Langmuir 2007, 23, 3953-3960
3953
Interactions at the Surface of Organophilic-Modified Laponites: A Multinuclear Solid-State NMR Study Silvia Borsacchi, Marco Geppi,* Lucia Ricci, Giacomo Ruggeri, and Carlo A. Veracini Dipartimento di Chimica e Chimica Industriale, UniVersita` di Pisa, V. Risorgimento 35, 56126 Pisa, Italy ReceiVed October 16, 2006. In Final Form: January 29, 2007 Organically modified clays are largely employed in the preparation of nanostructured materials. The structural and dynamic characterization of the clay surface appears very important in the perspective of understanding the molecular mechanisms determining the improvement of the material properties. To this aim, in this work, a synthetic clay, Laponite, was studied in its untreated hydrophilic Na+-form, after ion exchange with alkylammonium cations and after subsequent grafting reaction with an alkoxysilane. These three samples were characterized by IR, SEM, TGA, and X-ray techniques and were deeply investigated by means of a wide combination of 29Si, 13C, and 1H high- and low-resolution solid-state NMR experiments. The grafting reaction with alkoxysilanes, occurring at the clay platelet edges, resulted in a reduction of the clay inter-platelet distances, and in an increased disorder in both the arrangement of the platelets and the conformational structure of the intercalated organic cation chains, probably due to the relative twisting of adjacent platelets.
Introduction Polymer-clay nanocomposites are attracting much attention in the field of materials, since the nanodispersion of clays in a polymeric matrix proved to be successful in improving its mechanical, barrier, and thermal properties.1-3 The material behavior depends on the interactions occurring between clay and polymer, which in turn are strongly affected by the properties of the clay platelets surface. A key step of most preparation methods is the modification of the clay character from hydrophilic to organophilic, which improves the dispersion of the clay into the hydrophobic polymeric matrix. To this regard, the two most important methods consist in: (i) replacing the metallic cations originally present on the clay surface with large organic cations; and (ii) grafting alkoxysilanes onto the clay exploiting the reactivity of silanol groups present on the edges of the clay platelets. As far as the first method is concerned, the properties of the obtained organophilic clay are strongly dependent on the experimental method used for ion exchange, as well as on the type of clay and organic cation employed. This method was extensively applied, and several types of cations, mostly dialkyldimethyl- or alkyltrimethyl-ammonium ones, were used.4-11 Differently from its extensive employment in silica treatments (examples are reported in refs 12-15), the grafting of alkox* Corresponding author. Phone: +39-050-2219289. Fax: +39-0502219260. E-mail:
[email protected]. (1) Alexandre, M.; Dubois, P. Mater. Sci. Eng. 2000, 28, 1-63. (2) Sinha Ray, S.; Okamoto, M. Prog. Polym. Sci. 2003, 28, 1539-1641. (3) Ahmadi, S. J.; Huang, Y. D.; Li, W. J. Mater. Sci. 2004, 1919-1925. (4) Vaia, R. A.; Teukolsky, R. K.; Giannelis, E. P. Chem. Mater. 1994, 6, 1017-1022. (5) Leach, E. S. H.; Hopkinson, A.; Franklin, K.; van Duijneveldt, J. S. Langmuir 2005, 21, 3821-3830. (6) Breakwell, I. K.; Homer, J.; Lawrence, M. A. M.; McWhinnie, W. R. Polyhedron 1995, 14, 2511-2518. (7) Carrado, K. A. Appl. Clay Sci. 2000, 17, 1-23. (8) Xi, Y.; Ding, Z.; He, H.; Frost, R. L. Spectrochim. Acta, Part A 2005, 61, 515-525. (9) Hasegawa, N.; Kawasumi, M.; Kato, M.; Usuki, A.; Okada, A. J. Appl. Polym. Sci. 1998, 67, 87-92. (10) Mu¨ller, R.; Hrobarikova, J.; Calberg, C.; Je´roˆme, R.; Grandjean, J. Langmuir 2004, 20, 2982-2985. (11) Kubies, D.; Je´roˆme, R.; Grandjean, J. Langmuir 2002, 18, 6159-6163. (12) Bauer, F.; Ernst, H.; Decker, U.; Findeisen, M.; Gla¨sel, H.-J.; Langguth, H.; Hartmann, E.; Mehnert, R.; Peuker, C. Macromol. Chem. Phys. 2000, 201, 2654-2659.
ysilanes has been so far scarcely applied to layered silicates modifications;16-23 in particular, as far as smectites are concerned, this is mainly due to their usually low edge area to surface ratio, which in turn results in a small number of reactive edge silanols.19 Moreover, in both exchanging cations and alkoxysilanes, the presence of reactive groups is important for possible copolymerization reactions in the subsequent preparation of polymerclay nanocomposites. The clay here studied, Laponite, is a synthetic layered silicate similar in structure and composition to the natural smectite hectorite. Each Laponite platelet is composed of three sheets: a central sheet of magnesium ions in octahedral coordination with oxygen anions and hydroxyl groups and two outer tetrahedral silica sheets. The isomorphic substitution of some magnesium cations with lithium cations in the central sheet, as well as the presence of some vacant positions, gives rise to a partial negative charge, which is balanced by sodium cations absorbed on the platelets surface; the reported Laponite empirical formula is Na+0.7[Si8Mg5.5Li0.3]O20(OH)4]0.7-.24,25 The finite dimensions of the clay platelets are responsible for the occurrence of silanol groups at the “broken edges” of the sheets.26,27 The average dimensions of the Laponite sheets (diameter of 25 nm and height (13) Nishiyama, N.; Horie, K.; Asakura, T. J. Colloid Interface Sci. 1989, 129, 113-119. (14) Ek, S.; Iiskola, E. I.; Niinisto¨, L.; Vaittinen, J.; Pakkanen, T. T.; Root, A. J. Phys. Chem. B 2004, 108, 11454-11463. (15) Borsacchi, S.; Geppi, M.; Veracini, C. A.; Fallani, F.; Ricci, L.; Ruggeri, G. J. Mater. Chem. 2006, 16, 4581-4591. (16) de Prado, L. A. S.; Karthikeyan, C. S.; Schulte, K.; Nunes, S. P.; de Torriani, I. L. J. Non-Cryst. Solids 2005, 351, 970-975. (17) Herrera, N. N.; Letoffe, J.-M.; Putaux, J.-L.; David, L.; Bourgeat-Lami, E. Langmuir 2004, 20, 1564-1571. (18) Herrera, N. N.; Letoffe, J.-M.; Reymond, J.-P.; Bourgeat-Lami, E. J. Mater. Chem. 2005, 15, 863-871. (19) Wheeler, P. A.; Wang, J.; Baker, J.; Mathias, L. J. Chem. Mater. 2005, 17, 3012-3018. (20) Isoda, K.; Kuroda, K.; Ogawa, M. Chem. Mater. 2000, 12, 1702-1707. (21) He, H.; Duchet, J.; Galy, J.; Gerard, J.-F. J. Colloid Interface Sci. 2005, 288, 171-176. (22) Seckin, T.; Gultek, A.; Icduygu, M. G.; Onal, Y. J. Appl. Polym. Sci. 2002, 84, 164-171. (23) Bourlinos, A. B.; Jiang, D. D.; Giannelis, E. P. Chem. Mater. 2004, 16, 2404-2410. (24) Willenbacher, N. J. Colloid Interface Sci. 1996, 182, 501-510. (25) www.laponite.com (Rockwood Specialties). (26) Searle, A. B.; Grimshaw, R. W. The Chemistry and Physics of Clays and Other Ceramic Materials, 3rd ed.; Interscience Publishers: New York, 1959.
10.1021/la063040a CCC: $37.00 © 2007 American Chemical Society Published on Web 03/03/2007
3954 Langmuir, Vol. 23, No. 7, 2007
Figure 1. Chemical structures and atom labeling of 2C18 (top) and TSPM (bottom).
of 0.92 nm)25 confer to Laponite a quite high edge area to surface ratio (0.07), which makes it suitable for edge-modification by alkoxysilanes.16-19,23 In the present work, Laponite was subjected to both surface modification by cation exchange and edge modification by alkoxysilanes grafting. This double treatment, which so far has not been extensively exploited, appears important for obtaining multifunctional nanomaterials, as well as for achieving a better dispersion of the clay in nonpolar polymeric matrices.19,28 The double chain surfactant dimethyldioctadecylammonium chloride (2C18) was used for cation exchange, whereas the alkoxysilane employed (3-(trimethoxysilyl)propyl methacrylate, TSPM) contains a polymerizable group in the perspective of preparing nanocomposites with polyolefines. Laponite was studied in its untreated hydrophilic Na+-form (Laponite RD), and in the two organophilic forms obtained by ion exchange with 2C18 (Laponite2C18) and by the double treatment with 2C18 and TSPM (LaponiteTSPM). These three samples were characterized by X-ray diffraction, Fourier transform infrared spectroscopy, scanning electron microscopy, and thermogravimetric analysis. Moreover, solid-state nuclear magnetic resonance, which was already revealed to be extremely powerful in the study of the species interacting with a clay surface,29 was extensively applied. In particular, 29Si, 13C, and 1H high-resolution techniques, as well as 1H wide-line free induction decay analysis, were applied on the three Laponite samples. The combined interpretation of the different experimental results allowed us to obtain detailed information on the effects of the two Laponite organic modifications on the structural and dynamic properties of both the clay and the species interacting with it. Materials and Methods Samples. Laponite RD (Rockwood Additives, UK) is a synthetic clay similar in structure and composition to natural hectorite, belonging to the smectite group, with an empirical formula of Na+0.7[Si8Mg5.5Li0.3]O20(OH)4]0.7-. Laponite RD was modified by cation exchange with dimethyldioctadecylammonium chloride (2C18, Fluka) and then by grafting with 3-(trimethoxysilyl)propyl methacrylate (TSPM, Fluka) (see Figure 1). Synthesis of Laponite-2C18. In a three-neck flask equipped with a magnetic stirrer, thermometer, dropping funnel, and dropping cooling, 2 l of water was added. 10.0 g of Laponite RD was added (27) Worrall, W. E. Clays and Ceramic Raw Materials; Applied Science Publishers: London, 1975. (28) Wang, J.; Wheeler, P. A.; Jarrett, W. L.; Mathias, L. J. Polym. Prepr. 2005, 46, 564-565. (29) Urbanczyk, L.; Hrobarikova, J.; Calberg, C.; Je´rome, R.; Grandjean, J. Langmuir 2006, 22, 4818-4824.
Borsacchi et al. at a temperature of 60 °C by constantly and vigorously stirring the solution. The suspension was further stirred for 1 h, and then 500 cm3 of aqueous saturated solution (4.6 g/L) of dimethyldioctadecylammonium chloride (2C18) was added. The mixture was stirred for 21 h, filtered under vacuum, and washed with three portions of water (3 × 1 l) to remove all excess of 2C18. The so obtained organoclay (Laponite-2C18) was freeze-dried and characterized. Synthesis of Laponite-TSPM. In a three-neck flask equipped with a magnetic stirrer, thermometer, dropping funnel, and dropping cooling, 60 cm3 of a 95/5 vol/vol ethanol/water solution and 9 cm3 of 3-(trimethoxysilyl)propyl methacrylate (TSPM) were introduced. This mixture was stirred for 1 h to favor the hydrolysis of TSPM. 4.9 g of Laponite-2C18 was added, and the suspension was stirred at room temperature for 4 days. It was then filtered under vacuum and extracted with chloroform in Soxleth for 8 h to remove unreacted TSPM and possible polymethacrylate. The modified organoclay so obtained (Laponite-TSPM) was collected, dried, and characterized. Characterization Methods. FT-IR spectra were recorded on a Perkin-Elmer FT-IR 1760-X spectrometer on powder-pressed KBr pellets. Low angle X-ray diffraction measurements were carried out using a D 500/501 Siemens Kristalloflex 810 diffractometer with a CuK R source (λ ) 0.15406 nm). Scanning electron microscopy (SEM) was performed on a Jeol JSM T-300 microscope. Thermogravimetric analyses (TGA) were carried out on a Mettler Toledo TGA/STDA851 at a heating rate of 10 °C/min from 25 to 600 °C under nitrogen flow. Solid-State NMR. 29Si, 13C, and 1H magic angle spinning (MAS) NMR spectra were recorded on a double-channel Varian InfinityPlus 400 spectrometer, equipped with a 7.5 mm cross-polarization (CP)MAS probehead, working at 399.89 MHz for proton, at 79.44 MHz for silicon-29, and at 100.75 MHz for carbon-13. 1H, 13C, and 29Si 90° pulse lengths were always between 4.0 and 5.0 µs. All of the 13C and 29Si NMR spectra were recorded under high-power proton decoupling conditions. 29Si CP/MAS spectra were recorded for all of the samples with a MAS frequency of 6 kHz, a recycle delay of 3 s, and a contact time of 3 ms. To obtain quantitative results,19 29Si-single pulse excitation (SPE)/MAS spectra were recorded for samples Laponite-2C18 and Laponite-TSPM with a recycle delay of 180 s and a MAS frequency of 6 kHz. The 2D 29Si-1H FSLGHETCOR technique30 was applied using a contact time of 400 µs, 64-144 rows, and a MAS frequency of 6 kHz. 13C CP/MAS spectra were recorded for Laponite-2C18 and Laponite-TSPM with a MAS frequency of 6 kHz, a recycle delay of 3 s, and a contact time of 1 ms. The 1H MAS spectra were acquired at a MAS frequency of 6 kHz with a recycle delay of 3 s. Low-resolution 1H experiments were carried out on a singlechannel Varian XL-100 spectrometer interfaced with a Stelar DSNMR acquisition system and equipped with a 5 mm probehead. These measurements were performed on-resonance on static samples at a frequency of 25.00 MHz. Free induction decays were recorded for all of the samples using the solid-echo pulse sequence with a 90° pulse length of 2.6 µs, a recycle delay of 2 s, an echo delay of 13 µs, a dwell time of 1 µs, acquiring 8192 points. All of the experiments were performed at a temperature of 25 °C, controlled to within (0.1 °C.
Results and Discussion Characterization of Laponite-2C18. The presence of the ammonium cation in Laponite-2C18 could be identified by comparing the FT-IR spectra recorded on Laponite RD, Laponite2C18, and 2C18 (see Figure 2). The spectrum of Laponite-2C18 (Figure 2d) revealed three additional bands (2929, 2851, and 1468 cm-1) with respect to that of Laponite RD (Figure 2c), ascribable to C-H asymmetric and symmetric stretching and bending of 2C18 alkyl chains. The band at 3687 cm-1, arising from O-H stretchings of SiOH and MgOH groups,18 was better resolved in the spectrum of Laponite-2C18 due to the decrease (30) Van Rossum, B. J.; Forster, H.; De Groot, H. J. M. J. Magn. Reson. 1997, 124, 516-519.
Solid-State NMR Study of Modified Laponites
Figure 2. Infrared spectra of (a) 2C18, (b) TSPM, (c) Laponite RD, (d) Laponite-2C18, and (e) Laponite-TSPM.
Figure 3. SEM images of (a) Laponite-2C18 and (b) Laponite RD.
in the intensity of the partially overlapped broad peak at about 3500 cm-1, ascribable to the O-H stretching of water molecules,8 in agreement with the organophilic character of Laponite-2C18. From a morphological SEM analysis, a non-homogeneous distribution of the particle dimensions was observed for both Laponite RD and Laponite-2C18 (see Figure 3), but the particles of the latter presented a much rougher surface as a result of the treatment with 2C18. The thermogravimetric analysis of Laponite RD (see Figure 4) showed a weight loss with an onset temperature of about 72 °C, due to water weakly bonded to the clay. In the case of Laponite-
Langmuir, Vol. 23, No. 7, 2007 3955
Figure 4. Thermograms of Laponite RD (-), Laponite-2C18 (- - -), and Laponite-TSPM (- ‚ -).
2C18, the water content was strongly reduced, while a degradation step with an onset temperature of about 340 °C was observed, ascribable to the alkyl ammonium salt present in the modified clay, whose amount was estimated to be 24.9 wt %. Characterization of Laponite-TSPM. The comparison of the FT-IR spectra of Laponite-TSPM, Laponite-2C18, and TSPM (see Figure 2) confirmed the occurrence of TSPM grafting to the clay. In the spectrum of Laponite-TSPM (Figure 2e), the band at 1632 cm-1 is ascribable to the CdC stretching vibration, even though it overlapped with a band already present in bare Laponite. Moreover, a band, attributed to carbonyl groups of the silane moiety forming hydrogen bonds with hydroxyls present on the surface of the clay platelets, was observed at 1708 cm-1, shifted with respect to that present in the spectrum of pure TSPM (1720 cm-1), arising from nonbonded carbonyls.18 The thermogravimetric analysis of Laponite-TSPM (Figure 4) did not show the degradation step observed for pure TSPM (onset temperature of 99 °C, not shown), confirming that unreacted TSPM was completely removed. The degradation step ascribable to TSPM was clearly overlapped with that due to the alkyl ammonium cation. From the comparison with the dispersion curve of Laponite-2C18, the amount of TSPM grafted to the clay was estimated to be 1.7 wt %. X-ray. To estimate the influence of the two organic modifications on the clay inter-platelet distances, we performed an X-ray analysis of the three Laponite samples. In Figure 5, the X-ray diffractograms of Laponite RD, Laponite-2C18, and LaponiteTSPM are shown. The main peak in the diffraction patterns is attributable to the basal spacing d; it occurred at 2θ angles of 5.54°, 4.15°, and 4.92°, corresponding to d values of 15.9, 21.3, and 17.9 Å for Laponite RD, Laponite-2C18, and Laponite-TSPM, respectively. Passing from Laponite RD to Laponite-2C18, an increase of the basal spacing was observed, in agreement with the occurrence of the intercalation of the quaternary ammonium cations between the Laponite platelets. The 21.3 Å value found for Laponite2C18 is in agreement with literature data,11 and, considering the 9.6 Å intrinsic thickness of a single Laponite platelet, the effective gallery height available for the intercalated surfactant could be estimated to be 11.7 Å. Even though, to the best of our knowledge, previous X-ray diffraction data for double chain surfactants intercalated in Laponite are not available, for an analogous dioctadecyldimethyl ammonium salt in the crystalline state, where the alkyl chains are arranged in the all-trans conformation, the
3956 Langmuir, Vol. 23, No. 7, 2007
Borsacchi et al.
Figure 5. X-ray diffractograms of Laponite RD (-), Laponite-2C18 (- - -), and Laponite-TSPM (- ‚ -).
longest molecular axis length was reported to be 26 Å.31,32 Therefore, it is clear that in this case a perpendicular orientation of the surfactant with respect to the clay platelet surface must be ruled out, as already found for single chain ammonium surfactants.10,11 Moreover, from the value of 24 Å2 reported for the cross-sectional area of a rigidly packed all-trans alkyl chain,11 a maximum diameter of 11.2 Å could be estimated for the doublechain surfactant here employed, which is only slightly smaller than the 11.7 Å value found for the gallery height, thus suggesting an arrangement of the cation alkyl chains parallel to the clay platelet surface, and ruling out the possibility of a bilayer arrangement.33 After the TSPM grafting reaction, the interlamellar distance decreased to values that appear incompatible with an ordered arrangement of cation alkyl chains in the galleries between stacked clay platelets: an interpretation of this result will be given in the following, after the analysis of solid-state NMR results. At the present stage, it must be noticed that the diffraction peak investigated was characterized by a different distribution in the different samples, which was maximum in Laponite RD, and minimum in Laponite-2C18, indicating that the degree of disorder in the platelet arrangement decreased following the 2C18 cation exchange and increased again as a consequence of the TSPM functionalization. Solid-State NMR. 29Si. Before discussing the 29Si NMR spectra, it is useful to remind that silicon nuclei bonded to one, two, three, or four oxygen atoms are conventionally indicated as Mn, Dn, Tn, Qn, respectively, where n is the number of oxygen atoms further bonded to another silicon atom. In Figure 6, the 29Si CP/MAS spectra of Laponite RD, Laponite-2C , and 18 Laponite-TSPM are shown. All of the spectra were clearly dominated by an intense signal, resonating at -94.7 ppm, which was straightforwardly assigned to the Q3 Si(OMg)(OSi)3 silicon nuclei forming the two silica tetrahedral layers of the Laponite platelets. This signal was substantially the same in the three spectra, indicating that, from the NMR standpoint, the abundant Q3 silicon nuclei were almost unaffected by the two subsequent Laponite treatments. The only difference was a larger line width observed in Laponite RD, compatible with the more disordered platelets arrangement of this sample, as detected from X-ray (31) Osman, M. A.; Ernst, M.; Meier, B. H.; Suter, U. W. J. Phys. Chem. B 2002, 106, 653-662. (32) Okuyama, K.; Soboi, Y.; Iijima, N.; Hirabayashi, K. Bull. Chem. Soc. Jpn. 1988, 61, 1485-1490. (33) Wen, X.; He, H.; Zhu, J.; Jun, Y.; Ye, C.; Deng, F. J. Colloid Interface Sci. 2006, 299, 754-760.
Figure 6. 29Si CP/MAS spectra of (a) Laponite-RD, (b) Laponite2C18, and (c) Laponite-TSPM, acquired at a MAS frequency of 6 kHz. In the inset, a vertical expansion of the spectrum of LaponiteTSPM is reported, to show the T silicon signals.
diffraction, which can be ascribed to the presence of water. In all of the three spectra, an additional small and asymmetric signal was observed at about -85.5 ppm, ascribable to Q2 silicon nuclei, whose assignment will be more extensively discussed in the following. As expected, in passing from Laponite RD to Laponite-2C18, no additional 29Si signals were detected. On the contrary, in the spectrum of Laponite-TSPM, three new signals, resonating at about -47, -56, and -66 ppm, were observed (see inset of Figure 6), which were assigned to T1, T2, and T3 TSPM silicon nuclei, respectively. The presence of these T silicon signals confirmed the occurrence of the condensation reaction between Laponite-2C18 and TSPM, in agreement with the IR and TGA results. On the other hand, the absence of a T0 signal indicated the complete removal of unreacted TSPM molecules. Considering that a tridental anchoring of TSPM molecules to Laponite platelets is substantially unlikely, the presence of T3 and, in part, T2 signals has to be ascribed to condensation reactions occurring among different TSPM molecules. These gave rise to the formation of a siloxane network connecting different clay platelets. The presence of these siloxane oligomers had been already suggested in the case of bare Laponite functionalized with trialkoxysilanes,17 as well as for TSPM grafted silica.12,15 Moreover, the observation of the T1 signal, differently from what was previously reported for alkoxysilane-functionalized Laponites,17 indicated the presence of a non-negligible amount of mono-anchored TSPM molecules. As above-mentioned, in all of the spectra an asymmetric signal was observed at about -85.5 ppm, different in shape and intensity for the three Laponites. In the spectrum of Laponite RD (Figure 6a), only a broad signal was observed: in the literature, this was assigned to Q2 silicon nuclei, often referred to as uncondensed silicons, which are considered to be mostly present as Si(OMg)(OSi)2(OH) silanols, especially occurring at the clay particle edges.17-19,28,34 This signal had been previously found to decrease as a consequence of the reaction of Laponite RD with alkoxysi(34) Mandair, A.-P. S.; Michael, P. J.; McWhinnie, W. R. Polyhedron 1990, 9, 517-525.
Solid-State NMR Study of Modified Laponites
Figure 7. 29Si SPE/MAS spectra of (a) Laponite-2C18 and (b) Laponite-TSPM, acquired at a MAS frequency of 6 kHz.
lanes.17-19,28 Here, after the first treatment with 2C18, two components of the Q2 signal, with chemical shifts of about -84.4 and -86.5 ppm, became clearly distinguishable (see Figure 6b). After the functionalization reaction with TSPM, the intensity of the component at -84.4 ppm substantially decreased (see Figure 6c). This strongly suggests that the two signals arise from two different Q2 silicon species present in Laponite-2C18, only one of them being involved in the reaction with TSPM. To clarify this point and obtain more quantitative results, we acquired 29Si SPE/MAS spectra on Laponite-2C18 and Laponite-TSPM (see Figure 7). The deconvolution of these spectra allowed us to determine the intensities of the signals resonating at about -94.7 (Q3), -86.5, and -84.4 ppm (Q2), which were, respectively, 82%, 7%, and 11% in Laponite-2C18, and 91%, 6%, and 3% in Laponite-TSPM, where the percentages refer to the sum of the Q signals. On one side, this indicated that about 9% of Laponite silicons transformed from Q2 to Q3 as a result of condensation processes following the treatment with TSPM. On the other side, the Q2 silanols involved in such processes were almost exclusively those contributing to the signal resonating at about -84.4 ppm, whose intensity decreased from 11% to 3%. It is therefore reasonable to assign the signal at -84.4 ppm to the silanol silicon nuclei present on the Laponite platelet edges and involved in the reaction with the TSPM. Instead, the signal at -86.5 ppm should arise from uncondensed silicon nuclei not accessible to TSPM, either present as surface silanols located in the inter-platelets galleries or bonded to incompletely coordinated lattice oxygen anions. The difficulty in resolving these two signals in Laponite RD is probably due to the already discussed higher structural disorder induced by the presence of water, resulting in broader 29Si peaks. At last, it has to be noticed that the 29Si TSPM (T) signals were hardly detectable in the SPE/MAS spectrum of Laponite-TSPM; their total integral is surely much less than the increase of the Q3 signal observed as a consequence of the TSPM treatment. This strong discrepancy suggests that “self-condensation” reactions occurred among clay edge silanols. 1H MAS. In Figure 8, the 1H MAS spectra of the three samples are reported. In the spectrum of Laponite RD (Figure 8a), two signals were observed, resonating at 0.4 and 4.4 ppm. The first one can be assigned to the protons of the MgOH groups located in the octahedral layers of Laponite platelets; a similar signal, resonating at 0.35 ppm, was previously observed for a synthetic hectorite.35 The most intense signal at 4.4 ppm is ascribable to water, either (35) Hou, S.; Beyer, F. L.; Schmidt-Rohr, K. Solid State Nucl. Magn. Reson. 2002, 22, 110-127.
Langmuir, Vol. 23, No. 7, 2007 3957
Figure 8. 1H MAS spectra of (a) Laponite-RD, (b) Laponite-2C18, and (c) Laponite-TSPM, acquired at a MAS frequency of 6 kHz.
adsorbed on the outer platelets surface or intercalated between them. The signal due to clay silanol protons, involved in hydrogen bonds with water, was undistinguishable from that of water.36-38 A broad line was present in the 1H MAS spectra of Laponite2C18 and Laponite-TSPM, only partially resolved in spinning sidebands, resulting from an incomplete MAS averaging of the homonuclear dipolar coupling, arising from the most tightly dipolarly coupled protons of the samples, mostly due to the 2C18 chains. Apart from this broad line, which in the spectra shown (Figure 8b and c) was removed by data processing, the 1H MAS spectra of Laponite-2C18 (Figure 8b) and Laponite-TSPM (Figure 8c) were characterized by three main narrow signals, heavily superimposed, centered at about 0.4, 1.1, and 1.5 ppm. These narrow lines arose from protons experiencing minor dipolar couplings, efficiently averaged out by MAS, and could be mostly assigned to MgOH, 2C18 chains in mobile environments, and clay silanols, respectively. The signal at 4.4 ppm was no longer present in both spectra, according to the TGA results, which indicated a strong reduction in the water amount as a consequence of the ion exchange process. The large shift experienced by the clay silanol protons passing from Laponite RD to Laponite-2C18 was expected as a consequence of water removal.36,37 In the 1H MAS spectrum of Laponite-TSPM, the contributions of the TSPM protons were not clearly recognized, apart from two small and quite broad peaks resonating at 5.6 and 6.3 ppm, arising from alchenilic methylene protons, already observed in the 1H MAS spectrum of TSPM-grafted silica.12,15 This is probably due to the small TSPM concentration, to partial line overlaps with more intense signals, as well as to their main contribution to the broadest spectral line, in agreement with the rigid environment experienced by TSPM molecules after condensation. 1H-29Si HETCOR. In the attempt of obtaining additional information on the different silicon sites present in the three samples, 1H-29Si HETCOR spectra were acquired (see Figure 9). This bidimensional technique allows the signals of dipolarly coupled 1H and 29Si nuclei to be correlated. The intensity of the cross-peaks in the HETCOR map is higher for strongly dipolarly coupled nuclei: therefore, the most significant correlations involve 1H and 29Si nuclei that are both spatially close and in a quite rigid environment. In the maps of all three samples, the most intense cross-peak correlated Q3 silicons and MgOH protons, as already (36) Bronnimann, C. E.; Zeigler, R. C.; Maciel, G. E. J. Am. Chem. Soc. 1988, 110, 2023-2026. (37) Kinney, D. R.; Chuang, I.-S.; Maciel, G. E. J. Am. Chem. Soc. 1993, 115, 6786-6794. (38) Borsacchi, S.; Geppi, M.; Iuliano, A.; Veracini, C. A. Solid State Nucl. Magn. Reson. 2005, 28, 193-203.
3958 Langmuir, Vol. 23, No. 7, 2007
Figure 9. 2D 1H-29Si HETCOR maps of (a) Laponite RD and (b) Laponite-2C18, acquired at a MAS frequency of 6 kHz. The correlation peaks, labeled by roman numbers in the map, were sampled at different threshold values for the sake of clarity: in (a), the maximum peak intensity (in au) is 100 in I, 7 in II, and 4.4 in III; in (b), it is 100 in I, and 8.1 in II.
observed for a synthetic hectorite.35 The abundance of both the Q3 silicons and the MgOH protons, together with their spatial proximity and restricted mobility, contributed to the observed strong dipolar interaction. In the Laponite RD map (Figure 9a), another cross-peak correlated the broad Q2 silicons signal (at around -85.5 ppm) with the proton signal resonating at 4.4 ppm, in agreement with their assignment to the clay silanol silicons, and water and hydrogen-bonded silanol protons, respectively. A cross-peak correlating clay silanol silicons and MgOH protons was also present. In both Laponite-2C18 (Figure 9b) and Laponite-TSPM (data not shown) maps, the two Q2 silicon signals (-84.4 and -86.5 ppm) were correlated with proton signals resonating at two slightly different chemical shift values. The intrinsic problems present in the processing of the HETCOR data in the proton dimension, connected with the absence of certain reference frequencies and FSLG scaling factors, prevented us from establishing precise values for the chemical shifts of these two proton signals without performing any a priori assumption. However, it seems reasonable that the signal at higher chemical shift, correlated with the 29Si signal at -84.4 ppm, could correspond to silanol protons, and the other, correlated with the 29Si signal at -86.5 ppm, to 2C18 chain protons. This would be in agreement with the interpretation of both 29Si and 1H monodimensional spectra and, in particular, would strongly support the assignment of the two types of 29Si Q2 signals to edge silanol silicons and uncondensed unaccessible silicons. 1H FID Analysis. 1H free induction decay (FID) analysis is a useful NMR methodology, which allows motionally distinct
Borsacchi et al.
domains present in a sample to be characterized and quantified.39 The experimental 1H FID is fitted with a suitable linear combination of several analytical functions, here chosen among Gaussian, exponential, and Weibullian. The fit parameters were the 1H transverse relaxation time (T2), the weight percentage (wt %) of each function, and, in the case of the Weibullian function, the exponent n, ranging from 1 to 2, where the two limit values correspond to exponential and Gaussian functions, respectively. The proton transverse relaxation time is substantially determined by the extent of residual homonuclear dipolar interactions in which the protons are involved, being shorter when dipolar interactions are stronger; the residual dipolar interactions are reduced (giving rise to an increase in T2 values) by molecular motions with characteristic frequencies larger than tens of kHz, as well as by long inter-proton distances. The usefulness of this methodology consists in the possibility of establishing a rough correspondence between the different functions used to reproduce the experimental FID and motionally distinct domains of the sample. The results of the 1H FID analysis performed on the three Laponite samples are reported in Table 1, and an example of FID analysis is given in Figure 10. The 1H FID of Laponite RD was well reproduced by a linear combination of a Gaussian function, with a weight of 18.5% and a T2 of 73.6 µs, and an exponential function with a weight of 81.5% and a T2 about 1 order of magnitude longer. In this sample, at least two different “sources” of protons were present: the physisorbed water and the clay hydroxyl groups (MgOH and SiOH). It seems reasonable to associate the Gaussian FID component to the clay hydroxyl protons (mainly MgOH protons) and the exponential one to the water protons and to the water-hydrogen-bonded silanol protons. Indeed, MgOH protons, located inside the octahedral layers of the clay platelets, are expected to be in a more motionally restricted environment. This interpretation is in excellent agreement with the integrals of the two peaks observed in the 1H MAS spectrum (82% for the peak at 4.4 ppm and 18% for that at 0.4 ppm). As far as Laponite-2C18 and Laponite-TSPM are concerned, their 1H FIDs appeared quite similar between themselves and substantially different from that of Laponite RD. They were both well reproduced by the same set of three functions: a short-T2 (∼16 µs) Gaussian, an intermediate Weibullian with a T2 of about 50 µs and n ) 1.12, both with a 40-50 wt %, and a much smaller exponential with a longer T2 (about 200 µs) (see Table 1). First, it is worth noticing that in both Laponite-2C18 and Laponite-TSPM FIDs the long T2 (600 µs) exponential function was no longer present, confirming that this component has to be mainly ascribed to water protons. Because in both of these samples, most of the protons belong to the 2C18 cations intercalated among the clay platelets, they must be mainly responsible for both Gaussian and Weibullian components. Clay MgOH and SiOH protons (the latter no longer hydrogen bonded) should mainly contribute to the Weibullian and exponential components, respectively. In agreement with this assignment, the weight percentage of the exponential function decreased by 2.4% after the reaction with TSPM. In Laponite-TSPM, the slight transfer of population from Weibullian to Gaussian functions can be probably interpreted as due to the contribution of TSPM protons, which, therefore, would be in a restricted motional regime, in agreement with the formation of bridges between clay platelets by means of oligomeric structures. 13C. With the aim of gaining some additional insights about the arrangement and the conformational properties of the (39) Borsacchi, S.; Cappellozza, S.; Catalano, D.; Geppi, M.; Ierardi, V. Biomacromolecules 2006, 7, 1266-1273.
Solid-State NMR Study of Modified Laponites
Langmuir, Vol. 23, No. 7, 2007 3959
Table 1. 1H FID Analysis Results for Laponite RD, Laponite-2C18, and Laponite-TSPMa first component
second component
third component
sample
function
T2 (µs)
wt %
function
T2 (µs)
wt %
function
T2 (µs)
wt %
Laponite RD Laponite-2C18 Laponite-TSPM
G G G
73.6 15.8 16.1
18.5 42.1 49.2
E W W
616.8 51.1 47.3
81.5 49.4 44.7
E E
181.5 209.8
8.5 6.1
a
“G”, “W”, and “E” refer to Gaussian, Weibullian, and exponential functions, respectively.
intercalated alkylammonium ions and the grafted TSPM molecules, 13C CP/MAS spectra on both Laponite-2C18 and LaponiteTSPM samples were acquired (Figure 11). Solid-state 13C NMR spectroscopy was successfully employed in the investigation of the conformational properties of long chain organic cations, variously interacting with inorganic substrates. For instance, some recent studies concerned the organic chains arrangement of alkylammonium ions either forming self-assembled monolayers
Figure 10. 1H FID analysis of Laponite-TSPM. (a) Experimental and best-fit calculated FIDs. (b) Best-fit Gaussian (-), Weibullian (- ‚ -), and exponential (- - -) functions. Only the first 500 points of the FID are shown.
Figure 11. 13C CP/MAS spectra of (a) Laponite-2C18 and (b) Laponite-TSPM, acquired at a MAS frequency of 6 kHz.
on partially delaminated mica particles surface,31 confined in MCM-41 channels,40 or intercalated in several smectites.10,11 The assignment of the Laponite-2C18 13C CP/MAS spectrum was quite straightforward and in agreement with previous literature data.31 C1 and C2, resonating in the regions 50-55 and 65-70 ppm, respectively, showed broad lines and an asymmetric doublet shape, which must be ascribed to the 14N quadrupolar effect on the 13C-14N dipolar interaction.41 An additional contribution to the line broadening of these signals could also arise from the reduced mobility of these groups, due to their proximity to the polar head of the cation interacting with the clay surface.31,40 C18 and C19 resonated at 24.5 and 15.2 ppm, respectively. The chain terminal methyl carbons (C19) signal showed a low chemical shift value, if compared to that observed for similar alkylammonium salts in bulk form (15 ppm instead of 18 ppm): this is a common feature of the alkylammonium ions intercalated in clays or, more generally interacting with inorganic substrates,10,11,31,40 as well as of alkylic monolayers with methyl groups exposed at the organic-air interface.42-44 The C3-C17 methylene carbons gave rise to two distinguishable signals, resonating at 33 and 31 ppm, assignable to chains in the rigid all-trans and in the disordered trans-gauche conformation, respectively, due to the well-known γ-gauche effect. In the 13C CP/MAS spectrum of Laponite-TSPM, all of the cation signals resonated at approximately the same chemical shifts observed in the spectrum of Laponite-2C18; some additional weak signals ascribable to TSPM carbons, and therefore confirming the occurrence of the silylation reaction, could be detected at about 9, 18, 125, 137, and 169 ppm, corresponding to C2, C7, C8, C6, and C5.17 However, the most significant and interesting effect of the silylation reaction of Laponite-2C18 concerned the intensity ratio of the two signals, resonating at 33 and 31 ppm, arising from the C3-C17 carbons of the cation alkyl chains. For Laponite-2C18, their ratio was almost 1:1 in the CP/MAS spectrum recorded with a contact time of 1 ms, while in the corresponding spectrum of Laponite-TSPM the signal due to disordered trans-gauche conformations (31 ppm) became definitely more intense than that ascribed to the ordered all-trans conformation (33 ppm). Even if absolute quantitative considerations could not be drawn from the 13C CP/MAS spectra, the comparison between the spectra of the two samples, acquired at the same contact time, reliably indicated that in Laponite-TSPM a much larger fraction of surfactant chains experienced fast transgauche interconformational jumps. The coexistence, in both of the samples, of these two chain conformations is not surprising: many studies were reported concerning the distribution between the ordered all-trans and the disordered trans-gauche conformations for alkyl chains of surfactant cations intercalated in different (40) Simonutti, R.; Comotti, A.; Bracco, S.; Sozzani, P. Chem. Mater. 2001, 13, 771-777. (41) Naito, A.; Ganapathy, S.; McDowell, C. A. J. Chem. Phys. 1981, 74, 5393-5397. (42) Osman, M. A.; Suter, U. W. Chem. Mater. 2002, 14, 4408-4415. (43) Gao, W.; Dickinson, L.; Grozinger, C.; Morin, F. G.; Reven, L. Langmuir 1996, 12, 6429-6435. (44) Pawsey, S.; Yach, K.; Halla, J.; Reven, L. Langmuir 2000, 16, 32943303.
3960 Langmuir, Vol. 23, No. 7, 2007
Borsacchi et al.
Figure 12. Schematic representation of the platelets arrangement in Laponite RD (left) and after treatments with 2C18 (middle) and TSPM (right), as hypothesized on the basis of solid-state NMR and X-ray diffraction results. R′ indicates further TSPM units in the oligomeric chain.
kinds of clays, as well as dispersed and confined in different inorganic matrices. Despite the difficulty in establishing general rules, due to the strong dependence on the system features and their wide variety, the alkyl chains conformational properties turned out to be strongly dependent on temperature, surfactant properties (number and length of the alkyl chains), as well as on clay features (platelets size and interlayer charge).10,11,31,40,45 Here, the small size of the Laponite platelets (as compared to other clays of the smectite group), as well as the low charge density of the interlayer surface,10 favored a quite disordered clay structure, already highlighted by X-ray results, and a quite scarce constraining of the intercalated surfactant chains. A picture of the Laponite-TSPM system could be tentatively drawn considering: (i) the increase of the disordered fraction of cation chains after the TSPM grafting reaction, observed by 13C NMR spectra; (ii) the decrease in the average inter-platelet distance and the increase of disorder, observed by X-ray measurements; and (iii) the formation of TSPM oligomers connecting different platelets, observed by 29Si NMR spectra. The TSPM grafting reaction on the edges of the Laponite platelets appeared to be nonhomogeneous, and in particular the formation of siloxane oligomers connecting different clay platelets should give rise to a quite disordered disks arrangement, in which they were on average pushed closer, and, at the same time, partly twisted one with respect to the other (see Figure 12). This hypothesis, in agreement with the tendency of Laponite to establish edge-to-face interactions, would also explain why a noticeable fraction of surfactant molecules in Laponite-TSPM was less constrained in the inter-platelets spaces, therefore experiencing fast trans-gauche interconformational jumps.
Conclusions Two modified Laponites were prepared either using the usual ion exchange procedure for replacing Na+ cations with large organic ammonium cations (Laponite-2C18) or combining this (45) Grandjean, J.; Bujda´k, J.; Komadel, P. Clay Miner. 2003, 38, 367-373.
procedure with a subsequent silylation reaction with alkoxysilanes (Laponite-TSPM). These samples were characterized by combining several standard techniques (TGA, FT-IR, X-ray diffraction, SEM) with a wide range of multinuclear (29Si, 13C, 1H) solid-state NMR experiments. In particular, information was obtained on the arrangement of the clay platelets, the structural and dynamic behavior of the organic cations and alkoxysilane molecules, as well as their interactions with the clay surface. In Laponite-2C18, a remarkable fraction of the organic cations was arranged between the clay platelets with their chains lying on their surfaces in the all-trans conformation, resulting in an increase of the inter-platelets gallery height and in a quite ordered arrangement of the clay platelets. Despite the low amount of alkoxysilane molecules involved in the subsequent grafting reaction, the properties of Laponite-TSPM resulted noticeably modified: the arrangement of the clay platelets was less ordered, on average the platelets were much closer, and the organic cation chains experienced a much higher conformational freedom, most of them giving rise to fast trans-gauche jumps. A reliable hypothesis seems that the grafting occurred, as previously suggested, at the clay platelet edges, but in a quite dishomogeneous way: the platelets were kept together by alkoxysilane molecules or oligomers, which on one side brought closer two adjacent platelets, but on the other side twisted them one with respect to the other (Figure 12). This determined a higher conformational freedom of the organic cation chains, but also seems to favor inter-platelets interactions, through self-condensation among clay silanols at the platelet edges. Moreover, for the first time, silicon nuclei of silanol groups located at the clay platelet edges, available for the reaction with TSPM, were distinguished from unaccessible Q2 silicon nuclei, on the basis of their different 29Si chemical shift. Acknowledgment. Partial financial support from PRIN 2005 035119 is gratefully acknowledged. LA063040A
