Physical Adsorption

Journal of Colloid and Interface Science 327 (2008) 295–301

Contents lists available at ScienceDirect

Journal of Colloid and Interface Science www.elsevier.com/locate/jcis

Physical adsorption of organic molecules on the surface of layered silicate clay platelets: A thermogravimetric study V. Mittal ∗ , V. Herle Department of Materials, Institute of Polymers, ETH Zurich, 8093 Zurich, Switzerland

a r t i c l e

i n f o

a b s t r a c t

Article history: Received 14 June 2008 Accepted 17 August 2008 Available online 9 September 2008 Keywords: Physical adsorption Adsorbents Montmorillonite Thermal analysis

Physical adsorption of various adsorbents on the surface of premodified montmorillonite platelets was performed to fully organophilize the inorganic platelets for the purpose of their easy nanoscale dispersion in the polymer matrices during compounding. Different extents of adsorption could be achieved owing to the nature and the functionality of the adsorbents. High molecular weight adsorbents not only enhanced the organic coverage of the platelets but also were observed to contribute toward the thermal stability improvement of the organic modification, thus further fitting the use of such clays for high temperature compounding. The amount of adsorption could also be quantified with respect to the initial amount of adsorbent used in the process. The importance of a clean surface free from any excess surface modification or adsorbent molecules was emphasized. The adsorption process is an effective means to generate such high potential montmorillonites and is much simpler in technique than the common methods of grafting of polymer chains from the clay surface. © 2008 Elsevier Inc. All rights reserved.

1. Introduction Polymer nanocomposites have been a focus of vigorous research in the recent years owing to the tremendous property enhancement at very low inorganic filler volume fractions [1–7]. Layered silicate montmorillonites have been the inorganic material of choice because of their easy swelling in water, thus allowing subsequent surface modifications with alkyl ammonium ions [8– 11]. In the pristine state, these montmorillonite platelets are held together by electrostatic interactions, thus leaving an interlayer between them. The modification at the polar surfaces reduces the surface energy of the platelets and renders them organophilic and partially polar, thus helping to disperse them in the polymer matrices [8–11]. However, such modified montmorillonites are mostly useful for polar polymer matrices like polyurethane, epoxy, and polyamides, etc. [1–7]. The hydrophobic nature of polyolefin matrices like polypropylene and polyethylene hinders the nanoscale dispersion of these modified platelets as the completely nonpolar chains do not intercalate the partially polar interlayers of the stacks of montmorillonite platelets [12–16]. A number of approaches have been reported in the literature to overcome the negative interactions between the polyolefins and the modified montmorillonites. The most commonly used ap-

*

Corresponding author. Current address: BASF SE, Polymer Research, Ludwigshafen, Germany. Fax: +41 44 632 1082. E-mail addresses: [email protected], [email protected] (V. Mittal). 0021-9797/$ – see front matter doi:10.1016/j.jcis.2008.08.036

©

2008 Elsevier Inc. All rights reserved.

proach is to use low molecular weight surfactants or compatibilizers like polypropylene grafted maleic anhydride (PP-g-MA) or polyethylene grafted maleic anhydride (PE-g-MA), which because of their amphiphilic nature can bridge the polymer chains as well as clay platelets. This approach of partially polarizing the matrix by using polar compatibilizers like PP-g-MA and PE-g-MA leads to enhanced intercalation and exfoliation; however, the mechanical properties of the resulting nanocomposites suffer because of the low molecular weight of the compatibilizers [13–18]. Another approach is to fully organophilize the partially polar surface of the organically modified montmorillonite platelets so as to reduce the electrostatic forces between them to a minimum. Such loosely held stacks of the platelets can then be exfoliated in the polymer matrices by using shear and thus do not require the positive interactions for the polyolefins chains to enter the interlayers [12,19]. This system is more like a kinetic entrapment of the polymer chains along with clay platelets after cooling the melt in which clay platelets were exfoliated by shear rather than a thermodynamic process of filler exfoliation; however, it has been reported to be stable even after several cycles of processing of the composite material at high temperature [12,19]. The various techniques reported in the literature include modification of the clay platelets with ammonium ions of longer chain lengths [12], grafting of oligomeric chains “to” or “from” the clay surface [20–29], and chemical reactions on the clay surface with reactive surface modifications [30–37]. These routes lead to increased basal plane spacing of the clay platelets and reduced surface energy, thus making them susceptible to exfoliation when sheared along with the polymer matrix.

296

V. Mittal, V. Herle / Journal of Colloid and Interface Science 327 (2008) 295–301

As the cross-sectional area of the ammonium ion cation adsorbed on the clay surface is much smaller than the area available per cation in the montmorillonite with medium cation exchange capacity (CEC), the partially uncovered and hence partially polar surface can also be fully organophilized by increasing the chain density of the alkyl ammonium modification on the surface owing to its higher cross-sectional area [12,19]. For example, a better exfoliation was reported when the montmorillonite platelets were modified with tetraoctadecylammonium and trioctadecylmethylammonium ions as compared to conventional dioctadecyldimethylammonium (2C18 ammonium) ions. However, owing to the solubility limitations of these molecules, it is not very optimum to modify the clay platelets with high chain density ammonium ions. On the other hand, adsorption of long chain solvent-soluble molecules on the surface of montmorillonite already modified with conventional ammonium ions, e.g., dioctadecyldimethylammonium, can offer an easy route which does not suffer from the solubility concerns [38]. These molecules (e.g., long alkyl chain alcohols) can adsorb physically in between the gaps generated after modification with ammonium ions by forming H bonds with the OH groups present either in the inside structure of clay crystals or on the edges of the platelets. Also, the adsorption has been reported to take place on the preadsorbed water molecules in the clay interlayers [39– 42]. Apart from that, difunctional or multifunctional molecules can also be employed in order to generate much stronger adsorption on the clay surface. The present paper reports the trials of physical adsorption of various adsorbents on the montmorillonite surface. The montmorillonite was modified with conventional ammonium ions and was procured commercially. Adsorbents with different molecular weights and different numbers of functional groups in the molecule were used and their adsorption on the clay surface was quantified by using high resolution thermogravimetric analysis (TGA). Trials with different routes to achieve optimum adsorption on the surface were also carried out. The optimum amount of adsorption on the clay surface was also quantified. 2. Experimental 2.1. Materials Surface-modified bentonite with the trade name Tixogel VP was procured from Southern Clay Inc. (Gonzales, TX). The montmorillonite was premodified with dimethylditallow ammonium modification. The as-received clay had average particle size of 2.5 μm and narrow particle size distribution. The cation exchange capacity of the clay was in the range of 880 μeq g−1 of clay. 1-Octadecanol, 1,2-hexadecanediol, 1,2-hexanediol, Tween 85 (polyethylene glycol sorbitan trioleate), and poly(vinylpyrrolidone) (PVP) were received from Fluka (Buchs, Switzerland) and were used as received. Reaction solvents tetrahydrofuran (THF), ethyl acetate, and alcohol were purchased from Aldrich (Buchs, Switzerland). 2.2. Washing of surface-modified montmorillonite The commercially procured modified montmorillonite was thoroughly washed before physical adsorption trials to remove the excess of ammonium modification molecules generally present in the commercially modified montmorillonites [43]. The clay was suspended in 90:10 mixture of ethyl acetate and alcohol at 70 ◦ C and was sonicated for 10 min (ultrasonic horn at 70% amplitude). The dispersion was then sheared mixed by using high speed shear mixer for duration of 10 min (Ultra-Turax, T50, IKA, Staufen, Germany). The dispersion was kept at 70 ◦ C under stirring overnight and was subsequently filtered and thoroughly washed on filter. The wet clay was again suspended in ethyl acetate:alcohol mixture and

was similarly processed as before. Subsequently, the clay was filtered and dried under reduced pressure at 70 ◦ C. The extent of washing and the cleanliness of the surface were quantified by using high resolution thermogravimetric analysis. The washing step was repeated if the low temperature degradation peak associated with the excess surface modification was detected in the TGA thermogram. 2.3. Physical adsorption Two different methods were employed for physical adsorption trials. In the first method (no solvent method), clay (2 g) and adsorbent (1.5 g) were weighed and mixed in a mortar and heated in the oven to mix the contents or to melt the alcohol, in case it is in solid form. The mixture was stirred regularly and clay was subsequently washed with THF to remove the excess of alcohol molecules. In the second method (solvent method), the weighed amounts of clay and adsorbents were added in THF. The contents were thoroughly stirred and occasionally sonicated. The separation of the clay from the dispersion was achieved by centrifugation as it was not possible to filter the clay because of excessive swelling of clay in THF. The contents were centrifuged a number of times to clean the clay off the excess alcohol molecules. 2.4. Thermogravimetric analysis Hi-resolution (Hi-Res) thermogravimetric analysis, in which the heating rate is coupled to mass loss, i.e., the temperature is not increased until the mass loss is complete at that particular temperature, was performed on a Q 500 thermogravimetric analyzer (TA Instruments, New Castle, DE) to quantify the amount of adsorption. The dried clay samples were measured in air stream using a temperature range of 50–900 ◦ C. 3. Results and discussion Physical adsorption of polar functional groups with terminal long hydrophobic alkyl chains in the molecule can efficiently help to organophilize the clay surface and decrease the surface energy [39–42] so as to achieve their nanoscale dispersion when subsequently mixed with polymer matrices especially polyolefins. It is also synthetically less cumbersome to achieve adsorption as compared to more complex routes of grafting of polymer chains from the clay surface or other chemical reactions on the clay surface. However, it is very important to ensure that all the excess of the added adsorbent is completely washed off as the excess adsorbent may lead to thermal instability of the system when the inorganic filler is processed at high temperatures. Fig. 1 shows the schematic of such an adsorption process. The figure only serves to emphasize the process of adsorption and does not provide information on the order in which the adsorbent molecules are present along with ammonium ions ionically bound on the surface. Fig. 2 shows the chemical architecture of the various adsorbent molecules which were adsorbed on the clay surface. Alcohols with different chain lengths and different number of OH groups in the molecules were employed to characterize the effect of these parameters on the achieved adsorption. Apart from

Fig. 1. Schematic of physical adsorption process on the clay surface.

V. Mittal, V. Herle / Journal of Colloid and Interface Science 327 (2008) 295–301

297

Fig. 3. Three-dimensional model of Tween 85 indicating the potential adsorption sites.

Fig. 2. Chemical structures of the adsorbents used for adsorption on the clay surface: (a) 1-octadecanol, (b) 1,2-hexadecanediol, (c) 1,2-hexanediol, (d) Tween 85, and (e) PVP.

that, high molecular weight adsorbents with multiple functional groups in the molecule like Tween 85 (polyethylene glycol sorbitan trioleate) and poly(vinylpyrrolidone) were also chosen. These entities with a large number of potential adsorption sites present in their backbones can be expected to generate stronger and multiple bonds on the clay surface. As an example, in Fig. 3 is shown the three-dimensional model of Tween 85 showing the large number of prospective sites which can enter into a physical bond on the clay surface, thus securely bonding the high molecular weight molecule on the clay surface. This strong bonding also helps to avoid these molecules from being pulled away during washing or other processing steps. It has been reported that the commercially modified montmorillonites contain an excess of surface modifier molecules, which are thermally less stable and may induce unwanted interactions with the adsorbent molecules and subsequently polymer matrix, thus requiring cleaning the montmorillonite first [43]. Fig. 4a depicts derivative weight thermograms of Tixogel VP montmorillonite before and after washing in ethyl acetate:alcohol mixture. The asreceived surface-modified montmorillonite had two degradation peaks at 243 and 285 ◦ C (curve I). It is clear from the thermogram of the washed clay (curve II) that the low temperature degradation peak of 243 ◦ C owing to the presence of excess of surface modifier molecules not bound to the clay surface could be efficiently eliminated after washing. The total weight loss also reduced from 36.06 to 34.14% after washing, indicating the washing away of the excess molecules. The weight loss of these clays plotted as a function of temperature in Fig. 4b also confirms this notion.

(a)

(b) Fig. 4. (a) Derivative weight and (b) weight loss thermograms of organically modified montmorillonite Tixogel VP before (I) and after (II) washing.

298

V. Mittal, V. Herle / Journal of Colloid and Interface Science 327 (2008) 295–301

Table 1 Details of the physical adsorption on the clay surface Adsorbent

Molecular weight, g/mol

Total weight loss in TGA, %

Weight loss due to adsorption, %

Weight of adsorbed material, g/g of clay

Moles of adsorbed material, mol/g of clay

Corresponding CEC, μeq g−1

% of additional surface covered

1-Octadecanol 1,2-Hexadecanediol 1,2-Hexanediol Tween 85 PVP (1 g initial) PVP (1.5 g initial)

270.49 258.44 118.17 1838.58 10,000 10,000

40.17 40.26 47.88 42.20 54.72 59.84

6.03 6.12 13.74 8.06 20.58 25.70

0.0452 0.0459 0.1030 0.0605 0.1544 0.1928

1.74 e−4 1.78 e−4 8.72 e−4 5.11 e−4 0.15 e−4 0.19 e−4

237 126a 674a – – –

26.96 14.33a 76.59a – – –

a

Values in dialcohols have been halved for quantitative comparison of the adsorption with monoalcohols.

Fig. 5. TGA thermograms of organically modified montmorillonite before adsorption (I), after adsorption with 1-octadecanol and one cycle of washing (II), and after adsorption with 1-octadecanol and 10 cycles of washing (III). Inset shows the derivative weight curves of these clays as a function of temperature. Results obtained by the solvent mixing method are shown.

Trials with monoalcohol adsorption on the clay surface were performed with 1-octadecanol. Though di- or multifunctional alcohols are expected to adsorb on the clay surface strongly as compared to monoalcohols, owing to the single unbranched chain, monoalcohols also do not suffer from the steric hindrance and can easily adsorb on the clay surface. Fig. 5 shows the TGA thermograms of the clay after adsorption (III) by solvent mixing method compared with the reference clay (I). The results by using a mixing method without solvent were also similar. The change in the TGA curve as well as increase in the total weight loss indicates that a certain amount of alcohol is adsorbed on the surface. The effect of rigorous cleaning was once again demonstrated by comparing curves II and III. Curve II is the TGA thermogram of the clay after adsorption and with one cycle of washing, whereas the curve III is the resulting thermogram after 10 cycles of washing and subsequent washings yielding no change in the weight loss or the peak positions in the thermogram. The thermal degradation peak of the adsorbed alcohol molecules is also placed at higher temperature than the case when a lot of excess of alcohol was present, indicating better thermal stability of the alcohol molecules when adsorbed to the clay surface. The inset in Fig. 5 also shows the derivative weight thermograms of these clays and it is clear from the curves that though the thermal stability of alcohol increased after adsorption on the clay surface, the overall thermal stability of the clay is decreased owing to the low molecular weight alcohol molecules attached to the clay surface. A total weight loss of 40.17% was observed in the clay sample after adsorption as com-

pared to the 34.14% of the starting clay. Table 1 quantifies the amount of alcohol adsorption calculated from the increase in the weight loss in the thermogravimetric analysis. A value of 0.045 g or 1.75 e−4 mol of octadecanol could be adsorbed per gram of clay which represents a cation exchange capacity of 237 μeq g−1 . This value when compared with the cation exchange capacity of Tixogel VP of ∼880 μeq g−1 represents an additional 27% coverage of the available area on the clay surface. This result confirms that the alcohol adsorption method can be a good route to fully organophilize the clay platelets’ surface. 1,2-Hexadecanediol and 1,2-hexanediol were two diols used for adsorption on the clay surface. The shorter chain 1,2-hexanediol is expected to adsorb better owing to the less steric hindrance and ease of movement on the clay surface. Table 1 shows that a net weight loss of 40.26% was observed in the TGA thermograms of clay adsorbed with 1,2-hexadecanediol prepared by both mixing methods. The TGA thermograms of these washed clays are depicted in Figs. 6a and 6b. Though the thermal stability of the clay was partially reduced as in the case of octadecanol, significant amounts of alcohols could be adsorbed onto clay surface. This increased weight loss corresponded to an adsorption of 0.046 g or 1.78 e−4 mol of hexadecanediol per gram of clay. These values when corrected for the two OH groups in the alcohol molecule lead to a representative CEC of 126 μeq g−1 and 14.33% of additional platelet surface area covered per OH group adsorbed. In this correction, the cation exchange capacity values and surface coverage values are simply halved in order to compare these values with the cation exchange capacity and surface coverage of monoalcohols. A much higher amount of adsorption was observed when 1,2-hexadecanediol was replaced with 1,2-hexanediol. A total increase of 13.74% in the weight loss in TGA thermograms was translated to a 0.10 g or 8.7 e−4 mol of hexanediol adsorbed on the clay surface which is much higher than the adsorption achieved by other long chain alcohols. It confirms that the mobility of the adsorbent during the adsorption process has a very important role in the amount of adsorption achieved. However, hexanediol has a low molecular weight and has low temperature degradation; therefore, it is not suitable to be used for commercial applications. High molecular weight adsorbents with multiple adsorption sites are more preferable choices because of there high temperature stability as well as higher number of contacts on the clay surface. Fig. 7a details the adsorption trials of Tween 85 on organomontmorillonite. Curve I is the derivative weight thermogram of clay before adsorption, whereas curve II is the thermogram after adsorption. The presence of a sharp peak at 223 ◦ C indicates the presence of free Tween 85 on the clay surface. The peak is significantly reduced when the same material is further washed for more cycles (curve III), again stressing the need of a clean surface for correct development of the interface between the clay and the polymer when the modified clay is compounded with it. Fig. 7b also shows the weight loss plotted against temperature for these clays. A loss of 8.06% was observed for the clay modified

V. Mittal, V. Herle / Journal of Colloid and Interface Science 327 (2008) 295–301

(a)

299

(a)

(b) (b) Fig. 6. (a) TGA thermograms of organically modified montmorillonite before adsorption (I) and after adsorption with 1,2-hexadecanediol (II) when mixed without solvent and (b) TGA thermograms of the same clays when prepared by the solvent mixing method. Insets show the derivative weight curves of these clays as a function of temperature.

with Tween 85. The peak degradation temperature in the modified clay is observed at higher temperature of 309 ◦ C than the degradation temperature of 285 ◦ C in the starting clay, indicating that the adsorbent is synergistically modifying the clay surface and thus contributing to the improvement of thermal stability of the system. It is also worth noting that the onset thermal degradation temperature of the clay adsorbed with Tween 85 is somewhat lower than the unadsorbed clay, thus indicating that may be there are still some unattached molecules left in the clay interlayers which could not be washed. That is actually also the rationale of using the peak degradation temperature as a comparison between the two clays in this case to indicate that once the clay is fully cleaned off the excess, the thermal degradation temperature can be enhanced by 25 ◦ C or more. Similar to Tween 85, poly(vinylpyrrolidone) also showed synergism in the thermal behavior enhancement as the peak degradation temperature of the PVP-modified clay increased to 332 ◦ C (curve II in Fig. 8a), indicating much better thermal resistance than the parent modified clay. This increased thermal resistance is par-

Fig. 7. (a) Derivative weight and (b) weight loss thermograms of organically modified montmorillonite before adsorption (I), after adsorption with Tween 85 and few cycles of washing (II), and after adsorption with Tween 85 and rigorous washing (III).

ticularly beneficial when compounding the clay at high compounding temperatures required by polymer matrices like polypropylene. The mechanism of interaction leading to synergism can be expected to be same as the improvements in the thermal stability of ammonium modification molecules when they are bound to the clay surface because the clay surface shields these organic molecules, thus delaying their decomposition. The weight loss in the PVP-modified clay was increased by 25.7%, indicating that 0.2 g of the polymer could be adsorbed per gram of clay. It is by far the maximum adsorption achieved as compared to other adsorbents. It is expected to have large adsorption in the case of PVP owing to a large number of adsorption sites present in the backbones of the polymer chains which can adsorb at different locations in the clay platelets. The mechanism of adsorption can be defined in this case as follows: the more electronegative oxygen atoms in the PVP chains can form hydrogen bonds with OH groups present either on edges or in the internal structure of montmorillonite or also with the residual water molecules adsorbed on the surface of clay platelets.

300

V. Mittal, V. Herle / Journal of Colloid and Interface Science 327 (2008) 295–301

(a)

Fig. 9. Extent of PVP adsorption on the clay surface as a function of initial amount of PVP taken for the adsorption process.

(b) Fig. 8. (a) Derivative weight and (b) weight loss thermograms of organically modified montmorillonite before adsorption (I), after adsorption with 1.5 g of initially added PVP (for 2 g of organically modified montmorillonite), (II) and after adsorption with 1 g of initially added PVP (for 2 g of organically modified montmorillonite) (III).

It is important to note that as though the adsorbents were used in excess during the adsorption reaction, the final adsorption values may not represent the optimum or maximum achievable values by this process. Therefore, it was also of interest to correlate the amount of adsorbed material to the amount of adsorbent initially taken in the adsorption process. Curve III in Fig. 8a shows the TGA thermogram of the PVP adsorbed clay when the initial amount of PVP used in the adsorption reaction was 1 g. The increase in TGA weight loss corresponded to 0.15 g of PVP adsorbed per g of clay. The thermal stability of the resulting clay was also inferior to the case when 0.2 g of PVP was adsorbed on the clay surface by starting with 1.5 g of PVP for adsorption, confirming that PVP adsorption provides more thermal stability. Fig. 8b also shows the weight vs temperature plots of these clays, indicating the exact changes and stages of weight loss. Two more trials where higher initial amounts of PVP were used in the adsorption reaction (2 and 2.5 g) generated the same amount of adsorption as 1.5 g of initially taken PVP (0.2 g per g of clay) and the TGA plots are indistinguishable with curve II of Fig. 8a. Fig. 9 shows the phenomenon graphically. The adsorption first grows with increasing

Fig. 10. TGA thermograms of benzyl(2-hydroxyethyl)methyloctadecylammonium (BzC18OH)-modified clay (I) and modified clay after esterification reaction with dotriacontnoic acid (lacceroic acid) (II) [37].

the amount of initially added PVP; however, it reaches a plateau after a certain extent of adsorption has been reached and beyond that increasing the amount of initial PVP has no affect on the final extent of adsorption. This result clearly demonstrates that the amount of surface coverage of the clay with the adsorbent can be finely controlled. The results obtained with various adsorbents establish the fact that physical adsorption is an efficient and easier approach to achieve organic modification of clay platelets. The modified montmorillonites so obtained represent high potential materials for their susceptibility for nanoscale dispersion in the polymer matrix when mixed at higher temperatures. For reference, Fig. 10 also shows the TGA thermograms of benzyl(2hydroxyethyl)methyloctadecylammonium (BzC18OH)-modified clay and the clay after surface esterification reaction with dotriacontnoic acid (lacceroic acid) [37]. An increase of 18 wt% was achieved in the organic weight loss after the surface reaction. Physical adsorption results are also on similar lines and can be achieved with less effort along with enhancement of thermal stability, indicating the usefulness of this approach along with other chemical modification techniques.

V. Mittal, V. Herle / Journal of Colloid and Interface Science 327 (2008) 295–301

4. Conclusions Organically modified montmorillonite could be additionally adsorbed on the surface with adsorbents of different molecular weights and numbers of adsorbing moieties. The low molecular weight adsorbent could adsorb in large amounts owing to better mobility on the clay surface; however, as a result, the thermal degradation of the organic modification also onsets at a lower temperature. Adsorption of high molecular weight adsorbents was very effective in improving the organic coverage of the clay as these adsorbents contained multiple numbers of adsorption moieties in their molecules. The thermal resistance of the organic modification on the clay surface was enhanced synergistically too. It was also observed that there exists an optimum amount of adsorbent which can be adsorbed on the clay surface. After the optimum amount has been reached, no further adsorption occurs irrespective of the initial amount of adsorbent added during the adsorption process. The thermal behavior as well as interfacial interactions of the modified clays is dependent on the extent of cleanliness of the clay surface. The modified clays generated by the physical adsorption method represent high potential loosely held layered silicate platelets with minimal electrostatic forces between them and are susceptible to easy delamination by shearing during compounding with the polymer matrices. References [1] A. Usuki, Y. Kojima, M. Kawasumi, A. Okada, Y. Fukushima, T. Kurauchi, O. Kamigaito, J. Mater. Res. 8 (1993) 1179. [2] M. Alexandre, Ph. Dubois, Mater. Sci. Eng. R 28 (2000) 1. [3] A. Usuki, M. Kawasumi, Y. Kojima, A. Okada, T. Kurauchi, O. Kamigaito, J. Mater. Res. 8 (1993) 1174. [4] E.P. Giannelis, Adv. Mater. 8 (1996) 29. [5] J.W. Gilman, C.L. Jackson, A.B. Morgan, R. Harris, E. Manias, E.P. Giannelis, M. Wuthenow, D. Hilton, S.H. Phillips, Chem. Mater. 12 (2000) 1866. [6] M.A. Osman, V. Mittal, M. Morbidelli, U.W. Suter, Macromolecules 36 (2003) 9851. [7] P.C. LeBaron, Z. Wang, T.J. Pinavaia, Appl. Clay Sci. 15 (1999) 11. [8] S.W. Bailey, in: S.W. Bailey (Ed.), Reviews in Mineralogy, Virginia Polytechnic Institute and State University, Blacksburg, 1984.

301

[9] B.K.G. Theng, The Chemistry of Clay-Organic Reactions, Adam Hilger, London, 1974. [10] G. Lagaly, K. Beneke, Colloid Polym. Sci. 269 (1991) 1198. [11] S.W. Bailey, in: G.W. Brindley, G. Brown (Eds.), Crystal Structure of Clay Minerals and Their X-Ray Identification, Mineralogical Society, London, 1980. [12] M.A. Osman, V. Mittal, U.W. Suter, Macromol. Chem. Phys. 207 (2007) 68. [13] P. Reichert, H. Nitz, S. Klinke, R. Brandsch, R. Thomann, R. Mülhaupt, Macromol. Mater. Eng. 275 (2000) 8. [14] M. Kawasumi, N. Hasegawa, M. Kato, A. Usuki, A. Okada, Macromolecules 30 (1997) 6333. [15] N. Hasegawa, M. Kawasumi, M. Kato, A. Usuki, A. Okada, J. Appl. Polym. Sci. 67 (1998) 87. [16] D. Kaempfer, R. Thomann, R. Mülhaupt, Polymer 43 (2002) 2909. [17] A. Oya, Y. Kurokawa, H. Yasuda, J. Mater. Sci. 35 (2000) 1045. [18] W. Xu, G. Liang, W. Wang, S. Tang, P. He, W.P. Pan, J. Appl. Polym. Sci. 88 (2003) 3225. [19] M.A. Osman, J.E.P. Rupp, U.W. Suter, J. Mater. Chem. 15 (2005) 1298. [20] X. Fu, S. Qutubuddin, Polymer 42 (2001) 807. [21] L.P. Meier, R.A. Shelden, W.R. Caseri, U.W. Suter, Macromolecules 27 (1994) 1637. [22] U. Velten, R.A. Shelden, W.R. Caseri, U.W. Suter, Y. Li, Macromolecules 32 (1999) 3590. [23] U. Velten, S. Tossati, R.A. Shelden, W.R. Caseri, U.W. Suter, R. Hermann, M. Muller, Langmuir 15 (1999) 6940. [24] P. Uthirakumar, C.J. Kim, K.S. Nahm, Y.B. Hahn, Y.S. Lee, Colloids Surf. A 247 (2004) 69. [25] X. Fan, C. Xia, T. Fulghum, M.K. Park, J. Locklin, R.C. Advincula, Langmuir 19 (2003) 916. [26] A.S. Moet, A. Akelah, Mater. Lett. 18 (1993) 97. [27] M. Albrecht, S. Ehrler, A. Muhlebach, Macromol. Rapid Commun. 24 (2003) 382. [28] O. Prucker, J. Ruhe, Macromolecules 31 (1998) 602. [29] V. Mittal, J. Colloid Interface Sci. 314 (2007) 141. [30] H.H.J. Persson, W.R. Caseri, U.W. Suter, Langmuir 17 (2001) 3643. [31] A.J. Guiomar, J.T. Guthrie, S.D. Evans, Langmuir 15 (1999) 1198. [32] S. Lofas, B. Johnsson, J. Chem. Soc., Chem. Commun. (1990) 1526. [33] I. Wilner, A. Riklin, B. Shoham, D. Rivenson, E. Katz, Adv. Mater. 5 (1993) 912. [34] S. Pan, D.G. Castner, B.D. Ratner, Langmuir 14 (1998) 3545. [35] L. Yan, W.T.S. Huck, X.-M. Zhao, G.M. Whitesides, Langmuir 15 (1999) 1208. [36] S.D. Evans, R. Sharma, A. Ulman, Langmuir 7 (1991) 156. [37] V. Mittal, J. Colloid Interface Sci. 315 (2007) 135. [38] D.J. Greenland, R.H. Laby, J.P. Quirk, Trans. Faraday Soc. 61 (1965) 2024. [39] Q. Zhou, R.L. Frost, H. He, Y. Xi, H. Liu, J. Colloid Interface Sci. 307 (2007) 357. [40] Q. Zhou, H. He, R.L. Frost, Y. Xi, J. Phys. Chem. C 111 (2007) 7487. [41] Q. Zhou, R.L. Frost, H. He, Y. Xi, J. Colloid Interface Sci. 307 (2007) 50. [42] Q. Zhou, R.L. Frost, H. He, Y. Xi, M. Zbik, J. Colloid Interface Sci. 311 (2007) 24. [43] V. Mittal, J. Mater. Sci. 43 (2008) 4972.