Relationship of Swelling and Swelling Pressure on Silica−Water

532

Langmuir 2006, 22, 532-537

Relationship of Swelling and Swelling Pressure on Silica-Water Interactions in Montmorillonite Kalpana S. Katti* and Dinesh R. Katti Department of CiVil Engineering, North Dakota State UniVersity, Fargo, North Dakota 58105 ReceiVed June 9, 2005. In Final Form: NoVember 9, 2005 In this work, we have used the previously designed controlled uniaxial swelling (CUS) cell to obtain predetermined extents of swelling in montmorillonite. Using the CUS cell, a simultaneous measurement of swelling pressure is done with controlled swelling. Undisturbed clay samples at well-defined swelling (0%-75%) were removed from the CUS cell and analyzed using scanning electron microscopy (SEM) and Fourier transform infrared (FTIR) spectroscopy. In addition, orientation-dependent microattenuated total reflectance (micro-ATR) spectroscopic investigations are also conducted on the controlled swelled samples. Significant changes in the silicate (Si-O) stretching region (1150-950 cm-1) have been observed with changes in swelling and orientation. The band at 1005 cm-1 (attributed in the literature to arising from Si-O vibrations when montmorillonite platelets are normal to incident radiation) is most pronounced for the 0%-swelled sample and diminishes with swelling. The band associated with perpendicular vibration (at 1078 cm-1) increases with swelling. Thus, the intensity of this band increases with misorientation of clay particles. Our results indicate that the reduced particle size, as ascertained from SEM cryoimaging, with increased swelling is related to increased misorientation of the clay platelets. At 0% swelling, the clay platelets are most oriented and have largest particle size. The rearrangement of clay platelets as seen in the orientation-dependent spectra is a direct result of the breakdown of the clay particles with increased hydration resulting from increased swelling.

1. Introduction The understanding, modeling, and prediction of interaction of clays with water and other environmental fluids is an important issue in the field of geotechnical engineering, geoenvironmental engineering, as well as industrial applications such as muds for oil well drilling and water treatment.1-5 Swelling clays in particular have been difficult soils to deal with in construction of civil engineering structures such as foundation of structures, retaining walls, roads, irrigation canals, etc. due to damage caused by swelling and swelling pressure. Swelling pressure is defined as the maximum stress applied to expansive soil inundated with water to maintain a constant volume.6 On the other hand, the pure form of these clays are sought after as sealing materials (as a result of swelling) in environmental applications such as land fills and cutoff trenches and as drilling muds in drilling operations, etc. In both cases, the swelling clay-water interaction is responsible for the detrimental effects seen in the geotechnical applications and beneficial properties useful for environmental engineering and other applications. The swelling response of montmorillonite clay and corresponding development of swelling pressure when swelling is restrained is a result of fairly complex clay-water interactions between particles and within the particles themselves. The understanding of these fundamental interactions is critical for design of structures and developing solutions to prevent the detrimental effects, as well as for evaluating the feasibility and optimizing the performance of these materials for environmental * To whom all correspondences should be addressed. E-mail: [email protected]. (1) Anandarajah, A.; Zhao, D. J. Geotech. GeoenViron. Eng. 2000, 126, 148156. (2) Gray, G. R.; Darley, H. C. H. Composition and properties of oil well drilling fluids, 4th ed.; Gulg Publishing Co.: Houston, 1980. (3) Gromko, G. J. J. Geotech. Eng. 1974, 100 (6), 667-687. (4) Low, P. F. Soil Sci. Soc. Am. J. 1979, 43, 651-658. (5) Grim, R. E. Applied Clay Mineralogy; McGraw-Hill: New York, 1962. (6) Nelson, J.; Miller, D. J. ExpansiVe Soils; John Wiley & Sons: New York. 1992.

engineering and other applications. Both experimental and theoretical research has been done to evaluate the role of claywater interactions. The theoretical efforts have involved molecular modeling methods.7-12 Experimental studies have involved electron microscopy and vibrational spectroscopy. Analytical spectroscopy techniques with the scanning electron microscope (SEM) have been particularly useful for simultaneous microstructural characterization at high spatial resolution concurrent with elemental analysis using energy-dispersive X-ray spectroscopy (EDS). Earlier work by Collins13 on the study of microfabric forms in natural expansive soils using SEM indicated that accurate characterization of the microfabric of the clay is the key to assess properly the swelling behavior of expansive soils. Vibrational spectroscopic investigations yield useful information about hydration characteristics, interlayer cations, and moisture content in clays. The excitation of a purely vibrational transition by infrared light obeys the selection rules for electric dipole transitions as described by the classical electromagnetic theory. The interaction of infrared light with a molecule yields a spectrum that is characteristic of the symmetry and molecular structure of the molecule. Many different experimental techniques are used to obtain the vibrational spectrum of materials such as transmission, internal reflection, diffuse reflection, photoacoustics, etc.14 Transmission experiments involve making a thin pellet of powdered sample with KBr powder. This pellet is placed in the (7) Hwang, S.; Blanco, M., Demiral; E., Cagin, T.; Goddard, W. A. J. Phys. Chem. B 2004, 105, 4122-4127. (8) Katti, D. R.; Schmidt, S. R.; Ghosh, P.; Katti, K. S. Clays Clay Miner. 2005, 53, 171-178. (9) Schmidt, S.; Katti, D. R.; Ghosh, P.; Katti, K. S. Langmuir 2005, 21, 8069-8076. (10) Karaborni, S.; Smit, B.; Heidug; W., Urai; J.; van Oort, E. Science 1996, 271, 1102-1104. (11) Greathouse, J. A.; Refson, K.; Sposito, G. J. Am. Chem. Soc. 2000, 122, 11459-11464. (12) Chang, F.-R.; Skipper, N. T.; Sposito, G. Langmuir 1998, 14, 12021207. (13) Collins, K. Proceedings of the 5th International Conference on ExpansiVe Soils, Adelaide, South Australia, 1984, pp 37-41.

10.1021/la051533u CCC: $33.50 © 2006 American Chemical Society Published on Web 12/17/2005

Swelling and Silica-Water Interactions

Langmuir, Vol. 22, No. 2, 2006 533 Table 1. Vibrational Modes for Montmorillonite band position wavenumber (cm-1)

Figure 1. Model showing crystal structure of Na-montmorillonite.

path of the IR beam. Each band in the vibrational spectrum of a sample represents the excitation of normal modes of vibration in the molecular species comprising the sample. Infrared vibrational spectroscopy is thus a useful tool for studying the bonding of water molecules on clay surfaces. Vibrational spectroscopy has been extensively used for studying the uptake of water and other solvents by smectites. Vibrational spectroscopic study of smectite-water interactions were first reported in 1937.15 In addition, vibrational spectroscopy is a useful tool to study changes in the Si-O vibrations resulting from changes in crystal symmetry due to swelling. Clay Structure. Clay particles are in the micron to nanometer range length scale. The small dimensions of clay particles suggest a large influence of the molecular scale behavior and interactions (particle-particle, particle-water, and interlayer) on bulk mechanical properties of great consequence in civil engineering design in particular for prediction and understanding of swelling and swelling pressures in expansive clays. The basic structural units in clays consist of the silica sheet formed of silica tetrahedra (T) and the octahedral (O) units formed of octahedrally coordinated cations (with oxygens or hydroxyls) octahedra.16 Swelling clays, on the other hand, such as montmorillonite are composed of units made of two silica tetrahedral sheets with alumina octahedral sheet in between. Smectite refers to a family of clays primarily composed of hydrated sodium calcium aluminum silicate. Smectites have a T-O-T structure that is similar to that of the mineral pyrophyllite but may also contain significant amounts of Mg and Fe substituting into the octahedral layers. The smectites can thus be both dioctahedral and trioctahedral. One important smectite is montmorillonite, with a chemical formula: (1/2Ca,Na)(Al,Mg,Fe)4(Si,Al)8O20(OH)4‚ nH2O. Figure 1 shows the [010] projection of the montmorillonite structure. Nonswelling clays such as kaolinite consist of units of single sheets of silica tetrahedra between alumina octahedral sheets. In clays, several physiochemical forces are in effect,17 these include van der Waals attraction, electrical double layer repulsion, Born’s repulsive interaction and hydrogen bonding. One of the differences between kaolinite and montmorillonite is the much larger double layer repulsive energy barrier in montmorillonite. The negatively charged surfaces of clay hold or adsorb cations (14) Urban, M. W. Vibrational Spectroscopy of Molecules and Macromolecules on Surfaces; John Wiley & Sons Inc., New York, 1993. (15) Buswell, A. M.; Krebs, K.; Rodebush, W. H. J. Am. Chem. Soc. 1937, 59, 2603-2609. (16) Mitchell, J. K. Fundamentals of soil behaVior; John Wiley & Sons Inc., New York, 1993. (17) Israelachvili, J. N. Intermolecular and surface forces with applications to colloidal and biological systems; Academic Press: New York, 1985.

3634 3433 2470 1635 1204 1115 1078 1034 1005 918 888 847 778 671

band assignment O-H stretching H-O-H hydrogen bonded water O-D stretching H-O-H bend O-H deformation

} }

Si - O stretching

O-H deformation -linked to cation

Si-O deformation perpendicular to optical axis Si-O deformation parallel to optical axis

tightly, and the cations, which are in excess to neutralize the electro-negativity of the clay, are present as salt precipitates. When water infiltrates the clay, the precipitated salt goes into solution. The cation concentration near the clay surface is higher and the cations begin to diffuse away from the surface, since concentrations are much higher than that of the bulk solution. However, the electrostatic attraction between the negative clay surface and the cations inhibits this motion.18 Consequently, the density of cations near the surface is higher than that of the bulk solution with electrical potential maximum on the clay surface and zero at infinity.19 The balance of the interactions between the charged surface and the distributed charge in the adjacent phase results in the formation of the diffuse double layer or electrical double layer.18 The amount of overlap between adjacent double layers significantly impacts the long-range repulsive force between clay layers. The tendency for particles in suspension to flocculate reduces with increasing thickness of the diffuse layer, which results in higher swelling pressure in expansive soils.16 In recent work, by Anandarajah,20 it is shown that the deviation of clay platelets a few degrees from parallelism can lead to a decrease in the double layer force by as much as 100%, and changes in soil fabric orientation has significant effect on the compressive behavior of montmorillonite. Vibrational Spectroscopic Studies. A detailed infrared spectroscopic study on clays from the Clay Mineral Society is described in a recent work.21 Vibrational spectroscopic investigations yield useful information about hydration characteristics, interlayer cations, and moisture content in clays. Infrared vibrational spectroscopy is thus a useful tool for studying the interaction of water molecules with clay surfaces. In addition, vibrational spectroscopy is a useful tool to study changes in the Si-O vibrations resulting from changes in crystal symmetry due to swelling. Table 1 shows the band assignments of the different vibrations in the Si-O region. Earlier work using X-ray diffraction22,23 has indicated several structural changes in the montmorillonite during swelling, which are also observed as changes in the Si-O absorption bands. The Si-O region of the spectrum is rather complex with several overlapping spectral features. In the literature,24-26 four bands have been identified, at 1028, 1052, 1080, and 1122 cm-1. Out (18) Anderson, M.; Lu, N.; Mustoe, G. Proceedings of the 14th ASCE Engineering Mechanics Conference, Austin, TX, 2000, pp 1-6. (19) Chen, J.; Anandarajah, A.; Inyang, H. J. Geotech. GeoenViron. Eng. 2000, 126 (9), 798-807. (20) Anandarajah, A. J. Colloid Interface Sci. 1997, 194, 44-52. (21) Madejova, J.; Komadel, P. Clays Clay Miner. 2001, 49, 410-432. (22) Handy, R. L.; Turgut, D. Proceedings of the 6th International Conference on ExpansiVe Soils, New Delhi, India, 1987, pp 161-166. (23) Ravina, I.; Low, P. F. Clays Clay Miner. 1972, 20, 109-123.

534 Langmuir, Vol. 22, No. 2, 2006

Katti and Katti

Figure 2. (a) Drawing showing a section through the controlled uniaxial swelling (CUS) cell and (b) a photograph of the CUS cell placed in a load frame.

of these bands, the principle Si-O vibration band is at 1052 cm-1 and the increase in water content causes this band to be narrow and increase in intensity.24 Thus, this band is a good indicator of swelling characteristics. Further, changes in orientation of vibrations are associated with the band at 1086 cm-1. Some work in the literature also suggests that the changes in this band are related to changes in soil fabric resulting from disorientation of the layers in montmorillonite. Sposito and Prost25 observed an increase in the intensity of the 1122 cm-1 band with increasing clay water content. Shewring et al.26 observed small shifts in the position of the band at ∼1080 cm-1 with changes in hydration. A marked dependence of the intensity of the main Si-O band with hydration was suggested to be attributed to possible changes in particle size;27 however, experimental evidence for this is lacking. In addition, the Si-O stretching region (∼1040 cm-1) is also analyzed for influence of hydration of montmorillonite. Several researchers24-28 have identified a band at ∼1086 cm-1 and assigned it to a Si-O stretching vibration perpendicular to the surface of the clay platelet. They observed no shift in position but an increase in intensity of the band with water content. This was suggested to be related to disorientation of layers in montmorillonite with swelling. Further, a decrease in stretching frequencies corresponding to an increase in H-bonding of the water molecule has been observed. Also, the intensity of the O-H stretching vibrations increases with increasing hydrogen bonding. Spectral hydration features can be associated with O-H in the octahedral layer, water adsorbed on grain surfaces, and water bound in the interlayer regions. Further, several vibrational spectroscopic studies exist in the literature on the interaction and association of montmorillonite with water that focus on the O-H stretching and O-H bending regions of water.28 Some evidence for interaction of montmorillonite layers and interlayer water was shown recently.29 A correlation between the Si-O stretching vibration and H-O-H bending vibrations in the interlayer water is also observed. Water exhibits three vibrational modes: a symmetric stretch, an asymmetric stretch, and an O-H bend. It is also shown that the H-O-H bending vibration shifts to lower energy upon dehydration of montmorillonite.28 More recently, a correlation between water content (24) Lerot, L.; Low, P. F. Clays Clay Miner. 1976, 24, 191-199. (25) Sposito, G.; Prost, R. Chem ReV. 1982, 82, 553-573. (26) Yan. L.; Roth, C. B.; Low, P. E. Langmuir 1996, 12, 4421-4429. (27) Shewring, N. I. E.; Jones, T. G. J.; Maitland, G.; Yarwood, J. J. Colloid Interface Sci. 1995, 176, 308-317. (28) Russel, J. D.; Farmer, V. C. Clay Miner. Bull. 1964, 5, 443-464. (29) Johnson, C. T.; Sposito, G.; Erikson, C. Clays Clay Miner. 1992, 40, 722-730.

of montmorillonite films with different exchange cations has been presented.29 The role of interlayer cations, moisture content, and the nature of water on the vibrational spectrum of montmorillonite has also been illustrated.30,31 Further, attenuated total reflectance (ATR)-FTIR studies32 on an aqueous suspension of smectite show that smectite particles are more oriented in suspension than the dry deposit obtained after drying the suspension. Also FTIR experiments have been used to study the uptake of herbicides by smectites.33 2. Experimental Materials. Na-montmorillonite (Swy-2, Crook County, Wyoming) of cationic exchange capacity 76.4 mequiv/100 g was obtained from the Clay Minerals Repository at the University of Missouri, Columbia, MO. The SWy-2 montmorillonite has the chemical formula NaSi16(Al6FeMg)O20(OH)4.34 Deuterated water (D2O) of 99.9% purity was obtained from Aldrich Chemical Co. Double deionized water was used for all swelling experiments. Spectroscopic Techniques. Spectra were obtained on a Nicolet 850 FTIR spectrometer, using a KBr beam splitter and a NicPlan FTIR microscope. The FTIR microscope was equipped with a narrow band liquid nitrogen cooled MCT detector allowing spectra to be obtained between 4000 and 500 cm-1. Swelling and Swelling Pressure. A controlled uniaxial swelling (CUS) device (Figure 2a,b) was designed and fabricated to allow a predetermined amount of swelling, to measure the corresponding swelling pressure, and to allow for easy removal of the sample for consecutive microanalysis in the scanning electron microscope (SEM) and vibrational microspectroscopy. The percent swelling is defined as the change in the volume of the sample relative to the initial dry volume. Details of the construction of the CUS cell can be seen in our previous work.35 Montmorillonite samples were dry compacted in the cells using static compaction to a height of 14-mm and a diameter of 70 mm. An initial void ratio of 2.145 (dry unit weight of 8.33 kN/m3) was used for all samples. The void ratio is defined as the ratio of volume of voids to volume of solids in a given soil mass. The piston of the CUS cell was locked in position to maintain a zero volume change during saturation. The cell, immersed in a water tank, was then placed in a SATEC 22 EMF universal testing machine. (30) Karakassides, M. A. Clay Miner. 1999, 34, 429-438 (31) Bishop, J. L.; Pieters, C. M.; Edwards, J. C. Clays Clay Miner. 1994, 42, 702-716. (32) Johnston, C. T.; Premachandra, G. S. Langmuir 2001, 17, 3712-3718. (33) Johnston, C. T.; Sheng, G.; Teppen, B. J.; Boyd, S. A.; De Oliveira, M. F. EnViron. Sci. Technol. 2002, 36, 5067-5074. (34) van Olphen, H., Fritpiat, J. J., Eds. Data Handbook for Clay Materials and other Nonmetallic Minerals; Pergamon Press: New York, 1979. (35) Katti, D. R.; Shanmughasundaram, V. Can. Geotech. J. 2001, 38, 175182.

Swelling and Silica-Water Interactions The piston was maintained at zero volume change position by maintaining the position of the loading ram of the load frame through closed-loop electronics. The swelling pressure at no volume change condition was recorded. The loading ram was then moved up to a predetermined position of 7 mm for 50% swelling and 10.5 mm for 75% swelling. The percent swelling is defined as the change in the volume of the sample relative to the initial dry volume. Swelling pressure with time for each of the swelling conditions is recorded. Microstructural Characterization. Samples swollen at predetermined amounts of swelling (0%, 12.5%, 25%, 50%, and 75%) were removed from the cells and samples were axially sliced using a thin ceramic knife. The samples for microstructural study were sectioned (5 mm × 5 mm × 1 mm) from the central region of the controlled swelled sample from the CUS cell and rapidly frozen to -165 °C by dipping the sample in liquid nitrogen. This cryofixation technique is based upon the procedures described in Tovey et al.36-38 Tovey’s group has shown that this technique maintains the clay microstructure. The samples were mounted in a cryogenic stage of the SEM and images obtained at various magnification. Elemental analysis was performed using an energy dispersive spectroscopy (EDS) detector on the SEM to delineate particles and voids. The sample slices were prepared for SEM analysis using cyrofixation in liquid nitrogen. Samples were then placed in the cold stage of a JEOL 6300 (JEOL, Peabody, MA) analytical SEM operated at 5 kV maintained at -165 °C. Elemental analysis was performed using an EDS detector (NORAN Instruments, Middleton, WI) on the SEM. The high concentration of silicon in the solid phase as compared to the void allowed for the delineation between solids and voids. Micrographs were digitally recorded at 50×, 650×, 2000×, and 10 000× magnifications. The micrographs obtained were then analyzed using Scion Image (Scion Corp., Fredrick, MD; the Windows version of NIH Image, National Institutes of Health). Details of the custom modules developed to digitally delineate the void and solid regions in the electron micrographs and calculate the particle size dimensions can be seen in previous work.35 We used a minimum saturation time of 30 days for all samples in the CUS cell to ensure full saturation. Further, a fully saturated sample in the CUS cell was placed in the load frame and swelling pressure was measured at 0%, 1%, 2%, 4%, 10%, 12.5%, 17.5%, 25%, 50%, 75%, and 100% swelling. To understand the effect of swelling and swelling pressure on microstructural changes in the montmorillonite samples, three identical samples of montmorillonite clay were each swollen to 0%, 50% and 75% of their original volume uniaxially. These samples were removed from the cell and sectioned/frozen for microstructural analysis as described earlier. These samples were mounted in a cryogenic stage of the SEM. FTIR Vibrational Spectroscopy (Micro-ATR). Montmorillonite (Wyoming montmorillonite) was used as supplied. X-ray diffraction was used to verify the chemical characteristics of the montmorillonite as sodium montmorillonite. Swelled, saturated, undisturbed samples of the montmorillonite were removed from the CUS cell and samples tested immediately. Absorbance spectra were obtained for all the samples at varying stages of swelling. The 0%-75%-swelled samples were removed from the molds in an undisturbed state and sliced at 0°, 45°, and 90° to the swelling axis. Absorbance spectra from such samples were then obtained using micro-ATR. Deuterated Montmorillonite Slurry. Further, deuterated montmorillonite was prepared by dispersing 500 mg of montmorillonite in 5 mL of D2O (Aldrich Chemical Co., 99.9% purity) and stirred under dry nitrogen for 4 days. ATR-FTIR spectra were obtained for the deuterated montmorillonite in comparison with montmorillonite dispersed in H2O under identical conditions. (36) Tovey, N. K. 1970. Electron microscopy of clays. Ph.D. Thesis, Cambridge University, Cambridge, U.K. (37) Tovey, N. K.; Frydman, S.; Wong, F. K. Proceedings of the Third International Conference on ExpansiVe Soils, Haifa, 1973, Vol. 2, pp 45-54. (38) Tovey, N. K. J. Microsc. 1980, 1, 303-315.

Langmuir, Vol. 22, No. 2, 2006 535

Figure 3. Plot of relative swelling pressure vs percent swelling in montmorillonite.

3. Results and Discussion The variation of swelling pressure of dry montmorillonite clay sample during saturation was recorded over a period of 21 days. The swelling pressure rapidly increases to about 120 kPa within 1 day. This is followed by a gradual increase to about 160 kPa within 9-10 days. After a period of 10-11 days, the swelling pressure attains an almost constant magnitude of 163 kPa. On removal of the sample from the cell after 21 days on completion of the test, the degree of saturation was found to be 100%. Several samples were tested to verify complete saturation. The magnitude of swelling pressure of montmorillonite clay found in our laboratory experiments falls within the range of swelling pressure values for montmorillonite clay samples reported in the literature. The swelling pressure of these clays is a function of the initial void ratio and is known to increase exponentially with an increasing dry density (decreasing void ratio).39,40 The relative swelling pressure versus swelling percentage is shown in Figure 3. The swelling pressure is expressed relative to the maximum observed swelling pressure of 162 kPa. Clearly, as seen in Figure 3, the maximum swelling pressure is observed at 0% swelling, which decreases to about 6% of the maximum at 100% swelling. Further, the swelling pressure reduces to 50% of the maximum at 15 to 20% swelling, which decreases to 10% at 75% swelling. To understand the effect of swelling and swelling pressure on microstructural changes in the montmorillonite samples, three identical samples of montmorillonite clay were each swollen to 0%, 12.5%, 25%, 50%, and 75% of their original volume uniaxially. These samples were removed from the cell and sectioned/frozen for microstructural analysis as described earlier. These samples were mounted in a cryogenic stage of the SEM. Particles and voids were identified using elemental analysis of Si using energy dispersive spectroscopy. Large particle sizes are observed for the 0% swelling sample, and the particle sizes appear to diminish in the 75% swelling sample. We used the Scion Image software in combination with custom modules to process the images. First, the digitized images were used to obtain boundaries of particles in the images followed by counting the particles and measurement of the number of pixels along the narrow axis for each particle. A calibration factor for pixel size (39) Chen, F. H. Foundations on ExpansiVe Soils, 2nd ed.; Elsevier: Amsterdam, 1988. (40) Fredlund, D. G.; Rahardjo, H. Soil Mechanics for Unsaturated Soils; Wiley Publications: New York, 1993.

536 Langmuir, Vol. 22, No. 2, 2006

Katti and Katti

Figure 4. Plot of particle size vs swelling percent and swelling pressure in montmorillonite.

Figure 5. FTIR-micro-ATR spectra from Si-O region of the spectrum for 0%, 25%, and 50% swelling. For comparison, spectra of wet montmorillonite sample and slurry are also shown.

Figure 6. FTIR-micro-ATR spectra of montmorillonite suspension shown in (a) 1000-3600 cm-1 region and (b) 800-1300 cm-1 region. Traces A are for D2O and B for H2O.

enables the measurement of mean particle sizes. A plot illustrating the variation of particle size with swelling and relative swelling pressure is shown in Figure 4. The changes in the Si-O region of the spectrum for 0%, 25%, and 50% swelling as compared to that of a slurry as obtained using FTIR-micro-ATR spectroscopy are shown in Figure 5. As seen, in the 0%-swelled sample, the 1078 cm-1 appears as a shoulder but it has increased intensity with swelling, with the 50%-swelled sample exhibiting the highest intensity. This band is associated with increased distortion and misorientation in clay layers, and the increase in intensity of this band for the 50%swelled sample as compared to 0%-swelled sample signifies increased misorientation. The intensity of the 1115 cm-1 band is highest for the 0%-swelled sample. A close examination of the bands at 1005 and 1034 cm-1 reveals a relative increase in the intensity of the band at 1034 cm-1 with respect to the band at 1005 cm-1 with swelling. The overall decrease in the absorption bands in the Si-O region with swelling is expected due to the decrease in clay mineral content with increased hydration. The FTIR-ATR spectra of montmorillonite-H2O/D2O suspension are shown in Figure 6a,b. Traces A and B refer to the D2O and H2O dispersions, respectively. As seen, the characteristic shifts between O-H and O-D stretching and bending vibrations are observed. In addition, the Si-O (1150-950 cm-1) region of the spectrum does not show any apparent changes in peak positions

or features. Spectra shown in Figure 6b are that of slurries of montmorillonite in D2O and H2O. In a previous work, similar experiments showing no significant change in the Si-O region of the spectrum were suggested to imply that no significant interactions exist between the silica tetrahedral and water.26 Clear differences exist in the silica tetrahedral with increase in swelling, as indicated by Figure 5. Orientation-Dependent Experiments. Samples of 0%swelled samples were removed in an undisturbed state from the CUS cell and sliced at 0°, 45°, and 90° to the direction of swelling. Micro-ATR spectroscopic investigations were performed on each of these samples. The absorption spectra for these samples are shown in Figure 7a,b. These spectra are normalized with respect to the O-H vibration band. Figure 7a,b shows spectra obtained for 0%- and 12.5%-swelled samples. For the 0%-swelled sample, the spectrum collected from the 45°-sliced sample shows maximum intensity of the 1005 cm-1 band. Further, the 90°sliced sample shows maximum intensity of the 1078 cm-1 band. The principle Si-O vibration band at 1034 cm-1 is seen in addition to the 1005 cm-1 band and is more pronounced in the sample sliced at 90°. The overall intensities of the Si-O region appear similar for the 12.5%-swelled sample as compared to the 0%swelled sample. The 45°-sliced sample shows the highest intensity followed by the 90°- and 0°-sliced samples. This indicates that the principle Si-O excitation is largely aligned close to 45°. On

Swelling and Silica-Water Interactions

Figure 7. (a) FTIR-micro-ATR spectra of 0%-swelled sample sliced at 45°, 90°, and 0° to the swelling axis. (b) FTIR-micro-ATR spectra of 12.5%-swelled sample sliced at 45°, 90°, and 0° to the swelling axis.

the other hand, for the 12.5%-swelled sample, the principle Si-O vibration band at 1034 cm-1 shows higher intensity for the 45°sliced sample. The reversal in relative intensities of the 1005 and 1034 cm-1 bands as seen between the 45°- and 90°-sliced samples as seen for the 0%-swelled sample is not observed for the 12.5%swelled sample. Clearly, for the 0%-swelled sample the clay layers are more oriented and thus the 1034 and 1005 cm-1 bands reverse in intensity depending on the angle of slicing. In the 12.5%-swelled sample these differences are not apparent, since the clay layers are more misoriented. Thus these changes indicate a change in misorientation with swelling.

4. Summary and Conclusions A controlled uniaxial swelling (CUS) is utilized that allows simultaneous measure of swelling and swelling pressure as well as ease of removal of undisturbed samples under predetermined levels of swelling for scanning electron microscopy and Fourier transform infrared spectroscopy. Measurement of particle sizes at varying amounts of swelling using the CUS indicates that the particle sizes decrease as swelling increases in montmorillonite.

Langmuir, Vol. 22, No. 2, 2006 537

In this work, the FTIR spectra of montmorillonite were obtained for the controlled swelling stages of the clay samples. Orientationdependent micro-ATR spectroscopic investigations were also conducted on the controlled swelled samples. Russel and Farmer28 have observed a shoulder at ∼1015 cm-1 in the transmission spectrum of montmorillonite film whose platelets are normal to incident radiation. As seen in Figure 5, the shoulder appearing at 1005 cm-1 in the slurry trace is highly pronounced for the 0%-swelled sample. As swelling percent increases from 0% to 50%, the intensity of the 1005 cm-1 band slowly diminishes. In fact, the spectrum at 50% swelling appears to approach the slurry trace. This indicated that 0%-swelled samples are highly oriented as compared to 50%-swelled samples, although the dry, compacted sample before saturation was prepared from randomly oriented particles. The uptake of water by sodium montmorillonite followed by distortion of the silica tetrahedra appears to be influenced by the competitive hydrogen bonding between water molecules and between water molecules and silica tetrahedra.30 Hydration experiments have shown that increase in water content causes the appearance and development of the band associated with perpendicular vibration (1078 cm-1).23 In our experiments, increasing swelling results in increase in water content of the clay sample. As seen in Figure 5, 0% swelling shows a small shoulder near 1078 cm-1, but as swelling increases to 25% and 50%, the band also increases in intensity. Wet and slurry samples of the clay also show high intensity of the band at 1078 cm-1. This experiment not only illustrates the increase in intensity of the 1078 cm-1 band with hydration but also shows that the intensity of this band increases as misorientation of the clay platelets increases with increased swelling. We interpret this to mean that, as layers move apart, particles break down, which in turn leads to increased misorientation of the clay platelets. At 0% swelling, the clay platelets are most oriented and have largest particle size. Studies on aqueous suspension of smectite32 show clay platelets more oriented in suspension than dry samples obtained by drying the suspension. This is expected, since in suspension the clay platelets are free to orient themselves. However, clay platelets in samples swollen to 25% and 50% swelling show increase in misorientation with increase in swelling, possibly due to constraints imposed to freedom of movement of particles by adjacent particles. The orientation studies on clay samples further indicate the increase in misorientation with swelling. This study indicates that the magnitude of swelling plays an important role in the orientation or misorientation of clay platelets in swelling clays. Increased swelling reduces swelling pressure due to particle breakdown which is accompanied by increasing misorientation of clay layers. Acknowledgment. The authors would like to acknowledge the support from National Science Foundation (Grant # 0114622). The program manager at NSF is Dr. Richard J. Fragaszy. The authors also acknowledge Mr. Vijayakumar Shanmugasundaram for help with sample preparation. LA051533U