Organic Pillared Clays

Organic Pillared Clays

Journal of Colloid and Interface Science 238, 24–32 (2001) doi:10.1006/jcis.2001.7498, available online at http://www.idealibrary.com on Organic Pill...

123KB Sizes 0 Downloads 16 Views

Journal of Colloid and Interface Science 238, 24–32 (2001) doi:10.1006/jcis.2001.7498, available online at http://www.idealibrary.com on

Organic Pillared Clays L. P. Meier,∗,1 R. Nueesch,† and F. T. Madsen‡ ∗ Nonmetallic Materials, Swiss Federal Institute of Technology, 8092 Zuerich, Switzerland; †Institute for Technical Chemistry, Section Water and Geotechnology, Forschungszentrum Karlsruhe GmbH, P.O. Box 3640, D-76021 Karlsruhe, Germany; ‡R¨utistrasse 5, CH-8134 Adliswil, Switzerland Received April 11, 2000; accepted February 27, 2001

the organic cations fill the interlayer space and the layers are held apart permanently by the substituents of the organic cation. The organophilic smectites are compared to another type of modified smectites, the pillared clays, where the cations are exchanged against inorganic polyoxocations, e.g., polyoxoaluminates. Therefore, the organic modified clays are contemplated as “organic pillared clays.” Resulting from the organic ion exchange the hydrophilic ions in the interlayers are replaced, and the smectite becomes generally hydrophobic. Organophilic smectites are used as an adsorbent of organic pollutants. It is proposed to regard the organophilic interlayer as an organic solvent (2–6). Generally, the adsorption properties of smectites modified by alkylammonium increase with the length of the alkylsubstituents. Different pollutants were adsorbed by modified smectites with respect to the chain length of the alkyl groups and the layer charge (7– 9). The adsorption properties of organophilic smectites against different pollutants such as benzene, toluene, p-xylene, ethylbenzene, o-dichlorobenzene, o-xylene, and Lindan on tetramethylammonium smectite were investigated. The limiting adsorption factor of such modified smectites was suggested to be the remaining interlayer volume generated by the interlayer cation (10). Different methods for treating waste water based on organic clays were proposed (5, 11–24). The layer charge density of smectites can be determined via the interlayer expansion depending on the chain length of different exchanged alkylammonium ions (25–27), because the interlayer distance using alkylammonium depends on the size of the alkyl chains. The interlayers distance also is influenced by the presence of organic solvents, e.g., higher alcohols. An endothermal Van der Waals expansion energy of organic modified smectites was derived from these results to be around 5 kJ per mol of CH2 units (28–30). Our investigations focused on the mechanism of the pollutant intercalation into organophilic smectites. We propose that there are two types of organic pillared clays. The first type is expanded by flexible organic cations, such as dimethyldioctadecyl ammonium, and is denoted as flexible cation organic pillared clays (FPCs). The other one is pillared by rigid cations such as tetraphenylphosphouium and is called the rigid cation organic pillared clays (RPCs). FPCs usually show good adsorption at relatively high concentrations of 2-chlorophenol, whereas RPCs adsorb much better at low concentrations. Since these two types

Commonly used organophilic clays are modified by alkylammonium cations which hold apart the aluminosilicate layers permanently. The cations fill the interlayer space and are contemplated as flexible pillars, resulting from the mobility of the alkyl chains. Therefore, the interlayer distance varies depending on the layer charge and on the alkyl chain length. Contrary to these cations, rigid pillaring cations guarantee a constant interlayer distance without occupying the interlayer by themselves and show special adsorption properties such as hydrophilic behavior contrary to the generally hydrophobic ones. Smectites were modified by flexible organic cations, e.g., dimethyldioctadecylammonium, and by rigid ones, e.g., tetraphenylphosphonium. Their adsorption properties are compared. Our investigations showed improved adsorption properties for rigid organic cations on smectites using 2-chlorophenol as pollutant. Best adsorption results are achieved using pillaring cations in combination with low charged smectites, especially at low pollutant concentrations. The properties of organic modified smectites are discussed by a pollution intercalation model. The intercalation process of an organic pollutant into an organic modified smectite is expressed by a two-step Born-Haber cycle process: (i) the formation of an adsorbing position by layer expansion and (ii) the occupation of the adsorbing position by the pollutant. The first step of the formation of the adsorbing position is an endothermal transition state which lowers the total intercalation energy and therefore worsens the adsorption behavior. Thus, an already expanded organophilic smectite will show improved adsorption behavior. The formed adsorbing position state on organic modified smectites is comparable to the pillared state of inorganic pillared clays. °C 2001 Academic Press Key Words: smectite; organic pillared clay; pollutant; adsorption; 2-chlorophenol.

INTRODUCTION

Organic pollutants are still a problem for drinking water and the environment. Organic modified smectites are used for the removal of organic pollutants from waste water because of their hydrophobicity and their high surface area. They have a high capacity (uptake up to 100% weight) but the adsorption affinity at low pollutant concentrations usually is low compared that of to activated charcoal. The preparation of such smectites is carried out by the exchange of the naturally occurring interlayer cations by organic cations (1). After the cation exchange 0021-9797/01 $35.00

C 2001 by Academic Press Copyright ° All rights of reproduction in any form reserved.

24

ORGANIC PILLARED CLAYS

show major differences in the pollutant adsorption properties as well as in BET surface and layer-to-layer distance determined by XRD, we investigated these two types. A model was developed that explains the differences in adsorption behavior. MATERIALS AND METHODS

Smectites The following clays were used (see Table 1): Volclay (natural sodium ion form) (31) was from the American Colloid Company. The <2-µm fractions were separated (Ref. 1, p. 187). Hectorite (natural calcium form) was a Clay Minerals Society Source Clay having known properties (32). Calcite was removed by a standard procedure (33). The sodium exchange was achieved by exchanging 2× overnight in 1 mol/L NaCl and subsequently rinsing 2× in bidistilled water and then ethanol until the conductivity was below 5 µS/cm. The <2-µm particle fractions were separated (Ref. 1, p. 187). Syntheses of the low charged smectite stevensite was performed by the hydrothermal transformation at 300◦ C of sepiolite. The XRD spectra of the ethylene glycol-treated sample showed a fully swellable smectite and no sepiolite signals were observed, as described in literature (Ref. 1, p. 199 (34)). Organic Cations Didodecyldimethylammonium bromide, purum (CAT. 1) dimethyldioctadecylammonium bromide, purum (CAT. 2), tetraphenylphosphonium chloride, purum p.a. (CAT. 7), phenyltrimethylammonium bromide, purum (CAT. 8), benzyltriethylammonium chloride, purum (CAT. 10), and methyltriphenylphosphonium bromide, purum (CAT. 11), were all obtained from Fluka. Methyltrioctylammonium bromide, purum (Fluka; CAT. 3), was recrystallized 1× in hexane (free of water traces!) at −20◦ C to remove the brown color. Bis(α,α 0 -o-xylylene)ammonium dibromide (CAT. 4) was derived as follows. A mixture of 10.5 g α,α 0 -dibromo-o-xylene, 105 mL of 25% ammonia, and 105 mL ethanol was stirred in a glass bottle at 105◦ C overnight. Solvents were evaporated and the residue was dissolved in 50 mL of hot water. The cool mixture was filtered and recrystallized from ethyl acetate/ethanol. TABLE 1 Properties of the Purified Smectites (<2-µm Fractions)

Volclay (< 2-µm fraction) Hectorite (< 2-µm fraction) Stevensite (< 2-µm fraction)

CEC (mol/kg)

Total layer charge per half unit cell (calculated)

Interlayer cations (after exchange)

0.92

0.33

Na+

0.75

0.28

Mainly Na+

0.60

0.23

Mainly Na+

25

Yield: 1.12g = 19%; mp >300◦ C; Anal. Found: C, 63.3; H, 5.4; N, 4.6 (Calcd: C, 63.6; H, 5.3; N, 4.6). Methyl-tris-(2-phenylethyl)-ammonium bromide (CAT. 5) was derived as follows. Methylammonium chloride (2.50 g) (purum, Fluka) was dispersed in 200 mL nitromethane and after addition of 23.30 g 2-phenylethylbromid (purum, Fluka; 3.4 mol/mol) and 51.17 g of dry potassium carbonate (10 mol/mol) 48 h at 80◦ C. The cooled mixture was centrifuged and decanted to remove the potassium salt. The potassium salt was washed 2× with 60 mL of methylenchloride, centrifuged, and decanted. The unified organic phases were dried and the solvent was evaporated. It was recrystallized from 100 mL of ethyl acid ethylester/isopropanol 1/1. Yield: 5.1 g (32%); Mp 158–159◦ C [H-NMR (CDCl3 ): d = 3.1 (6H, m, HC–C–N), 3.9 (6H, m, HC–N), 3.6 (6H, m, HC–C–N), 3.45 (4H, s, H3 C–N), 7.4 (5H, m, H–Ar)]. Bis(2,20 -dimethylen-biphenylylen) ammonium bromide (CAT. 6) was derived as follows. Diphenyl-2,20 -dicarbonic acid (“diphenic acid”) was synthesized by coupling reaction of diazotized anthranilic acid after Ref. 35 (p. 197). It was then esterified to diphenyl-2,20 -dicarbonic acid dimethylester (“diphenic acid dimethylester”) (Ref. 35, p. 112). It was reduced to the diol 2,20 -bis(hydroxymethyl)–biphenyl (Ref. 35, p. 51). The hydroxy groups were exchanged to give 2,20 -bis (brommethyl)biphenyl (Ref. 35, p. 42). The cation bis(2,20 dimethylen-biphenylylen)ammonium bromide was achieved by direct alkylation of ammonia (36). Phenyltriethylammonium iodide (CAT. 9) was derived as follows. N,N-Diethylanilin (5 g), 3.5 g ethyliodide, and 15 mL nitromethane were stirred for 30 min at RT, 30 min at 50◦ C, and 1 h at 70◦ C. The solvent was evaporated, the crystals were filtered off, and were recrystallized from ethyl acetate/isopropanol (1/1). Yield: 5.6 g = 71% [H-NMR (CDCl3 ): d = 1.25 (6H, t, H3 C–C), 4.15 (6H, q, HC–N), 7.5–7.7 (3H, m, m/p-HAr), 8.0 (2H, d, o-HAr)]. Cation Exchange Procedure for the Organic Modification of the Smectites The cation exchange procedure is easily performed but excess salt has to be washed out accurately. The quality of the UV measurements is limited by the unsedimented finest clay fraction, probably due to traces of remaining organic salts. It is also important that the pH of the wash solution and the initial pH of the organic clay are in the range of 5–6, where the clay edges become uncharged (point of zero charge (PZO)). One gram of <2-µm smectite fraction (sodium-form) was dispersed in 80 mL 50% methanol in water by ultrasonic treatment for 5 min (200 W, Model Dr. Hielscher, 7-mm titanium probe) and 50% methanol was added to have a total volume of about 300 mL. A 1.2× molar amount of the cation exchange capacity was calculated for the organic cation and was added quickly in 50 mL 50% methanol under intensive mixing (IKA Ultra-Turrax blender). The dispersion flocculated immediately and was intensively mixed for an additional 15 min and then

26

MEIER, NUEESCH, AND MADSEN

slowly stirred overnight at 60◦ C under this condition the organic salts are much more soluble and are expected to reach the whole surface. The mixture was centrifuged and the supernatant solution was removed. The organic modified clay was washed out subsequently with 400 mL water by centrifuging and removing the supernatant solution. The pH was kept at 5.5 to 6 (bubbling CO2 through the mixture). The wash procedure was repeated until the electric conductivity was <5 µS/cm. The organophilic clay then was washed out 2× in ethanol (Fluka, purum). The Adsorption of Chlorinated Phenols by Organophilic Clays Was Determined by UV-Spectroscopy Different concentrations of 2-chlorophenol (purum, >98%, Fluka) were prepared (20/30/50/125/150/300/600/1000) using a buffer of 1 mM malonic acid adjusted to pH 5.7, where any influence of deprotonated form 2-chlorophenolate can be neglected (solubility in water, 2.7 wt% RT; pK s , 8.32 (37)). Thirty milligrams of the modified clay (organic cation on Hectorite, Volclay, and Stevensite, respectively) was transferred into a glass tube with a Teflon stopper. It is noteworthy that polypropylene tubes would adsorb a small amount of the 2chlorophenol, which influences the UV result. The testing concentrations were added to the tube. One 7-mm glass globe was added and the closed test tubes were moved overnight in a turnover mixer. The modified clay was centrifuged for 1 h at 2000G. The 2-chlorophenol concentration was determined in the supernatant solution at λ = 274 nm (εmax = 1980; Dr. Lange UV/vis spectrometer) except for the sample containing 20 ppm 2-chlorophenol (for treatment, see below) (Fig. 1). For the concentration determination of 125 ppm and below a 10-mm quartz cuvette was used. The concentrations

TABLE 2 The Calibration of the 2-Chlorophenol Extinction at λ = 274 nm

a

ppm 2-Chlorophenol

Extinction (274 nm)

20 30 50 125 300 1000

0.282 0.431 0.713 1.78a 4.39a 14.12a

Measured extinction in 1-mm cuvette recalculated for 10-mm cuvette.

of 150 ppm and higher were determined in a 1-mm quartz cuvette. Especially at low remaining 2-chlorophenol concentration (<1 ppm), the extinctions were significantly worsened by the background of the finest suspended clay particles. The extinction values at 20 ppm were determined on the deprotonated form: 3 mL of the centrifuged test solution was added to 1 mL of 2% sodium carbonate, mixed, and the extinction determined at λ = 294 nm (2-chlorophenolate: λmax = 294 nm, εmax = 3780 (38)) (Fig. 1). The values were determined by the corrected extinction ε(corr) after comparison of the values at ε(264 nm) and ε(324 nm) (see also Table 2): ε(corr) = ε(294 nm) −

ε(324 nm) + ε(264 nm) . 2

[1]

Results of the 2-Chlorophenol Adsorptions (Fig. 2) The following tables list the results of 2-chlorophenol adsorptions for cations 1–11.

FIG. 1. Extinction of 2-chlorophenol and 2-chlorophenolate at 20 ppm concentration depending on the wavelength. The maximum extinction of the anion is 50% higher and shifted to 294 nm, where the measurement is less sensitive to dust in the solution.

27

ORGANIC PILLARED CLAYS

FIG. 2. The tested cations.

Methyl-tris-(2-phenylethyl)-ammonium (CAT. 5)

Didodecyltrimethylammonium (CAT. 1) ppm 2-Cp

on Volclay

ppm 2-Cp

on Hectorite

20 30 50 125 150 1000

6.9 9.8 16.8 53 68 777

20 30 50 125 300

7.8 16.2 32 93 243

Bis-N ,N -(2, 20 -dimethylenbiphenylylen) ammonium (CAT. 6)

Dimethyldioctadecylammonium (CAT. 2) ppm 2-Cp

on Volclay

20 50 125 150 1000

3.5 10.1 34.5 45 725

ppm 2-Cp

on Hectorite

20 30 50 125 300 1000

10.6 17.2 32.4 100 255 893

Tetraphenylphosphonium (CAT. 7)

Methyltrioctylammonium (CAT. 3) ppm 2-Cp

on Volclay

on Hectorite

ppm 2-Cp

on Volclay

on Hectorite

on Stevensite

20 30 50 125 150 300 1000

4.7 7.5 15.3 59 78

6.7 11.3 21.7 76

20.000 50.000 125.00 150.00 300.00 1000.0

2.7 14.5 63 81

2.5 10.6 54.6

0.6 5.8 41.5 56.6

207 812

808

Bis-(α,α 0 -xylylen)-ammonium (CAT. 4) ppm 2-Cp

on Hectorite

20 30 50 300 1000

18.4 28.2 46.8 284 928

854

192 813

Phenyltrimethylammonium (CAT. 8) ppm 2-Cp

on Volclay

20 30 50 125 150 1000

12.1 17.6 30 76 91 796

817

28

MEIER, NUEESCH, AND MADSEN

Recalculation to Compare with Literature Data

Phenyltriethylammonium (CAT. 9) ppm 2-Cp

on Volclay

20 30 125 150 1000

10.3 15.5 84 104 910

Other graphs in literature (7) use n adsorbed depending on m clay csolution (n adsorbed , amount of adsorbed molecules; V , solutions volume; m clay , the mass of adsorbent). These plots can be calculated from the figures: n adsorbed V · cadsorbed V · (c0 − csolution ) = = m clay m clay m clay

N ,N ,N -Triethylbenzylammonium (CAT. 10) ppm 2-Cp

on Volclay

20 30 50 125 150 1000

11 16.1 28.8 83 103 908

=

on Hectorite

20 30 50 125 310

13 20.8 37.8 101 265

K =

[5]

Basal Distances of the Organic Modified Smectites Determined by XRD and Organic Content Determined by Thermogravimetrical Analysis (TGA)

The adsorption behavior of the different modified smectites is interesting over a large concentration range at low concentrations (affinity of the organosmectites). Langmuir derived his adsorption isotherm from the equilibrium constant K , explaining an adsorption as an adsorption equilibrium reaction. This constant is used in our plots. The advantage of this plot is the adsorption prediction in the low concentration regime:

[2-chlorophenol]adsorbed cadsorbed = [2-chlorophenol]solution csolution c0 − csolution c0 = = − 1. csolution csolution

n adsorbed . V · csolution

RESULTS AND DISCUSSION

Calculation of the Adsorption Equilibrium K and the Adsorption Isotherms Graphs

2-chlorophenoladsorbed ↔ 2-chlorophenolsolution

[4]

K can be calculated after the following equation:

Methyltriphenylphosphonium (CAT. 11) ppm 2-Cp

V · K · csolution . m clay

[2]

K =

[3]

The K values are calculated by determination of the equilibrium concentration of 2-chlorophenol in the supernatant solution csolution after the adsorption time and c0 , which is the initially added 2-chlorophenol concentration. Using log/log plots of K depending on the equilibrium concentration csolution made it possible to predict the adsorption behavior over a large range of concentrations, e.g., at low concentrations. Considering the organic phase volume the equilibrium constant K easily let compare the efficiency of the organophilic clays with phase partitioning.

Comparing the two types of organophilic clays, we expected different layer-to-layer expansion properties. Regarding the commonly used alkyl-type modified clays, the FPCs, the interlayer distance varies depending on the layer charge and the size of the alkyl chain of the alkylammonium ions, which makes this suitable for the layer charge determination. The expansion is possible due to the nonfixed conformation of the chains in the interlayer. It also swells in the presence of other molecules such as alcohols and phenols. For the other type, RPCs, the TABLE 3 Layer-to-Layer Distance and Organic Content of the Different Modified Smectites

Cation Volclay CAT. CAT. CAT. CAT. CAT. CAT. CAT. CAT. CAT. CAT. CAT.

1 2 3 4 5 6 7 8 9 10 11

Stevensite CAT. 2 CAT. 7

Clay

Organo-cation

Organic XRD content ˚ (A) (TGA)(%)

Didodecyldimethylammonium Dimethyldioctadecylammonium Methyltrioctylmethylammonium Bis-(α, α 0 -xylylene)-ammonium Methyl-tris-(2-phenylethyl)-ammonium Bis-(2, 20 -dimethylenbiphenylylen)ammonium Tetraphenylphosphonium Phenyltrimethylammonium Phenyltriethylammonium N ,N ,N -Triethyl-benzylammonium Methyltriphenylphosphonium

18.0 24.4 17.6 14.4 20.5 17.7

Dimethyldioctadecylammonium Tetraphenylphosphonium

18.6 18.4

18.4 14.4 14.9 14.6 15.6

34.1 22.3

14.2 11.4 13.7

14.6

ORGANIC PILLARED CLAYS

TABLE 4 The BET Surface of the Standard Type Organic Pillared Clays RPCs and FPCs Organic pillared clay

BET surface (m2 /g clay)

CAT. 7/Volclay CAT. 7/Stevensite CAT. 2/Volclay CAT. 2/Stevensite

131/131 208/209 Not defineda Not defined

RPC RPC FPC FPC

Note. The surface of the RPC was measured after heating out at 100 and 170◦ C and is the average of two measurements. a About 10 m2 /g after heating out at 100◦ C overnight (may differ depending on the heating out temperature).

dimensions of the cation are given and determine the interlayer expansion. The different layer to layer distances were determined by XRD. The distance of a completely dried and not expanded ˚ (1). The additional distance is a result sodium smectite is 10 A of the exchanged organic cations. The XRD results clearly show that interlayer distances of FPCs are not fixed. The layer-to-layer distance of CAT. 2 as the standard cation for FPCs on the higher charged Volclay is ˚ compared to 18.6 A ˚ on the lower charged Stevensite 24.4 A which corresponds to the layer charge of 0.92 mmol/g clay for Volclay and 0.6 mmol/g clay for the low charged Stevensite. On the other hand, CAT. 7, as the standard cation for RPCs, ˚ expands the interlayer to give layer to layer distance of 18.4 A independent on the smectite. This indicates that RPCs do not fill the interlayer space entirely and that the interlayers have porosity. Since pollutant molecules could occupy such spaces easily, they are of major interest. The molecules of lower symmetry usually show the lowest expansion of the clay layers, which is given by their lowest ˚ It is assumed that the dimension. CAT. 11 expands to 15.6 A. shortest dimension of any molecule determines the interlayer spacing.

29

Influence of the Layer Charge of the Smectite The amount of pillaring cations in the interlayer depends on the layer charge and increases with increasing smectite layer charge. To determine the interlayer porosity influence on the adsorption (affinity and capacity), we used different charged smectites, expecting an enhanced pollutant adsorption by RPCs on the low-charged layers. For this purpose a cation having a comparable organic amount (CAT. 3: 3× octyl-substituent) was chosen. Any effect connected to the organic volume in the interlayer will thus be in the same range. The crucial difference of the pollutant intercalation mechanism of FPCs and RPCs indeed effects observeable differences. Better properties of FPCs are achieved on higher charge densities. The lower charge density of Hectorite compared to Volclay lowers K for CAT. 3 (FPC) by about 40%. For RPCs lowering the charge and therefore the carbon content increases their adsorptivity K by a factor of 5 on Hectorite (see Fig. 3). Comparing the best systems at 20 ppm initial 2chlorophenol concentration, CAT. 7/Stevensite has a almost 10 times higher affinity than CAT. 3/Volclay. Thus, concerning the inorganic clay layers the following rule is valid: the lower charged the smectite the better it is suited for a good adsorbing RPC. The modified low-charged smectite Stevensite shows much better results than Hectorite and Volclay with higher CEC. Although the cations provide interlayer space, they also consume it by their own volume. It is expected that the best suited smectites for RPC have a very low charge, where the volume loss through the cations can be neglected. Unfortunately, the swellability of any smectites is related to the layer charge (41). Below a certain value of layer charge only parts of the

Determination of the BET Surface To determine the porosity the BET surface of the clays was measured. It was expected that RPCs of the lower charged stevensite have a higher surface area than RPC/Volclay. The surface area of the swellable smectite is 798 m2 /g (39) and the BET surface is 40 m2 /g. The surface area of the RPCs depends on the charge of the smectite. The lower the charge the higher the surface area. The results imply that RPCs indeed have porosity. On the other hand, FPCs have no defined BET surface. It varies depending on the heating out temperature (the modified smectite may still contain traces of water (40)). It is also possible that the alkyl chains in the interlayer become mobile due to the increased temperature and any accessible surface gets lost. However, RPCs have a much higher surface area which does not depend on the heating out temperature.

FIG. 3. Dependence of the equilibrium constant K on the layer charge and therefore on organic content of FPCs and RPCs.

30

MEIER, NUEESCH, AND MADSEN

FIG. 4. Adsorption isotherm of 2-chlorophenol by FPCs (CAT. 3, 5) and RPCs (CAT. 4, 6, 7, 11) on Hectorite.

In general the clays modified by small organic cations (CAT. 8, CAT. 9, CAT. 10) show an approximation of the equilibrium constant toward a constant value as described by the Langmuir adsorption equation. The limited adsorption volume is given by the small pillar˚ (Fig. 5). Thus, at ing cation (resulting interlayer space <5A) higher 2-chlorophenol concentrations CAT. 8 has a higher K than CAT. 9 or CAT. 10. At lower concentrations the ethylated ammonium derivates have higher affinity, probably due to their higher hydrophobicity, but they become saturated earlier as a consequence of the larger substituents and the lower adsorption space. CAT. 2, used for commercial organophilic clays, shows the much better properties of long-chain cations. A comparison of FPCs modified with different cations (CAT. 1–CAT. 3) to RPC CAT. 7/Volclay shows higher adsorption capacity constants (high concentration) for FPCs but lower affinity (low concentration) (Fig. 6). FPCs generally show Langmuir isotherms. For FPCs the carbon content is connected to K and, thus, CAT. 2 is the best adsorbing cation of the FPC type. The isotherm of CAT. 7/Volclay shows no Langmuir-type leveling for the constant K at low concentration and has higher affinity than CAT. 2/Volclay. Adsorption of Pollutants as a Two-Step Reaction

interlayers are swellable and such organophilic clays cannot be produced. Adsorptionisotherms of Different Cations (FPC versus RPC) (see Fig. 4) Different cations on Hectorite were tested to observe the effects of the pillaring cations on the 2-chlorophenol adsorption. The smectite was hectorite with a CEC of 75 mmol/100 g. The FPCs show higher adsorption capacity than RPCs. In general RPCs show that the adsorption constant strongly depends on the adsorption volume achieved by the cation. The characteristic volume of CAT. 4 is big in comparison to the small expansion ˚ Nonflexible cations do not adsorb better if they can(14.4 A). not produce adsorbing volume by expansion (CAT. 11: basal ˚ The expansion is given by the smallest dimendistance 15.6 A). sion of the molecule. The flexibility of seven-membered rings lets CAT. 6 not expand the interlayer to a fixed spacing and thus it is comparable to flexible substituents. Smectite modified by CAT. 6 has not better adsorption behavior then FPC. Tetraphenylphosphonium CAT. 7/Volclay, when compared with cations having the same carbon content (CAT. 3 and CAT. 1), adsorbs better, especially at low pollutant concentrations (4× better). The adsorption properties of organophilic clays depend not only on the type (RPC or FPC), but on the expansion and generated adsorption volume as well. Other cations like trimethylphenylammonium (CAT. 8), triethylphenylammonium (CAT. 9), dimethyldioctadecylammonium (CAT. 2), and triethylbenzylammonium (CAT. 10) were tested as a comparison (7, 9, 20).

We regard an adsorption of a pollutant into an organic pillared clay as a two-step reaction. The adsorption is split up in a endothermal and an exothermal partial reaction (Born-Haber cycle process). During the reaction the two types of clay react different as can be demonstrated.

FIG. 5. Adsorption isotherm of 2-chlorophenol by CAT. 2, CAT. 8, CAT. 9, and CAT. 10 on Volclay.

ORGANIC PILLARED CLAYS

31

FIG. 8. RPC. The water has to be removed from the porous interlayer. This is considered to be less endothermal and the total energy balance is more favorable.

cations does not change during the uptake. The second step is the exothermal intercalation of the pollutant and the driving force.

FIG. 6. Adsorption isotherm of 2-chlorophenol by different FPCs and RPCs on Volclay.

Model. During the adsorption, of molecules into organic pillared clays the uptake reaction passes a transition state. To reach this state is the endothermal first part of the reaction. At this activated first state a pore volume is generated, into which the molecule is intercalated in the second step, which is exothermal and happens spontaneously. Whereas the first step is independent of the adsorbing molecule, the second step is strongly influenced by the hydrophobicity of the molecule and is assumed not to vary much for FPCs and RPCs. Thus, for a given molecule the total adsorption energy balance is mainly influenced by the first state. For FPCs. Since FPCs do not have porosity, the absorptive volume for intercalating a pollutant molecule has to be provided by increased interlayer distance (Fig. 7). The flexible alkyl substituents change their position and conformation to reach the activated state, which is a transitional state and is not observable. The formation of the adsorbing site is distinctively endothermal. In the second step the pollutant will fill this space. The conformation and arrangement of the alkyl-chains of the interlayer

FIG. 7. FPC. The intercalation of a molecule into an organophilic clay modified by a FPC. The first step is endothermal.

For RPCs. The organic pillared clay dispersed in water already has structural porosity. On the other hand the pores of RPC are not really empty. If adsorbing pollutants from water, the pores will contain competing water as well. To reach the transitional comparable state, the water molecules have to be removed from the interlayer. The pollutant then can be intercalated into the empty sites (Fig. 8). The higher equilibrium constant K of RPCs is explained by following energetical arguments: For FPCs, the arrangement of the organic alkyl substituents in the interlayer is not fixed and many conformations are possible. Thus, they have a conformation of low energy excluding water since they are hydrophobic. They form a certain structure (pseudocrystalline structure) (42). For the formation of an adsorbing site this structure has to be broken up. When trying to follow the adsorption process with RPCs, the endothermal expansion step is not necessary since RPCs already have their porous volume. They actually represent the activated transition state of FPCs, differing just by having the pores filled with water. The energy of this first step (removal of the interlayer water) for RPC is related to the binding energy of the water in the interlayer (cations and silicate layer surface), which is assumed to be lower for the quite hydrophobic organic cations in RPCs than changing the conformation and expanding the layers in the case of FPCs. The activated state of FPCs is reached less easily than for RPCs and thus, the total adsorption energy is less exothermal. Since the second step (intercalation) depends on the phase transfer of the pollutant (2-chlorophenol) from the aqueous to the organophilic interlayer phase, the energy is not expected to differ much between FPCs and RPCs. Therefore, RPCs are expected to adsorb better. As a consequence of this the adsorption properties of RPC depend on the energy for the provision of the pore volume. The amount of pore volume in the interlayer space is higher using low charged smectites. Additionally, a low layer charge of the smectite also decreases the hydrophilic properties of the interlayers (41). Both effects increase the adsorption using lowcharged smectites.

32

MEIER, NUEESCH, AND MADSEN

Contrary to the FPCs, the macroscopic behavior of RPCs is not hydrophopic. RPCs are wetted by water but remain flocculated. We assume this to be due to the porosity of the interlayer; water is taken up and their macroscopic behavior is not hydrophobic. SUMMARY

The new type of organic pillared clays, the rigid cation organic pillared clay (RPC), was tested with clays having different layer charge and their fundamental adsorption properties were investigated. They are compared to the commercially used type pillared by organic flexible cations (FPCs). After modification of the clay using pillaring cations, the RPCs show increased BET surface area compared to FPCs. RPCs also show improved adsorption behavior in comparison to any organic modified clay investigated in literature. The adsorption experiments with different RPC systems showed the factors involved in the adsorption. RPCs have improved adsorption qualities especially at low pollutant concentrations up to more than 10 times better than the commonly used flexible cations organic clay. The adsorption properties of the RPCs are promising for the preparation of drinking water because they adsorb especially well at low pollutant concentrations. Further investigations on RPC will try to further improve the adsorption properties by increasing the porous volume and decreasing the cations size. ACKNOWLEDGMENTS We thank Dr. Stelian Grigoras (Dow Chemicals) for the fruitful discussions. We are truly indebted to Prof. L. J. Gauckler (Institute of Ceramics, ETH Z¨urich) and Prof. U. W. Suter (Institute of Polymers, ETH Z¨urich) for providing the support for this work. This study has received financial support from the KTI Project, Kommission of Technology and Innovation and Sulzer Chemtech, Switzerland.

REFERENCES 1. Moore, D., and Reynolds, R., “X-Ray Diffraction and the Identification and Analysis of Clay Minerals.” Oxford University Press, Oxford, 1997. 2. Boyd, S., Mortland, M., and Chiou, C., Soil Sci. Soc. Am. J. 52 , 652 (1988). 3. Smith, J., Jaff´e, P., and Chiou, C., Environ. Sci .Technol. 24, 1167 (1990). 4. Smith, J. A., Tuck, D. M., Jaff´e, P. R., Mueller, R. T., in “Organic Substances and Sediments in Water” (R. Baker, Ed.), p. 201. Lewis, Chelsea, MI, 1991. 5. Boyd, S., “Method of Removing Hydrocarbon Contaminants from Air and Water with Organophilic, Quaternary Ammonium Ion-Exchanged Smectite Clay,” EPO 474386B1 (1992). 6. Smith, J., and Jaff´e, P., J. Environ. Eng. 120, 1559 (1994). 7. Mortland, M., Shaobai, S., and Boyd, S., Clays Clay Miner. 34, 581 (1986). 8. Smith, J., and Galan, A., Environ. Sci .Technol. 29, 685 (1995). 9. Lee, J., Mortland, M., et al., Clays Clay Miner. 38, 113 (1990). 10. Lee, J., Mortland, M., Boyd, S., and Chiou, C., J. Chem. Soc. Faraday Trans. I 85, 2953 (1989).

11. Weiss, A., Appl. Clay Sci. 4, 193 (1989). 12. Park, J.-W., and Jaff´e, P., Environ. Sci. Technol. 27, 2559 (1993). 13. Holsen, T., Taylor, E., Seo, Y.-C., and Anderson, P., Eviron. Sci. Technol. 25, 1585 (1991). 14. Beall, G., “Process for Treating Organics Contaminated Water,” US 4517094 (1985). 15. Meo Ill, D., “Vapor-Extraction System for Removing Hydrocarbons from Soil,” US 5360067 (1994). 16. Deans, J., Heinsohn, G., and Link, B., “Removal of Organic Contaminants from Aqueous Media,” US 5393435 (1995). 17. Bolsin, G., “Method of Immobilizing Contaminants in the Soil or in Materials Similar to Soil.” UF 5413616 (1995). 18. Hughes, J., “Method of Improving Absorption of Salt Water by Water Swellable Clay by Redrying.” US 5427990 (1995). 19. Hooykaas, C., and Newton, J., “Fixant for Mixed Organic and Inorganic Contaminated Materials and Method for Using the Same,” US 5430235 (1995). 20. Katjia, L., “Method of Improving the Contaminant Resistance of a Smectite Clay by Rewetting and Impregnating the Clay with a Water-Soluble Polymer, and Redrying the Polymer-Impregnated Clay,” US 5556547 (1996). 21. Beall, G. A., “Method of Removing Water-Insoluble Organic Contaminants from an Acidic Aqueous Stream,” US 5567318 (1996). 22. Katjia, L., “Method of Improving the Contaminant Resistance of a Smectite Clay by Rewetting and Impregnating the Clay with a Water-Soluble Polymer, and Redrying the Polymer-Impregnated Clay,” US 5578219 (1996). 23. Boyd, S., “Method of Removing Hydrocarbon Contaminants from Air and Water with Organophilic, Quaternary Phosphonium Ion-Exchanged Smectite Clay,” US 5635075 (1997). 24. Kruse, K., “Process for Immobilizing Organic and Inorganic Pollutants in a Contaminated Soil Material on a Remediation Site,” US 5651831 (1997). 25. Lagaly, G., Fernandez-Gonzales, M., and Weiss, A., Clay Miner. 11, 173 (1976). 26. Lagaly, G., Clay Miner. 16, 1 (1981). 27. Lagaly, G., Identifizierung und Charakterisierung von Tonmineralien, Erkennung und Identifizierung von Tonmineralen mit organischen Stoffen. in “DTTG Jahrestagung” (H. Tributh and G. Lagaly, Eds.), pp. 86–130. Giessen, 1991. 28. Lagaly, G., and Malberg, R., Colloids Surf. 49, 11 (1990). 29. Weiss, A., and Lagaly, G., Z. Naturforsch. 24b, 1057 (1969). 30. Stul, M., and De Bock, J., Clays Clay Miner. 33, 351 (1985). 31. Schultz, L., Clays Clay Miner. 17, 115–149 (1969). 32. Van Olphen, H., and Fripiat, J. (Eds.) “Data Handbook for Materials and Other Non-metallic Minerals.” Pergamon, Oxford, 1979. 33. Jackson, M., “Soil Chemical Analysis—Advanced Course.” M. Jackson, Madison, WI, 1969. 34. G¨uven, N., and Carney, L., Clays Clay Miner. 27, 253 (1979). 35. Tierze, L.-F., and Eicher, T., “Reaktionen und Synthesen im organischchemischen Praktikum.” Georg Thieme Verlag. Stuttgart, 1981. 36. Wittig, G., K¨onig, G., and Clauss, K., Liebig’s Ann. Chem. 599, 127. 37. Beilstein EII, 6, 170. 38. Beilstein EIV, 6, 783. 39. Madsen, F. T., Thermochim. Acta 21, 89 (1979). 40. Engelhardt, T., pers. commun., Suedchemie. 41. Meier, L. P., and Nueesch, R., J. Colloid Interface Sci. 217, 77 (1999). 42. Israelachvili, J., “Intermolecular and Surface Forces.” Academic Press, London, 1992.