Mechanism of Smectic Arrangement of Montmorillonite and Bentonite

Mechanism of Smectic Arrangement of Montmorillonite and Bentonite Clay Platelets Incorporated in Gels of Poly(Acrylamide) Induced by the Interaction w...
0 downloads 0 Views 277KB Size
Langmuir 2006, 22, 369-374

369

Mechanism of Smectic Arrangement of Montmorillonite and Bentonite Clay Platelets Incorporated in Gels of Poly(Acrylamide) Induced by the Interaction with Cationic Surfactants S. G. Starodoubtsev,*,† E. K. Lavrentyeva,† A. R. Khokhlov,† G. Allegra,‡ A. Famulari,‡ and S. V. Meille‡ Physics Department, Moscow State UniVersity, Leninsky Gory, Moscow 119992, Russia, and Department of Chemistry, Materials and Chemical Engineering, Polytechnic of Milan, “G. Natta” Via L. Mancinelli 7, Milan, I-20131, Italy ReceiVed March 4, 2005. In Final Form: October 18, 2005 Structure transitions, induced by the interaction with the cationic surfactant cetylpyridinium chloride in nanocomposite gels of poly(acrylamide) with incorporated suspensions of the two closely related layered clays bentonite and montmorillonite, were studied. Unexpectedly, different behaviors were revealed. X-ray diffraction measurements confirm that, due to the interaction with the surfactant, initially disordered bentonite platelets arrange into highly ordered structures incorporating alternating clay platelets and surfactant bilayers. The formation of these smectic structures also in the cross-linked polymer gels, upon addition of the surfactant, is explained by the existence of preformed, poorly ordered aggregates of the clay platelets in the suspensions before the gel formation. In the case of montmorillonite, smectic ordering of the disordered platelets in the presence of the surfactant is observed only after drying the suspensions and the clay-gel composites. Rheology studies of aqueous suspensions of the two clays, in the absence of both surfactant and gel, evidence a much higher viscosity for bentonite than for montmorillonite, suggesting smaller clay-aggregate size in the latter case. Qualitatively consistent results are obtained from optical micrographs.

Introduction Cross-linked elastomers with colloidal fillers are widely used in different areas of modern industry. Swollen gels with incorporated dispersions of clays are a very specific class of such composite materials.1-5 Typically, the flat platelets of layered clays such as montmorillonite (MONT) have large lateral size in comparison with the mesh size of cross-linked, swollen polymer networks. Therefore, they cannot diffuse within the swollen polymer network. Due to this, the network prevents the flocculation of the clay particles, which retain their ability to participate in ion-exchange reactions and to absorb different substances. Among clays, various types of MONT and closely related minerals are the most important and widely used absorbents for organic compounds6-9 also because of their ability to swell remarkably in water. Due to this swelling behavior, MONT has a large active surface area (700-800 m2/g).7 Pauling gave fundamental contributions to demonstrate the layered structure of clay minerals.10 In particular, the layers of * To whom the correspondence should be addressed. † Moscow State University. ‡ Polytechnic of Milan. (1) Gao, D. Preparation and Property ImproVements of a Superabsorbent Polymer Composite; Alberta Research Council: Edmonton, Canada, 1992; Chapter 1. (2) Gao, D.; Heimann, R. B. Polym. Gels Networks 1993, 1, 225. (3) Churochkina, N. A.; Starodoubtsev, S. G.; Khokhlov, A. R. Polym. Gels Networks 1998, 6, 205. (4) Starodoubtsev, S. G.; Churochkina, N. A.; Khokhlov, A. R. Macromol. Symp. 1999, 146, 193. (5) Evsikova, O. V.; Starodoubtsev, S. G.; Khokhlov A. R. Vysokomolek. Soed., A 2002, 44, 1. (6) Grim, R. E. Clay Mineralogy; McGraw-Hill Series in Geology: New York, London, Toronto, 1953. (7) Hoffmann, U. Angew. Chem., Int. Ed. Engl. 1968, 7, 681. (8) Theng, B. K. G. Formation and Properties of ClaysPolymer Complexes; Elsevier: Amsterdam, 1979; Chapter 1. (9) Kukovsky, E. G. Structure and Properties of Clay Minerals; Naukova Dumka: Kiev, 1966. (10) Pauling, L. Proc. Nat. Acad. Sci. U.S.A. 1930, 16, 578.

MONT each consist of two silica sheets with one alumina sheet between them.6,7 Silica sheets have two planes of oxygen/hydroxyl ions, one of which is formed by faces and the other by vertexes of linked Si(O, OH)4 tetrahedrons. Due to isomorphous substitution of a fraction of silicon atoms by trivalent metal cations, the neighboring groups are ionized to ≡SiO-. The charge of these groups is compensated by positively charged counterions capable of ion exchange with the cations in the solution. In aqueous media, sodium or lithium cations give rise to high osmotic pressure in the galleries between the platelets, which may lead to infinite swelling in water. Figure 1a schematically shows a single platelet of the sodium salt of MONT surrounded by the counterions. The thickness of single-layer platelets in the dry state is about 1 nm, while their lateral size can exceed several tens of nanometers. In the dry state, the crystals of layered clays have a smectic structure schematically shown in Figure 1b where individual lines represent silica-allumina-silica triple-layer platelets. Swelling in water results in disordering of the platelets. Due to attraction forces between the edges and the surface of the sheets, clays form highly viscous liquids (gels) for which card-house structures, schematically illustrated in Figure 1c, are generally assumed.6-8 At high dilution, clays form low-viscosity, completely disordered stable colloidal dispersions. It should be mentioned that MONT clays are usually mixtures of MONT with some other clay particles. Some clays that contain, together with MONT, a small fraction of impurities (typically smaller than 30 wt%) are called bentonites (BENT).7 The interaction with cationic surfactants of MONT platelets in suspensions leads to the formation of organoclays, which normally consist of layered structures where the clay platelets alternate with surfactant bilayers,11-14 as schematically shown in Figure 1d. The investigation of aqueous suspensions of BENT incorporated in poly(acrylamide) (PAM) gels by X-ray diffraction did not show significant order in the arrangement of the clay

10.1021/la0505869 CCC: $33.50 © 2006 American Chemical Society Published on Web 11/17/2005

370 Langmuir, Vol. 22, No. 1, 2006

StarodoubtseV et al.

platelets in the gel is possible when a threshold concentration of the surfactant and the clay in the gel phase is achieved.15,16 Furthermore, surfactants with longer hydrophobic tails like hexadecylpyridinium chloride and hexadecyltrimethylammonium bromide induce the formation of larger smectic domains than surfactants with shorter tails such as dodecyltrimethylammonium bromide.18 The smectic ordering of clay platelets incorporated in the crosslinked gels was observed up to now only for gel-clay composites based on BENT obtained from a single source. In the present study, we will demonstrate that some features of the clay play an important role in the formation of smectic aggregates in the gel. We shall demonstrate that there is a strong correlation between the rheological characteristics of the clay suspensions, their microscopic structure, and the ability of clay platelets to form smectic aggregates in the presence of the cationic surfactants. Experimental Section

Figure 1. Schematic representation of the structure of clay suspensions in water and in the gel phase. (a) Negatively charged single platelet with counterions; (b) the dried clay powder; (c) the card-house structure; (d) the structure of lamellas with intercalated cationic surfactant; (e) dispersion of BENT embedded in PAM gel; (f) expected structure of lamellas after the removal of the surfactant; (g) preformed aggregates of BENT platelets in the suspension. Straight black lines are the platelets and individual platelet layers; dotted lines are negative charges of the clay platelets; winging thin black lines are PAM chains; gray circles with tails are surfactant ions; open circles are sodium cations. Sodium cations are shown only in (a). Each straight line in (b-g) represents the single platelet that is composed of three layers.

platelets in the gel phase.15,17 Thus, the structure of gel-BENT composites can be schematically represented as shown in Figure 1e. Further X-ray diffraction studies have shown that, unexpectedly, the disordered platelets of BENT suspensions embedded in PAM gels form highly ordered smectic structures after treatment with cationic surfactants. The structure and the extension of the ordered domains in the suspensions of free platelets and that of the platelets embedded in the gel were shown to be similar (Figure 1d).4,15-18 In addition, it was demonstrated that moderately crosslinked gels obtained in rather diluted solutions of monomers (4.5-6.0 wt%) practically do not inhibit the formation of smectic organoclay aggregates. However, when the concentration of the polymer chains of the network becomes high enough (15.0 wt% PAM gel), they begin to inhibit the process of self-assembling of the organoclay platelets. This result was explained by a model which assumes that a fraction of PAM chains is incorporated in the galleries between the collapsed platelets in the smectic aggregates and coexists there with the hydrocarbon tails of the surfactant.18 It was also shown that the smectic ordering of the (11) Dekany, I.; Haraszti, T. Colloids Surf., A 1997, 123-124, 391. (12) Lagaly, G.; Weiss, A. Kolloid Z. Z. Polymer. Sci. 1971, 243, 48 (13) Lagaly, G.; Weiss, A. Kolloid Z. Z. Polymer. Sci. 1971, 248, 979. (14) Lagaly, G.; Stange, H.; Weiss, A. Kolloid. Z. Z. Polymer. Sci. 1972, 250, 675. (15) Starodoubtsev, S. G.; Churochkina, N. A.; Khokhlov, A. R. Macromol. Symp. 1999, 146, 193. (16) Starodoubtsev, S. G.; Churochkina, N. A.; Khokhlov, A. R. Langmuir 2000, 16, 1529. (17) Starodoubtsev, S. G.; Ryabova, VA. A.; Dembo, A. T.; Dembo, R. A.; Aliev, L. I.; Wasserman, A. M.;.Khokhlov, A. R. Macromolecules 2002, 35, 6362. (18) Starodoubtsev, S. G.; Ryabova, A. A.; Khokhlov, A. R.; Allegra, G.; Famulari, A.; Meille, S. V. Langmuir 2003, 19, 10739.

Materials. BENT clay was purchased from Fluka ChemikaBioChemika Corp. while the MONT sample was a natural sodium montmorillonite (Cloisite Na+) from Southern Clay Products, Inc. X-ray diffraction powder patterns showed that the main impurities in BENT were small amounts of quartz and of the SIO2 mineral cristobalite, both of which are not expected to interfere in the present study. Acrylamide (AM), N,N′-methylene-(bis)-acrylamide (BAA), ammonium persulfate, N,N,N′,N′-tetramethylethylenediamine, and cetylpyridinium chloride (CPC) were obtained from Aldrich. The suspensions of the clays were prepared by mixing the clay powder on a magnetic stirrer in closed flasks for several days. Some suspensions were sonicated for 10 min using a 50 W high-intensity ultrasonic processor. The o.d. of the sonication rod was 3 mm, the i.d. of the sonicated reservoir was 6 mm, the power was kept at 20 W, and the time of sonication was 10 min. During sonication, the cylindrical reservoir was cooled by ice-water. Below we will identify the suspensions and the corresponding gels using the values of the concentration of the clays in the suspensions implying that they are correlated to the concentrations during the synthesis of the gels. Composite gels were synthesized by three-dimensional radical polymerization of monomers dissolved in the suspensions of the clays before the polymerization.11,12 The concentration of monomer, AM, during polymerization was 6.0 wt%, and the molecular ratio of the cross-linker, BAA to AM, was 1/500. Drying of the gels for SAXS measurements was carried out inside glass capillaries in a vacuum; the drying process was typically completed in 2 h. The clay-surfactant and polymer-clay-surfactant complexes were prepared in 0.01 M solutions of the surfactant. Methods. The content of sodium ions in the clays was determined with reference to an external standard by flame photometry using the 589 nm wavelength at maximum of transmission. The measurements were performed using a Carl Zeiss Jena AAS1 atomic absorption spectrometer. Reference samples were prepared by dilution of standard Merck solutions. The concentration of sodium in BENT and MONT was estimated at 3.57 and 3.71 wt%, respectively. Optical density was measured with a Hewlett-Packard 8452 spectrophotometer, while electroconductivity was determined with a CDM83 conductometer (Radiometer, Copenhagen). Optical microscopy was performed using an Axiolab Pol polarizing microscope (Carl Zeiss) equipped with a CCD camera. The thickness of the samples for optical microscopy was 100 µm. Small-angle X-ray scattering (SAXS) measurements were carried out with a Bruker NANOStar diffractometer with crossed Go¨bel mirrors. The measurements were performed under vacuum at 0.154 06 nm (Cu KR1). Samples were enclosed in glass capillaries for SAXS measurements. Acquisition times for diffraction patterns were typically of about 2-3 h. The mean long-range order dimension, L, in the clay-surfactant complexes was estimated from half-maximum widths of the Bragg peaks in the diffraction patterns using the Scherrer formula:19

Mechanism of Smectic Arrangement of Clay Platelets L)

λ βs cos θ

Langmuir, Vol. 22, No. 1, 2006 371 (1)

where βs is the full width at a half-maximum (fwhm) intensity of the peak (in radians) observed at the mean 2θ scattering angle. This procedure can still afford acceptable coherent domain size indications for scattering angles of about 2°, typical of the systems under study. Experimental fwhm’s were empirically corrected for instrumental broadening (ca. 0.10°) determined from the line breadth in diffraction patterns of crystals of silver behenate.

Results Smectic Ordering in Composite Gels with Incorporated BENT and MONT Suspensions Induced by the Interaction with CPC. Figure 2 shows wide-angle X-ray (WAXS) profiles obtained from the dry powders of BENT and MONT. The two clays provide slightly different characteristic silicate layer stacking maxima corresponding, respectively, to d-spacings of 1.25 nm for BENT and 1.18 nm for MONT. The BENT powder diffraction pattern reveals an additional broad weak peak at 2.1 nm and the presence of minor crystobalite and quartz impurities. The values of mean long-range order dimension, L determined applying the Scherrer equation to the diffraction profile of the silicate layer stacking peak in BENT and MONT samples, are 17.8 and 10.3 nm, respectively. The substantially larger propensity of platelets of BENT to stack suggests that they are more extended orthogonally to the stacking direction than those of MONT. Quantitative conclusions about platelet size from the analysis of the fwhm’s of higher-angle diffraction maxima are however more difficult to achieve, also because of different degrees of disorder within MONT and BENT layers. After the preparation of the aqueous suspensions, all the peaks in WAXS scattering curves, including the peak corresponding to the minor component, become negligible for all practical purposes. In our previous paper,18 we compared the structure of organoclays obtained from aqueous suspensions of BENT and BENT-PAM composite gels treated with cationic surfactants having different lengths of hydrocarbon tail. It was shown that suspensions of BENT with high enough concentration of BENT (2.4-4.0 wt%), in the presence of cationic surfactants, form smectic aggregates whose structure is schematically shown in Figure 1d. Unexpectedly, the incorporation of the clay platelets in 6 wt% PAM gel has not decreased their ability to form smectic aggregates under the action of the cationic surfactants. SAXS studies have shown that the positions of the maxima for composite gels practically coincide with those obtained from 4.0 and 2.4 wt% organoclay suspensions. Despite incorporation of the large clay platelets in the cross-linked PAM network, the value of the mean long-range order dimension, L, was also practically the same as for the suspensions of the free platelets of the same concentration. In a very concentrated 15 wt% PAM gel, the relative intensity of the maxima on the scattering curves significantly decreases, leaving, however, their width unchanged. In Figure 3, SAXS profiles obtained from the swollen BENTPAM and from MONT-PAM composite gels treated with CPC are compared. Figure 3a (curves 1 and 2) shows SAXS profiles obtained from the swollen BENT-PAM composites with 4.0 and 2.4 wt% of BENT. Both of the scattering curves show two maxima located at 2θ ) 2.1° and 4.25°, respectively. It can be estimated from the scattering curves that in both gels the ordered domains schematically shown in Figure 1d are built of 5-6 alternating surfactant bilayers and clay platelets. The decrease of the concentration to 1.0 wt% of BENT in the gel at synthesis leads (19) Vainshtein, B. K. Diffraction of X-rays by chain molecules; Elsevier Publishing Company: Amsterdam, London, New York, 1966.

Figure 2. X-ray intensity profiles of dry MONT (1) and BENT (2) powders, and aqueous suspensions of MONT (3) and BENT (4) with 4.0 wt% of the clays.

Figure 3. (a) SAXS profiles obtained from the composite gels with 4.0 (1, 4), 2.4 (2) and 1.0 wt% (3, 5) BENT in the swollen (1-3) and dried (4, 5) state. (b) SAXS profiles obtained from the swollen composite gels with 4.0 wt% of MONT (1) from the same gel after drying (2) and after the second swelling (3). SAXS profiles obtained from PAM gel with 2.4% suspension of BENT treated with 0.01 M solution of CPC (4), from PAM gel containing the preformed BENTCPC complex (5), and from PAM gel containing the preformed BENT-CPC complex after extraction with 50% ethanol (6). The concentration of sodium chloride was 0.01 M.

to disappearance of the maxima on the scattering curve (curve 3 in Figure 3a). The decrease of the concentration of BENT in the suspension also leads to disappearance of the maxima on the scattering curves.18 The drying of the 4 wt% BENT gel practically does not affect the parameters of the scattering curve (curve 4 in Figure 3). For the gel with 1.0 wt% of the clay, the drying results in the appearance of the highly ordered smectic aggregates (curve 5 in Figure 3a). The positions of the maxima for the dried 1 wt% BENT gel coincide with those of the 4 wt% gel, implying a similar structure.

372 Langmuir, Vol. 22, No. 1, 2006

StarodoubtseV et al.

Table 1. Conductivity of BENT and MONT Suspensions in Water substance

conductivity, F (au)

H2O MONT 1 wt% BENT 1 wt% MONT 0.33 wt% BENT 0.33 wt%

44 935 812 320 310

Figure 3b (curve 1) shows SAXS profiles obtained from a PAM-MONT composite gel containing 4.0 wt% of the clay after treatment with a 0.01 M solution of CPC. In contrast, with BENT, no smectic ordering of the clay platelets in the presence of the surfactant is observed. After drying, the clay platelets in the PAM network form smectic aggregates (curve 2 in Figure 3b). The scattering curve shows two peaks located at 2θ ) 2.2° and 4.5°, respectively. These values are close to those obtained for PAM-BENT-CPC complexes, but the peaks are less intense and much broader. These features suggest that the PAMMONT-CPC composites differ significantly from the BENT analogues. Curve 3 in Figure 3b shows a SAXS profile obtained from the same MONT gel after drying and repeated swelling for 72 h. This time interval is significantly longer than the time needed for equilibrium swelling in water of our gel composite samples. The curve presents a weakly pronounced maximum at 2θ ) 2.4-2.6° that is absent in the curve obtained from the same gel before the drying. This feature manifests the formation of additional stable interactions between the clay platelets in the dry state which persist to some degree upon swelling the gel again, as discussed in some recent papers.5, 18 Comparison of the Properties of BENT and MONT Suspensions. Although the predominant component in both our BENT and MONT samples is an expandable 2:1-type alumino silicate MONT mineral, the ability of the platelets of the two clays to arrange in highly ordered smectic aggregates under the action of the cationic surfactant is quite different. The platelets of BENT form such aggregates already in suspensions and in PAM gels containing 2.4 wt% of the clay, while the analogous self-ordering of MONT platelets was observed only during the drying of the gel. The higher ability of BENT to form smectic aggregates could arise in principle from a much higher charge density in comparison with MONT. If this was the case, the amount of the mobile counterions in the two clays should be different. The relative amount of the mobile counterions in the suspensions of the two clays should be roughly proportional to their electroconductivity, F. The measured values of F for suspensions of BENT and MONT are listed in Table 1. They practically coincide for diluted suspensions containing 0.3 wt% of the clays. For a 1.0 wt% suspension of BENT, the value of the conductivity is slightly lower in comparison with MONT. Thus, the existence of a large difference in the surface charge density of BENT and MONT platelets is unlikely. Additional support to this notion is the very similar content of mobile sodium cations in the two clays (see Experimental Section). An alternative reason for the observed difference in the behavior of BENT and MONT systems could be a larger average size of the platelets of BENT. Evidences consistent with this second hypothesis were given in the diffraction patterns of the two untreated clays (see Figure 2), which indicate larger size of the silicate stacks in the direction orthogonal to the stacks while the situation is less clear as far as individual platelets are concerned. We remark that it is reasonable to propose some direct correlation between the lateral size of the platelets and their tendency to stack, which may well determine larger size orthogonally to the alumina-silicate layers. Size differences between the platelets

Figure 4. (a) Dependence of viscosity on velocity gradient for BENT (1) and MONT (2) suspensions. (b) Dependence of complex viscosity on the frequency gradient for BENT (1) and MONT (2) suspensions. The concentration of the clays was 2.4 wt%.

of two clays should also determine substantial differences in the rheological behavior of the clay suspensions without the gel. Figure 4a shows the dependence of the viscosity, η, on the velocity gradient, dγ/dt, for BENT and MONT suspensions with a 2.4 wt% concentration of the two clays. Due to technical reasons, the experimental points could not be measured at the same values of velocity gradient. However, a large (more than 2 orders of magnitude) difference between the viscosities of the suspensions is apparent. Figure 4b shows the dependence of complex viscosity, |η|, on frequency, f, for the same clay suspensions. The values of complex viscosities at the same frequencies for BENT suspensions are 2-3 orders of magnitude higher than for MONT. The observed much higher viscosity of BENT suspensions, as well as the X-ray evidence, point clearly to a larger average diameter of BENT platelets in comparison with those of MONT.

Discussion The formation of a smectic phase in solutions of large thin disks was analyzed by De Gennes.20 For noninteracting thin disks of radius B, the concentration for the formation of the smectic phase is given by the formula:

B3 × Cd ) 0.67

(2)

where Cd is the number of disks per unit volume. The calculations show that disks with a thickness of 1.2 nm, a density of 2.6 g/cm3 (typical for clay platelets), and a radius of 50 and 100 nm form smectic phases when their concentration exceed 13 and 1.5 wt%, respectively. Thus, eq 2 predicts the right order of magnitude of (20) De Gennes, P. G.; Prost, J. The Physics of Liquid Crystals, 2nd ed; Oxford University Press: Oxford, 1993; p 63.

Mechanism of Smectic Arrangement of Clay Platelets

Langmuir, Vol. 22, No. 1, 2006 373

Figure 5. Optical micrographs of BENT (a) and MONT (b) suspensions immobilized in PAM gels. The concentration of the clays was 1.0 (i, ii), 2.4 (iii, iv), and 4.0 wt% (v, vi). Images on the right were obtained in polarized light mode.

the concentration for the transition of clay platelet suspensions to the ordered state even if, in the model, the attraction forces6-9 between the edges and the surface of the clay’s platelets are ignored. The structure and the properties of PAM-BENT composite gels modified with cationic surfactants were described in some detail in our previous papers.15-18 The question of how the large clay platelets manage to stack in highly ordered smectic aggregates, although constrained by a cross-linked gel, was however still open. The first and main problem is that the crosslink density of the networks studied is too high to allow completely disordered platelets (as shown in Figure 1e) to arrange in the ordered structure shown in Figure 1d. Second, it is well known, that PAM can form rather strong complexes with MONT.8 Due to this, in the presence of PAM the aggregates of MONT should dissociate. Moreover, some amount of highly ordered aggregates is observed by SAXS, even for the gel containing 15 wt% of acrylamide and 2.4 wt% of BENT during synthesis.18 At these concentrations, the volume fraction of PAM in the composite is 12.6 times higher than that of the clay. Assuming a random distribution of the platelets of the clay even in the dried state, they should be divided by a thick layer of PAM. The more plausible answer is that (i) the absence of crystalline diffraction maxima in X-ray profiles obtained from the gel-clay-surfactant composite should not be interpreted as implying complete disorder in the arrangement of the platelets and (ii) in the initial suspension and during the three-dimensional polymerization of AM, the

platelets form somewhat ordered aggregates of relatively loosely packed platelets. To substantiate this explanation, we have prepared composite PAM gels containing preformed suspensions of the claysurfactant smectic aggregates. Suspensions of BENT with a 2.4 wt% clay content were treated with CPC and sonicated. Then the PAM-BENT-CPC gel composite was prepared by the standard procedure described in the Experimental Section. A SAXS intensity profile obtained from these gel samples is shown in Figure 3b (curve 4). The observed pattern is closely similar to the one obtained from the sample prepared by addition of CPC to prepared PAM-BENT composite (Figure 3b, curve 5). Thus, the clay platelets in the gel were ordered and formed smectic aggregates schematically shown in Figure 1d. Then, the surfactant was subsequently removed from the gel via the extraction with a water-ethanol mixture containing sodium chloride.16 After the removal of the surfactant and washing with water, the SAXS profile of the gel showed no maxima. This seems quite unlikely that the preformed ordered aggregates of the large clay platelets were able to completely disorder because the mesh size of the gel is significantly smaller that the size of the platelets. Thus, the expected structure of the platelets aggregates should be roughly similar to the one schematically shown in Figure 1f. However, the disordering of the platelets due to removing of the CPC surfactant was enough to make it impossible to detect significant degrees of order in the assemblies of the platelets by X-ray diffraction.

374 Langmuir, Vol. 22, No. 1, 2006

Additional evidence that the platelets of the clay in the investigated suspensions are oriented in the gel phase comes from the direct observation of the gel-clay composites by optical microscopy. Figure 5 shows optical micrographs of BENT (a) and MONT (b) suspensions immobilized in PAM gels. The concentration of the clays was varied from 1.0 up to 4.0 wt%. Large aggregates with characteristic size of several micrometers are observed for both clays. Images obtained in polarized light evidence the marked optical anisotropy of the aggregates. It should be noted that there is no direct correlation between the optical observation of clay aggregates in the suspensions and the observation of the maxima on SAXS profiles obtained from the same samples after treatment with cationic surfactant. For instance, the anisotropic aggregates are observed optically in the suspensions and gel composites of MONT, while the maxima on the scattering SAXS curves are absent. It follows from the above discussion that the idealized, random card-house model (Figure 1c and e), often assumed for clay dispersions in water, is probably too simplistic: it appears that some order with approximate parallel orientation of portions of adjacent clay platelets is probably maintained also in clay suspensions. This structure is schematically shown in Figure 1g. The degrees of order and the size of the semi-ordered aggregates may vary substantially in suspensions of different clays and are very plausibly influenced by the lateral size of the clay platelets.

StarodoubtseV et al.

Conclusion The interaction of PAM-BENT gel composites with the cationic surfactant CPC results in the formation of smectic aggregates composed of alternating layers of BENT platelets and of surfactant bilayers. In the PAM-MONT-CPC gels, analogous aggregates are observed only after drying. Rheology studies indicate that the observed difference in the ability of the two closely related clays BENT and MONT to form smectic aggregates in the presence of CPC can be explained by a much larger size of BENT platelets. The transitions occur despite the relatively large size of clay platelets as compared to the mesh size of the PAM network. The surprising formation, upon addition of CPC surfactant, of highly ordered smectic structures in the cross-linked polymer gels can be explained by the existence in the initial aqueous suspensions of preformed, partially ordered, clay platelet aggregates. These assemblies, which remain after the synthesis of the gels, are not detected by SAXS but are apparent by optical microscopy of the suspensions. Acknowledgment. S.G.S. gratefully acknowledges the CARIPLO Foundation and the Landau Network-Centro Volta for financial support and for the opportunity to work at the Politecnico di Milano. Support from the Italian MIUREnDash-PRIN 2003 for G.A., A.F., and S.V.M. is also acknowledged. LA0505869