the exchange reaction was incomplete, showing that these ions are less competitive than HDTMA for exchange sites in soils. Exchanged HDTMA ions, once bound, were held essentially irreversibly on the clay surface. This is shown by the inability of high ionic strength CaC1, solutions to displace exchanged [“CIHDTMA. In this work we c o n f i i and expand our previous results (19) on the increased sorptive capabilities of HDTMAexchanged soil by examining several additional organic compounds that are representative of common groundwater contaminants. Treatment of surface or subsurface soil with HDTMA dramatically increased the uptake from water of all compounds tested. For example, the sorption coefficients of the treated subsurface soil increased by -200 times for toluene, PCE, ethylbenzene, and dichlorobenzene. The more water soluble compounds (e.g., benzene and TCE) were not measurably removed from water by the untreated subsurface soils, but high uptake was observed in the treated soils. For the surface soils, sorption coefficients of the untreated samples were higher than for the untreated subsurface soil because of the high indigenous soil OM content of the A horizon soil. However, the HDTMA-exchanged A horizon soil also displayed enhanced uptake of all NOCs tested. For HDTMA-treated surface soil the increase in K was 15 times, although the organic matter content of the Marlette A horizon soil was only approximately doubled by exchange with HDTMA. That the K values for the treated surface soil increased
-
to a much larger extent ( 15X) than the OM levels (-2X) demonstrates that the HDTMA-derived organic matter is N
a much more effective sorbent for NOCs than natural soil
OM. A more accurate evaluation of the sorptive capabilities of natural soil OM and synthetic OM derived from exchanged cations can be obtained by comparing the corresponding KO, values, which indicate the effectiveness of different organic phases on a unit weight basis. For the untreated A horizon soil, all OM is natural, whereas for the treated B, horizon soils, OM is derived primarily from the exchanged organic cations. The K , values of benzene, TCE, toluene, PCE, ethylbenzene, and o-DCB on the HDTMA-treated Marlette B, soil are 10-30 times higher than the corresponding Kom values on the untreated Marlette A horizon soil. It can be therefore concluded that OM derived from added HDTMA is approximately 10-30 times more effective on a unit weight basis than natural soil OM for removing NOCs from water. Thus, not only can the OM content of soil be increased significantly by this simple soil modification technique, but the added OM is highly effective and can be expected to greatly enhance the retention of NOCs by soil. The data presented here strongly suggest that the OM derived from the exchanged quaternary ammonium ions behaves as a powerful partition medium in the uptake of NOCs from water. This is a process of solubilization of NOCs in the synthetic organic phase formed by the conglomeration of the alkyl tails of exchanged organic cations. This process is essentially identical with the partitioning of solutes from water into an immisible bulk organic solvent phase such as hexane or octanol. The linearity of the isotherms, and the similarity of the log Komvalues of the HDTMA-treated Marlette B, horizon soil to the corresponding log KO,values (Table 1111,strongly support this mechanism. Although natural soil OM and the synthetic (HDTMA) organic matter phases are mechanistically similar in uptake of NOCs from water, the effectiveness of the latter is much greater because the HDTMA-derived OM provides a better nonpolar solvent environment than natural soil OM, which has a high polar functional group content. An additional objective of the work was to examine the sorptive behavior of soil exchanged with three different organic cations containing alkyl moieties of different size, ranging from Cgto CI6. Obviously, as the size of the organic cation decreases, the increase in OM after cation exchange is less. However, the effectiveness of the organic phases derived from the different cations can be compared by comparing the Km values. This comparison of the treated Marlette Bt horizon soils reveals that the organic phases composed of HDTMA and DDTMA were similar, with HDTMA forming a slightly more effective partition medium. When the Cg ion (NTMA) was used, a decrease in the KO, value occurred, although these values were still considerably higher than Km values for natural OM. Apparently, as the alkyl chain length decreases below CI2, the hydrocarbon moieties become more isolated and form a less effective partition medium than with the larger cations, i.e., HDTMA and DDTMA. The sorptive capabilities for NOCs of three different soils exchanged with HDTMA were also compared. These subsurface soils differed primarily in their CECs, which were relatively high for the St. Clair and Marlette soils and low for the Oshtemo soil. As expected, the OM contents of the HDTMA-exchanged Marlette and St. Clair soils were higher than the Oshtemo soil, and as a result, the K values for benzene, toluene, and ethylbenzene decreased Environ. Sci. Technol., Vol. 23,No. 11, 1989
1371
in the order St. Clair N Marlette > Oshtemo (Table IV). However, the KO, values for these three HDTMA-treated soils (Table IV) were nearly identical, indicating that the HDTMA-organic phases formed on each soil were equally effective. The results presented here demonstrate that soil modification by exchange reactions of organic cations is highly effective for enhancing the sorptive capabilities (for NOCs) of soils of widely different textural classes, including those with relatively low clay contents. For the actual field application of this work, organic cations with large (CIJ hydrocarbon moieties would be most effective because (1) they result in nearly stoichiometric ion-exchange reactions, (2) they are bound irreversibly, (3) they increase the OM content significantly, and (4) they form a highly effective partition medium for removing NOCs from water. The soil modification described here represents the only plausible way to enhance the sorptive capabilities of soils and subsurface materials for organic contaminants. This technology has direct application in the stabilization of contaminated soils and could be used to improve the containment capabilties of clay barriers. Registry No. HDTMA, 6899-10-1; DDTMA, 10182-91-9; NTMA, 3581923-9; TCE,79-01-6;PCE, 127-184; O-DCB, 95-50-1; 1,2,4-TCB,120-82-1; benzene, 71-43-2; toluene, 108-88-3; ethylbenzene, 100-41-4.
Literature Cited (1) Mortland, M. M. Adv. Agron. 1972,22, 75-117. (2) Theng, B. K. G. The Chemistry of Clay-Organic Reactions; Wiley: New York, 1974; p 343. (3) Parris, G. E.Residue Rev. 1980, 76, 1-30. (4) Chiou, C. T.;Peter, L. J.; Frees, V. H. Science 1979,206, 831-832. (5) Chiou, C. T.; Porter, P. E.;Schmedding, D. W. Environ. Sci. Technol. 1983, 17, 227-231.
1372
Environ. Sci. Technol., Vol. 23, No. 11, 1989
(6) Briggs, G. G. Nature 1969, 223, 1288. (7) Karickhoff, S. W.; Brown, D. S.; Scott, T. A. Water Res. 1979,13,241-248. ( 8 ) Garbarini, D. R.; Lion, L. W. Environ. Sci. Technol. 1986, 20, 1263-1269. (9) Hassett, J. J.; Means, J. C.; Banwart, W. J.; Wood, S. A.; Khan, A. J . Environ. Qual. 1980,9, 184-186. (10) Boucher, F. R.; Lee, G. F. Environ. Sci. Technol. 1972,6, 538-543. (11) Means, J. C.; Wood, S. G.; Hassett, J. J.; Banwart, W. L. Environ. Sci. Technol. 1982, 16, 93-98. (12) Dzombak, D. A.; Luthy, R. G. Soil Sci. 1984,137,292-308. (13) Chiou, C. T.; Shoup, T. D.; Porter, P. E. Org. Geochem. 1985, 8, 9-14. (14) Bowman, M. C.; Schechter, M. S.; Carter, R. L. J . Agric. Fd. Chem. 1965,13, 360-365. (15) Lambert, S. M.; Porter, P. E.; Schieferstein, H. Weeds 1965, 13, 185-190. (16) Lamert, S. M. J. Agric. Food Chem. 1967, 15, 572-576. (17) Lamert, S. M. J. Agric. Food Chem. 1968, 16, 340-343. (18) Chiou, C. T.; Shoup, T. D. Environ. Sci. Technol. 1985,19, 1196-1200. (19) Boyd, S. A.; Lee, J. F.; Mortland, M. M. Nature 1988,333, 345-347. (20) Boyd, S. A.; Mortland, M. M.; Chiou, C. T. Soil Sci. SOC. Am. J . 1988,52, 652-657. (21) Mortland, M. M.; Shaobai, S.; Boyd, S. A. Clays Clay Miner. 1986, 34, 581-585. (22) Lee, J. F.; Mortland, M. M.; Chiou, C. T.; Boyd, S. A. J . Chem. SOC.,Faraday Trans. 1, in press. (23) Brunauer, S. The Adsorption of Gases and Vapors;Oxford University Press: New York, 1944; p 150.
Received for review Janaury 30,1989. Accepted July 11,1989. The research contained in this paper was supported by Grant 14-08-0001-G1622 from the U.S. Geological Survey, Department of the Interior; however, this does not necessarily represent the policy of that agency or of the Federal Government.
