ction

bentonitecontent

solutes tof the clayf benzenesolubility.t alone maye sorptionon contentic-carboning organic-

sed by the

Journal of Colloid and Interface Science 266 (2003) 251–258www.elsevier.com/locate/jcis

Nonionic organic solute sorption onto two organobentonites as a funof organic-carbon content

Shannon L. Bartelt-Hunt, Susan E. Burns, and James A. Smith∗

Program of Interdisciplinary Research in Contaminant Hydrogeology, Department of Civil Engineering, University of Virginia, P.O. Box 400742,Charlottesville, VA 22904-4742, USA

Received 1 October 2002; accepted 6 June 2003

Abstract

Sorption of three nonionic organic solutes (benzene, trichloroethene, and 1,2-dichlorobenzene) to hexadecyltrimethylammonium(HDTMA bentonite) and benzyltriethylammonium bentonite (BTEA bentonite) was measured as a function of total organic-carbonat quaternary ammonium cation loadings ranging from 30 to 100% of the clay’s cation-exchange capacity. Sorption of all threeHDTMA bentonite was linear and sorption of all three solutes by the HDTMA bentonite increased as the organic-carbon content oincreased. 1,2-Dichlorobenzene sorbed most strongly to HDTMA bentonite, followed by benzene and TCE. The stronger sorption oto HDTMA bentonite compared to TCE was unexpected based on a partition mechanism of sorption and consideration of soluteLogKoc values for all three solutes increased with organic-carbon content. This suggests that the increased organic-carbon contennot explain the observed increase in sorption capacity. Sorption of the three solutes to BTEA bentonite was nonlinear and solutincreased with decreasing organic-carbon content, with a peak in the magnitude of solute sorption occurring at an organic-carbcorresponding to 50% of CEC. Below 50% of CEC, sorption of all three solutes to BTEA bentonite decreased with decreasing organcontent. Surface area measurements indicate that the surface area of both organobentonites generally decreased with increascarbon content. Since nonionic organic solute sorption to BTEA bentonite occurs by adsorption, the reduced sorption is likely caureduction in surface area corresponding to increased organic-cation loading. 2003 Elsevier Inc. All rights reserved.

Keywords: Adsorption; Bentonite; Benzene; 1,2-Dichlorobenzene; Organic cations; Organoclay; Partitioning; Trichloroethylene

benin ao-

tionsnicale rethaten-cityun-

forare

larun-

laysnt fa-rplingcialr en-

tood-tionor-

evi-rnary

nc-tion

of

1. Introduction

Numerous studies have investigated the use of organotonites as potential sorbents for organic contaminantswide variety of environmental applications [1–5]. Organbentonites are produced by the exchange of organic ca(typically a quaternary ammonium cation) for the inorgacations (e.g., Ca2+, Na+) that naturally occur on the internand external mineral surfaces of bentonite. This exchangsults in a material with physical and chemical propertiesare vastly different from unmodified bentonite. Organobtonites are organophillic and have a large sorption capafor aqueous nonpolar organic compounds. In contrast,modified bentonite is hydrophilic and a poor sorbentthese compounds [6–8]. Additionally, organobentonites

* Corresponding author.E-mail address: jsmith@virginia.edu (J.A. Smith).

0021-9797/$ – see front matter 2003 Elsevier Inc. All rights reserved.doi:10.1016/S0021-9797(03)00617-9

-

-

able to retain low permeability in the presence of nonpoliquids [9] and are stronger and less compressible thanmodified clay [10]. Because of these properties, organochave been investigated as additives to waste containmecilities, landfill liners [1], slurry walls [2], and wastewatetreatment processes [3,4] and as adsorbents for air samof airborne organic contaminants [5]. Several commercompanies manufacture and market organobentonites fovironmental applications.

Although sorption of nonionic organic compoundsmany, if not all, organobentonites is stronger than to unmified bentonite, the magnitude and mechanism of sorpis highly dependent on the molecular structure of theganic cation that is used to modify the clay surface. Prous research has shown that when the exchanged quateammonium cation has one or more long-chain alkyl futional groups (e.g., hexadecyltrimethylammonium), sorpof relatively nonpolar organic pollutants is several orders

252 S.L. Bartelt-Hunt et al. / Journal of Colloid and Interface Science 266 (2003) 251–258

chareous15].n hathe

anicam

ort-o-

ear17].n ha

hersorpandentbytternicthe

beecitytra-hatoils.c-edeneiitysedorp-ve ahas

andwithizedthat

forthesur-

ef-ionsthisnictwoiten,

eact ofluteos-

2nd

syn-er-on-ies,

ofaseandbase

syn-

had

erendrop-was

-) to-free).

nat-them-

ofthe

n,de-

of

tained in

magnitude greater than on untreated bentonite and isacterized by linear, noncompetitive isotherms and aqusolubility dependence of the sorption process [1,6,7,11–Based on these observations, the mechanism of sorptiobeen attributed to solute partitioning between water andorganic phase created by the alkyl chains of the orgcations [6]. Conversely, when the exchanged quaternarymonium cation has benzyl, phenyl, and/or relatively shchain alkyl functional groups (e.g., benzyltriethylammnium), sorption is characterized by competitive, nonlinisotherms with no clear solubility dependence [1,6,8,15–Based on these observations, the mechanism of sorptiobeen attributed to a physical adsorption process [6].

In general, natural soils and clays having a higorganic-carboncontent are considered to have a greatertive capacity for nonionic organic compounds than soilsclays with a comparatively lower organic-carbon contsince nonionic organic solute sorption generally occursa partition mechanism between natural soil organic maand water. Few studies have examined the effect of orgacarbon loading on sorption to organobentonites, andresults from the studies that have been performed haveinconsistent. Smith et al. [18] found that sorptive capaincreased with increasing organic-cation loading for techloromethane sorption to BTEA bentonite, a result tmight be expected considering the behavior of natural sSimilarly, Boyd et al. [8] found that the sorption capaity of HDTMA bentonite for TCE and benzene increaswith organic-carbon content. In a study of perchloroethylsorption to zeolite modified with the HDTMA cation, Land Bowman [19] found that while the sorption capacof a surfactant-modified zeolite increased with increaorganic-cation loading, there was a minimal increase in stion for further increases in organic-carbon content abomonolayer coverage. However, a recent technical noteshown that for certain combinations of organobentoniteorganic solute, sorption capacity may actually decreaseincreasing organic-carbon content [20]. It was hypothesthat the reduction in sorptive capacity for organoclayssorb by an adsorption mechanism was caused by themation of positively charged dimers on the surface ofclay that block access to the sorptive sites on the clayface.

To date, no study has systematically investigated thefects of the loading of quaternary ammonium organic caton the magnitude and mechanism of solute sorption. Instudy, we examine the sorption of three nonionic orgasolutes (TCE, benzene, and 1,2-dichlorobenzene) toorganobentonites: BTEA bentonite and HDTMA bentonat cation loadings from 30% to 100% of CEC. In additiospecific surface area was measured for both clays atcation loading. Our objective was to elucidate the effecorganic-cation loading on the magnitude of nonionic sosorption to organobentonites from water and to identify psible mechanisms responsible for the observed results.

-

s

-

s

-

-

n

-

h

Table 1Chemical and physical properties of organic solutesa

Compound Molecular weight Solubility, 25◦C logKow(g/mol) (mg/l)

1,2-Dichlorobenzene 147.00 148 3.34–3.55Trichloroethylene 131.39 1100 2.29–3.30Benzene 78.11 1750 1.56–2.20

a All values taken from Groundwater Chemicals Desk Reference,ed. [24].

2. Experimental

2.1. Materials

The 16 organobentonites used in this study werethesized from Wyoming bentonite (sample TG-50, Amican Colloid Company). The natural organic-carbon ctent of the base clay was 0.2% (Huffman LaboratorIncorporated) with a cation-exchange capacity (CEC)69.1 meq/100 g (Hazen Research, Incorporated). The bbentonite soil was composed of 3.6% sand, 7.3% silt,89.1% clay. The exchangeable inorganic cation on thebentonite clay was primarily sodium.

Two quaternary ammonium compounds were used tothesize the clays: HDTMA bromide [(CH3)3NC16H33Br]and BTEA chloride [(C2H5)3NCH2C6H5Cl]. Both com-pounds were obtained from Aldrich Chemical Company,a chemical purity of 99%, and were used as received.

Benzene, 1,2-dichlorobenzene (1,2-DCB), and TCE wobtained from Aldrich Chemical Company (99% purity) awere used as received. Select chemical and physical perties of each solute are given in Table 1. Each solutemixed with its corresponding [14C] isotope (Aldrich Chemical Company, 99% chemical and radiochemical puritycreate a radiolabeled stock solution. Deionized, organicwater was used in all experiments (Barnstead Nanopure

2.2. Methods

The organoclays were prepared by exchanging theurally occurring inorganic cations on the surface ofWyoming bentonite with either of the two quaternary amonium cations at 30, 40, 50, 60, 70, 80, 90, and 100%the clay’s CEC. The quantity of organic cation added tobentonite was determined by

(1)f = Mcation

CECMclayGMWcationZ,

wheref = fraction of CEC satisfied by the organic catioMcation= mass of organic cation required to achieve thesired fraction of CEC, CEC= cation-exchange capacitythe base clay,Mclay = mass of the base clay, GMWcation=gram molecular weight of the organic cation, andZ = molesof charge per equivalent.

The mass of organic cation required to satisfy a cerpercentage of the cation-exchange capacity was dissolv

S.L. Bartelt-Hunt et al. / Journal of Colloid and Interface Science 266 (2003) 251–258 253

riedml

hetin-redionwassthecol-ture

e tocon-ing-ps..0 gsureilib-t thhadsol-

2ffi-ve

andund

stra-

rred.liq-

asThewas

re-ctorotheto

ter-atedri-

om-oot-21].ing azer,ea-

us-her-

had

the-letes atic-to

ultsarehe

n-mayply-nemsndin

delbeby

tolso

on

water at room temperature. The volume of water used vawith the mass of cation to be dissolved, ranging from 500to 1 l. A quantity of 500 g of base clay was mixed into taqueous cation solution and the solution was stirred conuously for 20 min. The clay mixture was covered and stoat room temperature for 24 h. After the 24-h equilibratperiod, the standing water was poured off and the clayoven-dried at 110◦C. Once dry, the clay was ground firwith a mortar and pestle and then again in a jar mill. Tclay particles passing through a number 40 sieve werelected and stored in air-tight containers at room temperauntil use.

Sorption of TCE, 1,2-dichlorobenzene, and benzenthe sixteen organobentonites were quantified using aventional batch equilibration technique [6,7,14,15]. Varyamounts of sorbent, water, and14C-labeled solute were combined in 15-ml glass centrifuge tubes with Teflon-lined caThe mass of sorbent was 0.5 g for BTEA bentonite and 2for HDTMA bentonite; these masses were chosen to inthat 30 to 90% of the added solute was sorbed at equrium. The masses of solute added were chosen so thafinal range of equilibrium aqueous solute concentrationsan upper limit approaching 50% of the solute’s aqueousubility in order to identify any isotherm nonlinearities.

After assembly, the batch reactors were shaken at 2◦Cin the dark for 24 h. This was considered to be a sucient time to reach equilibrium, as Deitsch et al. [11] hashown previously that sorption of carbon tetrachloride1,2-dichlorobenzene onto three organobentonites was foto be very rapid, with equilibrium occurring within the firfew hours of solute-sorbent contact time. After equilibtion, the reactors were centrifuged at 2400g for 60 min at22◦C. A sample of 500 µl of the supernatant was transfeto 5 ml of scintillation cocktail in a 7-ml scintillation vialThe radioactivity was quantified with a Packard 1900TRuid scintillation analyzer and the measured radioactivity wrelated to aqueous concentration by a standard curve.sorbed concentration of the solute in each batch reactorthen calculated by difference.

For each isotherm experiment, three additional batchactors were prepared for quality assurance. These reawere used to quantify solute losses caused by processesthan sorption to the sorbent (e.g., volatilization, sorptionTeflon), to quantify the background radiation, and to demine if the sorbent, water, or reactor has been contaminwith radioactivity. Average solute recovery in all expements was greater than 95%.

The isotherm model chosen for each solute–sorbent cbination was based on an evaluation of the residual rmean-square error (RMSE) as described by Kinniburgh [

The surface area of the materials was measured usmultipoint BET method (Gemini 2360 surface area analyMicromeritics) with N2 as the adsorbate. Surface area msurements were performed in triplicate.

Organic-carbon content was measured by combustioning a Carlos Erba NA 2500 series elemental analyzer (T

e

sr

moQuest Italia S.p.A.). The unmodified base bentonitean organic-carbon content of 0.2%.

3. Results

Measured organic-carbon contents agree well withoretical values (e.g., values calculated assuming compexchange of all of the added organic cations) for both clayevery level of cation loading (Table 2). The natural organcarbon content of the unmodified bentonite was foundbe 0.2%.

Figure 1 shows a subset (for visual clarity) of the resof benzene sorption to BTEA bentonite. The isothermsdistinctly nonlinear and were fit to a Langmuir model of tform

(2)q = qmbCe

1+ bCe

,

whereq = amount of solute sorbed,b andqm are isothermparameters, andCe is the equilibrium aqueous solute cocentration. Equilibrium sorbed benzene concentrationsbe converted to sorbed molar concentrations by multiing by 1.28× 10−5. The isotherm parameters for benzesorption to BTEA bentonite are given in Table 3. Isotherfor TCE sorption to BTEA bentonite are also nonlinear aa subset (for visual clarity) of these isotherms is shownFig. 2. The isotherm data were fit to a Langmuir mo(Table 4). Equilibrium sorbed TCE concentrations mayconverted to sorbed molar concentrations by multiplying7.6× 10−6.

Sorption isotherms for 1,2-dichlorobenzene sorptionBTEA bentonite are given in Fig. 3. These data were anonlinear and fit with a Freundlich model of the form

(3)q = KC1/ne ,

Table 2Theoretical and actual organic-carbon content

Sorbent Theoretical organic-carbon Measured organic-carbcontent (%) content (%)

30 BTEA 3.1 3.140 BTEA 4.1 4.150 BTEA 5.1 5.060 BTEA 6.0 6.170 BTEA 7.0 7.380 BTEA 7.9 8.090 BTEA 8.8 8.7100 BTEA 9.6 8.730 HDTMA 4.5 4.640 HDTMA 5.9 5.850 HDTMA 7.2 7.660 HDTMA 8.5 8.970 HDTMA 9.8 9.880 HDTMA 11.0 11.090 HDTMA 12.2 12.2100 HDTMA 13.3 14.0

254 S.L. Bartelt-Hunt et al. / Journal of Colloid and Interface Science 266 (2003) 251–258

withrep-

and

and

on-ionsr

ven

ngedines

ex-con-

in

Fig. 1. Isotherms for the sorption of benzene to bentonite exchangedBTEA cations at 3.1, 5.0, 7.3, and 8.7% organic-carbon content. Linesresent Langmuir model fits.

Table 3Isotherm parameters for the sorption of benzene to BTEA bentoniteHDTMA bentonite with varying organic-carbon content

Sorbent qm (mg/kg) b (l/mg) R2 Sorbent Kd (kg/l) R2

30 BTEA 12,500 0.04 0.99 30 HDTMA 10.5 0.9940 BTEA 14,300 0.1167 0.99 40 HDTMA 8.27 0.9850 BTEA 14,300 0.0636 1.00 50 HDTMA 12.5 0.9860 BTEA 16,600 0.0429 0.99 60 HDTMA 16.5 0.9970 BTEA 12,500 0.0500 0.99 70 HDTMA 20.8 0.9880 BTEA 14,300 0.0318 0.99 80 HDTMA 24.3 0.9890 BTEA 11,100 0.0196 0.98 90 HDTMA 29.0 0.98100 BTEA 11,100 0.0321 0.99 100 HDTMA 31.6 0.99

Table 4Isotherm parameters for the sorption of TCE to BTEA bentoniteHDTMA bentonite with varying organic-carbon content

Sorbent qm (mg/kg) b (l/mg) R2 Sorbent Kd (kg/l) R2

30 BTEA 12,500 0.0308 0.98 30 HDTMA 2.39 0.9840 BTEA 12,500 0.0182 0.94 40 HDTMA 3.51 0.9950 BTEA 10,000 0.0054 0.89 50 HDTMA 5.99 0.9860 BTEA 11,100 0.0300 0.98 60 HDTMA 8.46 0.9970 BTEA 10,000 0.0119 0.99 70 HDTMA 11.0 0.9980 BTEA 10,000 0.0053 0.95 80 HDTMA 10.7 0.9890 BTEA 6,250 0.0167 0.96 90 HDTMA 13.7 0.98100 BTEA 10,000 0.0164 0.97 100 HDTMA 18.1 0.99

whereq andCe are as previously defined andK and 1/n

are fitted parameters. Equilibrium sorbed 1,2-DCB ccentration data may be converted to molar concentratby multiplying by 6.8 × 10−6. Isotherm parameters fo1,2-dichlorobenzene sorption to BTEA bentonite are giin Table 5.

Fig. 2. Isotherms for the sorption of trichloroethene to bentonite exchawith BTEA cations at 3.1, 5.0, 7.3, and 8.7% organic-carbon content. Lrepresent Langmuir model fits.

Fig. 3. Isotherms for the sorption of 1,2-dichlorobenzene to bentonitechanged with BTEA cations at 3.1, 5.0, 7.3, and 8.7% organic-carbontent. Lines represent Freundlich model fits.

Sorption of benzene to HDTMA bentonite is shownFig. 4. These data were fit to a linear model of the form

(4)Cs = KdCe,

S.L. Bartelt-Hunt et al. / Journal of Colloid and Interface Science 266 (2003) 251–258 255

Table 5Isotherm parameters for the sorption of 1,2-dichlorobenzene to BTEA bentonite and HDTMA bentonite with varying organic-carbon content

Sorbent K (mg/kg) 1/n R2 Sorbent K (mg/kg) 1/n R2

30 BTEA 429 1.10 0.97 30 HDTMA 11.4 1.38 0.9140 BTEA 545 0.981 0.99 40 HDTMA 37.7 1.19 0.8850 BTEA 474 1.09 0.97 50 HDTMA 60.2 1.25 0.9660 BTEA 624 0.921 0.98 60 HDTMA 52.2 1.17 0.9970 BTEA 475 0.906 0.98 70 HDTMA 140 1.09 0.9980 BTEA 141 1.23 0.98 80 HDTMA 170 1.16 0.9790 BTEA 273 0.870 0.99 90 HDTMA 293 1.01 0.98100 BTEA 330 0.893 0.99 100 HDTMA 524 0.875 0.98

withines

n toEa-tiontlyre-

ionofclay

gree

)ree

ch isfor

MAesent

alue2]-lue

tudy

. Fored as

Ato

sthatyle-cre-ityata

Fig. 4. Isotherms for sorption of benzene to bentonite exchangedHDTMA cations at 4.6, 7.6, 9.8, and 12.2% organic-carbon content. Lrepresent linear model fits.

whereCs = equilibrium sorbed concentration,Ce = equi-librium aqueous concentration, andKd = sorption distribu-tion coefficient. Isotherm parameters for benzene sorptioHDTMA bentonite are given in Table 3. Sorption of TCto HDTMA bentonite was also linear (Fig. 5). Isotherm prameters are given in Table 4. Isotherm data for sorpof 1,2-dichlorobenzene to HDTMA bentonite were slighconcave up (Fig. 6). These sorption data were fit with a Fundlich model with parameters given in Table 5. For sorptof all three solutes to HDTMA bentonite, the magnitudesorption increased as the organic-carbon content of theincreased.

The isotherm parameters measured in this study awell with values observed in previous studies. logKoc(whereKoc = Kd divided by the organic-carbon fractionvalues for HDTMA bentonite measured in this study agwell with those measured by Boyd et al. [8]. logKoc forbenzene at 7.6% organic-carbon content was 2.22, whicomparable with a value of 2.0 reported by Boyd et al. [8]HDTMA smectite at 7.1% organic carbon. logKoc for TCE

Fig. 5. Isotherms for sorption of TCE to bentonite exchanged with HDTcations at 4.6, 7.6, 9.8, and 12.2% organic-carbon content. Lines reprlinear model fits.

at 7.6% organic-carbon content was 1.9 compared to a vof 1.7 reported by Boyd et al. [8]. Gullick and Weber [2measured a logKoc value of 2.0 for TCE on HDTMA bentonite at an organic-carbon content of 14.2%. This vacorresponds closely to a value of 2.1 observed in this sat an organic-carbon content of 14.0%.

The surface area measurements are given in Table 6both clays, the measured surface area generally decreasthe organic-cation loading increased.

4. Discussion

The magnitude of sorption of all three solutes to HDTMbentonite increased with increased cation loading up100% of the clay’s CEC. All HDTMA bentonite isothermwere linear or slightly concave up, supporting the ideasorption to clays modified with cations with long alkchains such as HDTMA occurs primarily by partition btween the aqueous solution and the organic mediumated by the alkyl chains [6]. The slight upward concavobserved in some of the HDTMA bentonite isotherm d

256 S.L. Bartelt-Hunt et al. / Journal of Colloid and Interface Science 266 (2003) 251–258

withines

nite

en athe

then of

ute’sav-glysedthe

at at

MA

CBore

sedtes

ultandb-nvidezeneich

e andow-cedup-CE.ac-y

acesac-anarces,CE.ater

Fig. 6. Isotherms for sorption of 1,2-DCB to bentonite exchangedHDTMA cations at 4.6, 7.6, 9.8, and 12.2% organic-carbon content. Lrepresent Freundlich model fits.

Table 6Measurements of surface area for BTEA bentonite and HDTMA bento

Sorbent Surface area (m2/g) Standard deviation

30 BTEA 36.90 0.8940 BTEA 25.91 2.8250 BTEA 28.65 0.9660 BTEA 22.53 0.4870 BTEA 22.91 0.3580 BTEA 13.23 2.7090 BTEA 28.48 1.05100 BTEA 21.52 0.0730 HDTMA 10.88 0.3640 HDTMA 8.82 0.2150 HDTMA 10.28 1.9860 HDTMA 11.49 0.3370 HDTMA 5.59 0.2980 HDTMA 4.59 0.1590 HDTMA 3.63 0.27100 HDTMA 4.57 0.07

has been observed by previous researchers and has betributed to the formation of a discrete solute phase onsurface of the clay [12]. This explanation may explainslight upward concavity in the data presented for sorptio1,2-DCB to HDTMA bentonite.

Since the partition process is dependent on the solaqueous solubility, it is generally believed that solutes hing a high aqueous solubility will not be sorbed as stronas solutes with comparatively low aqueous solubility. Baon this idea, for the three solutes used in this study,magnitude of sorption would be expected to be 1,2-DCB�TCE> benzene. Results of the sorption tests indicate th

t-

Fig. 7. Sorption of benzene, 1,2-dichlorobenzene, and TCE to HDTbentonite at 7.6 and 14.0% organic-carbon content.

all levels of cation loading examined in this study, 1,2-Dwas sorbed the most strongly, but benzene sorbed mstrongly than TCE, a result that would not be expected baon the aqueous solubility of the solutes. Figure 7 illustrathis result for two different cation loadings. A similar reswas reported by Boyd et al. [8] in a study of benzeneTCE sorption onto HDTMA smectite. Boyd et al. [8] oserved a larger logKoc for benzene than for TCE for a giveorganic-carbon content, although the authors did not proan explanation for this result. Compounds such as benare partly polarizable due to electron delocalization, whmay cause an electrostatic attraction between benzenthe clay mineral surface that does not occur for TCE. Hever, we cannot offer an explanation for why this indudipole moment would have a stronger effect on solutetake than the permanent dipole moment exhibited by TAnother possible explanation for the lower sorptive capity of HDTMA bentonite for TCE than for benzene mabe that TCE is unable to access the interlamellar spof the HDTMA bentonite, and therefore does not havecess to the entire partition medium. Benzene, being a plmolecule, would be able to access the interlamellar sparesulting in a higher sorptive capacity for benzene than TIt is also possible that TCE may have a lower micelle–w

S.L. Bartelt-Hunt et al. / Journal of Colloid and Interface Science 266 (2003) 251–258 257

Efnd

aseterto

-d3]taco-

ackticen-than

ic-tenttheAt-ns.

-paconll as

clayis, theg innic

ar-orp-e ofrbonEC

rre-of

tent.ion,de-tionde-EAtivesed.

as40

tlyat ain-asedea of. Wefacellerp-

ofarlyeat-gan-iumthis

thanca-nts,ari-m avead-

EC.-

withn-rgerrgerith

entTheanu-

ramonlly

Fig. 8. logKoc as a function offoc for 1,2-DCB, benzene, and TCsorption to HDTMA bentonite, wherefoc = organic carbon fraction othe clay, Koc = organic carbon normalized distribution coefficient, aCe = equilibrium aqueous concentration.

partition coefficient than benzene when the micelle phis composed of HDTMA cations. In general, micelle–wapartition coefficients are thought to be linearly relatedoctanol-water partition coefficients(Kow). Based on this assumption, TCE, having a largerKow than benzene, shoulbe more soluble in HDTMA. Valsaraj and Thibodeaux [2found that while this relation usually holds, very few dahave been collected to determine micelle–water partitionefficients for aliphatic hydrocarbons such as TCE. The lof data on micelle–water partition coefficients for aliphahydrocarbons, combined with the partly polar nature of bzene, raises the possibility that benzene is more solubleTCE in a cationic surfactant phase such as HDTMA.

For all three solutes examined in this study, logKoc val-ues for HDTMA bentonite increase with increasing organcarbon content to a maximum at an organic-carbon concorresponding to a cation loading of about 100% ofclay’s CEC (Fig. 8). Because 1,2-DCB sorption to HDTMbentonite is nonlinear,Koc values were calculated and ploted for three different equilibrium aqueous concentratioThe observed increase inKoc with organic-cation loading indicates that for a given solute, the increase in sorption caity is not solely a function of the increased organic-carbcontent of the clay. Perhaps cation arrangement as wecation loading affects the sorption capacity of an organofor a particular solute. When more of the HDTMA cationloaded onto the internal and external surfaces of the clayalkyl chains may become more densely packed, resultina more effective partitioning medium for nonpolar orgasolutes.

Sorption of all three solutes to BTEA bentonite was chacterized by strong, nonlinear uptake, indicating that stion occurred by an adsorption process. The magnitudsorption for all three solutes increased as the organic-cacontent of the clay decreased from 100% of the clay’s C

-

to approximately 50%. At organic-carbon contents cosponding to below 50% of the clay’s CEC, the magnitudesorption decreased with decreasing organic-carbon conBecause sorption to BTEA bentonite occurs by adsorptit is likely these sorption results may be explained by thecrease in surface area observed with increasing BTEA caloading on the surface of the clay. As the surface areacreased, fewer sorptive sites on the surface of the BTbentonite were available, thus reducing the overall sorpcapacity of the clay as the cation loading was increaFor all three solutes, the maximum sorption capacity wachieved at a cation loading corresponding to betweenand 60% of the CEC of the BTEA bentonite. A significansmaller sorption capacity was observed for all solutesloading equivalent to 30% of the clay’s CEC. This resultdicates that there may be a trade-off between the increorganic-carbon content and the decreased surface arorganobentonites that sorb by an adsorption mechanismwould expect that at loadings less than 30%, the surarea of the clay would continue to increase, but the smaamount of organic carbon would effectively limit the sortive capacity of the clay.

The results of this study have implications for the useorganoclays in environmental applications, as cost is nealways a consideration in designing remediation and trment schemes. The commercial cost of synthesizing oroclays is affected by the amount of quaternary ammoncation exchanged onto the clay surface. The results ofstudy indicate that BTEA bentonites exchanged at less100% of the CEC of the clay result in larger sorptionpacities for several common nonpolar organic pollutapotentially reducing the cost of using organoclays in vous types of remediation schemes and thus making themore attractive option. Conversely, HDTMA bentonites haa higher sorption capacity when synthesized at cation loings equivalent to a higher percentage of the clay’s CBased on the logKoc values measured for HDTMA bentonite, it appears that the increase in sorption capacitycation loading may be nonlinear. As a result, HDTMA betonites exchanged at cation loadings equivalent to a lapercentage of the CEC of the clay may have a much lasorption capacity for nonionic organic contaminants wonly a small increase in cost.

Acknowledgments

This research was supported in part by the Departmof Defense Graduate Research Fellowship Program.authors thank Margaret Miller, Eric Anderson, and BryDick for their assistance in collecting laboratory data. Ssan Burns gratefully acknowledges support of the progdirector, Dr. Cliff Astill, at the National Science Foundati(Grant No. 9984206) through which this work was partiafunded.

258 S.L. Bartelt-Hunt et al. / Journal of Colloid and Interface Science 266 (2003) 251–258

I,

90)

2

9.001)

ch-

ys

erhers,

In-

.

ewis

References

[1] J.A. Smith, P.R. Jaffé, J. Environ. Eng. 120 (1994) 1559.[2] R.W. Gullick, Dissertation, University of Michigan, Ann Arbor, M

1998.[3] G. Sheng, S.A. Boyd, Clays Clay Miner. 48 (2000) 43.[4] J.-W. Park, P.R. Jaffé, Environ. Sci. Technol. 27 (1993) 2559.[5] M. Harper, C.J. Purnell, Environ. Sci. Technol. 24 (1990) 55.[6] J.A. Smith, P.R. Jaffé, C.T. Chiou, Environ. Sci. Technol. 24 (19

1167.[7] J.A. Smith, P.R. Jaffé, Water Air Soil Pollut. 72 (1994) 205.[8] S.A. Boyd, M.M. Mortland, C.T. Chiou, Soil Sci. Soc. Am. J. 5

(1988) 652.[9] J. Li, J.A. Smith, A. Winquist, Environ. Sci. Technol. 30 (1996) 308

[10] N.M. Soule, S.E. Burns, J. Geotechnol. Geoenviron. Eng. 127 (2363.

[11] J.J. Deitsch, J.A. Smith, M.B. Arnold, J. Bolus, Environ. Sci. Tenol. 32 (1998) 3169.

[12] W.F. Jaynes, S.A. Boyd, Soil Sci. Soc. Am. J. 55 (1991) 43.[13] W.F. Jaynes, G.F. Vance, Clays Clay Miner. 47 (1999) 358.[14] J.A. Smith, P.R. Jaffé, Environ. Sci. Technol. 25 (1991) 2054.[15] J.A. Smith, A. Galan, Environ. Sci. Technol. 29 (1995) 685.[16] J.-F. Lee, M.M. Mortland, C.T. Chiou, D.E. Kile, S.A. Boyd, Cla

Clay Miner. 38 (1990) 113.[17] G. Sheng, S.A. Boyd, Clays Clay Miner. 46 (1998) 10.[18] J.A. Smith, D.M. Tuck, P.R. Jaffé, R.T. Mueller, in: R.A. Bak

(Ed.), Organic Substances and Sediments in Water, Lewis PublisChelsea, MI, 1991, p. 392.

[19] Z. Li, R.S. Bowman, Environ. Sci. Technol. 32 (1998) 2287.[20] A.Z. Redding, S.E. Burns, R.T. Upson, E.F. Anderson, J. Colloid

terface Sci. 250 (2002) 261.[21] D.G. Kinniburgh, Environ. Sci. Technol. 20 (1986) 895.[22] R.W. Gullick, W.J.J. Weber, Environ. Sci. Technol. 35 (2001) 1523[23] K.T. Valsaraj, L.J. Thibodeaux, Sep. Sci. Technol. 25 (1990) 369.[24] J.H. Montgomery, Groundwater Chemicals Desk Reference, L

Publishers, Chelsea, MI, 1996.