Modification of Kaolinite Surfaces with Cesium Acetate at 25, 120

Vijay Bahadur Yadav , Ranu Gadi , Sippy Kalra. Applied Clay Science 2018 155, ... Nouha Jaafar , Hafsia Ben Rhaiem , Abdesslem Ben Haj Amara. 2014,1-6...
0 downloads 0 Views 85KB Size
Langmuir 1999, 15, 8787-8794

8787

Articles Modification of Kaolinite Surfaces with Cesium Acetate at 25, 120, and 220 °C Ray L. Frost,*,† Janos Kristof,‡ Erzsebet Horvath,§ and J. Theo Kloprogge† Centre for Instrumental and Developmental Chemistry, Queensland University of Technology, 2 George Street, GPO Box 2434, Brisbane, Queensland 4001, Australia, Department of Analytical Chemistry, University of Veszprem, PO Box 158, H8201 Veszprem, Hungary, and Research Group for Analytical Chemistry, Hungarian Academy of Sciences, PO Box 158, H 8201 Veszprem, Hungary Received December 24, 1998. In Final Form: August 23, 1999 Kaolinite hydroxyl surfaces have been modified upon intercalation with cesium acetate under a range of conditions. Upon intercalation of both low- and high-defect kaolinites with cesium acetate at 25 °C, additional infrared hydroxyl stretching bands are observed at 3606 and at 3603 cm-1. New infrared hydroxyl deformation modes were observed at 898 and 910 cm-1 upon intercalation with cesium acetate. These changes are attributed to the hydrogen bonding of the acetate ion to the inner surface hydroxyls. The intercalation process involves the incorporation of water into the interlayer structure. When treatment of the kaolinites is carried out under hydrothermal conditions, intercalation is incomplete. The 001 d spacing showed strong asymmetry on the low angle side indicative of some expansion of the kaolinite layers. Changes in both the hydroxyl stretching and deformation regions indicate that water is being incorporated into the kaolinite structure at the elevated temperatures of 120 and 220 °C. Raman spectra of the 25 °C cesium acetate intercalated kaolinites show additional bands at 3606 and 3598 cm-1. These bands are attributed to the inner surface hydroxyls hydrogen bonded to the acetate ions. The Raman spectra also suggest that some structural rearrangements of the kaolinite layers are occurring with consequential alteration in the defect structures.

Introduction Kaolinite and the other kaolin polytypes (halloysite, dickite, nacrite) contain two types of hydroxyl groups: hydroxyl groups at the internal surface, OHou, often referred to as inner surface hydroxyl groups, and hydroxyl groups, OHin, often referred to as inner hydroxyl groups, within the layer. The OHou groups are situated in the outer, unshared plane whereas the OHin groups are located in the inner-shared plane of the octahedral sheet, often termed the gibbsite-like sheet.1,2 The OHou groups are located on the outside of a single kaolinite layer. However the kaolinite crystal consists of a stacking of many of these layers, which means that actually the OHou groups are located in the space between two adjacent layers; hence the term inner surface hydroxyl is normally used in the literature. Figure 1 shows a model of kaolinite unit cell with the OHin and OHou hydroxyl groups. The OHin hydroxyl is within the unit cell and not exposed at the surface of the kaolinite; consequently it is difficult to modify the kaolinite surfaces by interaction with this hydroxyl group. The OHou hydroxyl groups are at the gibbsite-like surface of the kaolinite and are readily modified. * To whom correspondence should be addressed: e-mail, [email protected]; fax, +61 7 3864 1804; phone, +61 7 3864 2407. † Queensland University of Technology. ‡ University of Veszprem. § Hungarian Academy of Sciences. (1) Brindley G. W.; Chih-Chun K.; Harrison J. L.; Lipsiscas M.; Raythatha R. Clays Clay Miner. 1986, 34, 233. (2) Frost, R. L.; Kristof, J.; Tran, T. H. T.; Kloprogge, J. T. Am. Mineral. 1998, 83, 1182.

Figure 1. Model of the unit cell of kaolinite.

When the kaolinite surface is modified through interaction with any number of bonding molecules, changes are reflected in the vibrational spectrum, in either the infrared or Raman spectrum or both.2,3 Infrared spectroscopy has been extensively used for the study of kaolinite polytypes, and the four distinct infrared bands are normally assigned as follows: the three higher frequency vibrations (ν1, ν2, ν3) at 3695, 3670, and 3650 cm-1 are due to the three inner surface hydroxyls (OHou) and the band (ν5) at 3620 cm-1 is due to the inner hydroxyl (OHin). The hydrogens of the inner hydroxyl groups are bound to the oxygen below the aluminum atoms and directed toward the intralayer cavity of the kaolinite (Figure 1). Any modification of kaolinite surfaces is unlikely to alter the inner hydroxyl group. The OHin (3) Frost, R. L.; Kristof, J.; Paroz, G.; Kloprogge, J. T. J. Colloid Interface Sci. 1998, 208, 478.

10.1021/la981755a CCC: $18.00 © 1999 American Chemical Society Published on Web 11/16/1999

8788

Langmuir, Vol. 15, No. 26, 1999

hydroxyls essentially lie parallel to the 001 plane. This hydroxyl points toward the vacant dioctahedral site in the kaolinite structure.4-7 The inner surface hydroxyl groups (OHou) point toward the adjacent siloxane layer. Hydrogen bonds are formed between the gibbsite-like layer and the adjacent siloxane layer. It is these hydrogen bonds which hold many kaolinite layers in a crystal. Upon modification of the surfaces of kaolinite, this interlayer hydrogen bonding is broken and is normally replaced with hydrogen bonding to the reacting molecule.8 Such alteration in the hydrogen bonding is reflected in the changes in the vibrational spectrum of the kaolinite. The application of infrared spectroscopy to the study of the changes in the kaolinite structure has proved worthwhile.9 Remarkable changes in the intensities of the infrared bands in the low-frequency region for kaolinite also occurred upon intercalation. In particular, the 935 cm-1 band changed in intensity upon intercalation and provided evidence for the acetate ion being hydrogen bonded to the inner surface hydroxyl group. The application of Raman microscopy to the study of intercalated kaolinites has also proved most useful.10-13 An additional Raman band, attributed to inner surface hydroxyl groups strongly hydrogen bound to the acetate, was observed at 3605 cm-1 for the potassium acetate intercalate with the concomitant loss of intensity in the OHou bands at 3652, 3670, 3684, and 3693 cm-1. Intercalation of a high defect kaolinite resulted in a Raman spectrum similar to that of an intercalated ordered kaolinite. Thus the intercalated halloysite resembled the intercalated kaolinite, at least on a molecular level.11 Here the conclusion was made that the intercalation process resulted in a changes in the kaolinite surfaces such that a decrease in the defect structures of the kaolinite occurred. In a previous study the effect of pressure on the intercalation of both an ordered and disordered kaolinite with potassium acetate induced disordering as evidenced by both X-ray diffraction and Raman spectroscopy.13 Intercalation of the ordered kaolinite with potassium acetate (KCH3COO) under a pressure of 20 bar and 220 °C, induced new Raman bands at 3590, 3603, and 3609 cm-1 in addition to the normal kaolinite bands. These bands were attributed to the inner surface hydroxyls hydrogen bonded to the acetate. It was proposed that the intercalation under 20 bar pressure at 220 °C caused the differentiation of the inner surface hydroxyl groups, resulting in the appearance of these additional bands. Diffuse reflectance infrared spectra of the potassium acetate intercalated kaolinite pressure formed at 20 bar and at 220 °C showed new bands at 3595 and 3605 cm-1. Upon formation of the intercalate at 2 bar and at 120 °C, additional infrared bands were found at 3592, 3600, and 3606 cm-1. These infrared bands correspond well with the observed Raman spectra. It was proposed that the (4) Brindley, G. W. Order-disorder in clay mineral structures. In Crystal Structures of Clay Minerals and their X-ray Identification; Brindley, G. W., Brown, G., Eds.; Mineralogical Society Monograph No. 5; Mineralogical Society: London, 1984; Chapter 2, pp 125-189. (5) Giese, R. F. Bull. Mineral. 1982, 105, 417. (6) Giese, R. F. Kaolin minerals: structures and stabilities. In Reviews in Mineralogy Volume 19 Hydrous Phyllosilicates; Bailey, S. W., Ed.; Mineralogical Society of America BookCrafters, Inc.; Chelsea, MI, 1988; Chapter 3. (7) Hess, C. A.; Saunders, V. R. J. Phys. Chem. 1992, 96, 4367. (8) Constanzo, P. M.; Giese, R. F. Clays Clay Miner. 1990, 38, 160. (9) Barrios, J.; Planc¸ on, A.; Cruz, M. I.; Tchoubar, C. Clays Clay Miner. 1977, 25, 422. (10) Frost, R. L.; Tran, T. H.; Kristof, J. Clay Miner. 1997, 32, 587. (11) Frost, R. L.; Kristof, J. Clays Clay Miner. 1997, 45, 68. (12) Frost, R. L.; Van Der Gaast, S. J. Clay Miner. 1997, 32, 293. (13) Frost, R. L.; Kristof, J.; Paroz, G.; Kloprogge, J. T. J. Colloid Interface Sci. 1998, 204, 478.

Frost et al.

effect of intercalation of the highly ordered kaolinite under pressure caused the kaolinite to become disordered and this disordering was dependent on the temperature of intercalation. It was suggested that when pressure is applied to the kaolinite crystal in the presence of an intercalating agent, the hydrogen bonds between adjacent layers are broken in order to create space for the intercalating agent between the layers. A direct result is that the order of the kaolinite crystals shows a decrease resulting in more defect structures. The additional bands in both the Raman and infrared spectra evidence this. In the kaolin production industry, molecules such as urea and potassium acetate are used to chemically delaminate the kaolinite before reconstitution of the kaolinite for use as a coater and filler in, for example, the paper industry. The use of these molecules for delamination creates the problem of removal of the inserting molecule, and several washing treatments are required before the kaolinite is clean. Cesium acetate is also an intercalating agent and incompletely intercalates the kaolinite. The reason is due to the different size of the potassium and cesium cations. It is considered that whereas urea and potassium acetate modify kaolinite surfaces to a large extent, the cesium acetate incompletely modifies the kaolinite surface and provides a mechanism for milder treatment of the kaolinite surfaces. Our aim in the current study is to research the reordering of the kaolinite layers when the kaolinite is intercalated with cesium acetate over a range of temperatures and pressures. In the kaolin industry, kaolinites of different structural orders are mined, i.e., with different stacking arrangements. Typically kaolinites with low defects or high order and kaolinites with high defects or low order are obtained. In the kaolin production industry in Georgia, USA, these two types of kaolinites are known as soft and hard kaolinites, respectively. Therefore to undertake this work, it is best to study both low- and high-defect kaolinites. In this paper we report the changes in the surfaces of both low- and high-defect kaolinites induced through intercalation with cesium acetate both at ambient temperatures and under hydrothermal conditions using vibrational spectroscopy. Experimental Methods Intercalation at Room Temperature. The clay minerals used in this study are from Kiralyhegy and Szeg, Hungary. Samples were analyzed for phase purity using X-ray diffraction techniques and Raman microprobe spectroscopic analysis. Some quartz and traces of a potassium feldspar were observed. Quartz was removed by sedimentation techniques, and the mineral was dried in a desiccator to remove adsorbed water and used without further purification. The air-dried kaolinite was intercalated according to Weiss.14,15 Three hundred milligrams of kaolinite was treated with 30 cm3 of 3.0 M saturated cesium acetate solution. The samples were shaken for 80 h in a constanttemperature bath at ambient temperature. The excess solution was removed by centrifugation. The intercalated kaolinite was allowed to dry in air, before diffuse reflectance IR Fourier transform (DRIFT) and Raman spectroscopic analysis. Intercalation under Pressure. Intercalation under high pressure and temperature was carried out in a High-Pressure Asher (HPA) (Anton-Paar, Austria). Two hundred milligrams of Kiralyhegy or Szeg kaolinite together with 1.6 g of cesium acetate and 20 cm3 of water were placed in the quartz bomb of the HPA equipment that is closed by a quartz lid. A Teflon gasket between (14) Weiss, A.; Thielepape, W.; Ritter, W.; Schafer, H.; Goring, G. Zur Kenntnis von hydrazin-kaolinit. Anorg. Allg. Chem. 1963, 320, 183. (15) Weiss, A.; Thielepape, W.; Orth, H. Intercalation into kaolinite minerals. In Proceedings of the International Clay Conference; Heller, L., Weiss, A., Eds.; Jerusalem: Israel University Press: Jerusalem, 1966; pp 277-293.

Modification of Kaolinite Surfaces the bomb and the lid ensured the gastight sealing of the bomb. Then the bomb with the lid was placed in the metal heating block of the equipment and a nitrogen pressure of 80 bar was applied over the lid to prevent escape of vapors from the bomb. In 30 min the temperature was increased to 220 °C and kept constant for 8 h. After the bomb was cooled to room temperature, the pressure was released and the clay was separated from the solution by centrifugation. During this treatment the computer controlling the process recorded the temperature and the outside pressure. At 220 °C the nitrogen pressure increases to 120 bar. It is not possible to measure the pressure inside the bomb, but it does not exceed 23 bar. The pressure inside the bomb is taken to be nominally 20 bar. In addition a parallel experiment was conducted in which the same kaolinites were heated only to 120 °C. In this case the inside pressure did not exceed 2 bar. X-ray Diffraction. X-ray diffraction (XRD) analyses were carried out on a Philips wide-angle PW 1050/25 vertical goniometer equipped with a graphite-diffracted beam monochromator. The d spacing and intensity measurements were improved by application of a self-developed computer-aided divergence slit system enabling constant sampling area irradiation (20 mm long) at any angle of incidence. The goniometer radius was enlarged from 173 to 204 mm. The radiation applied was Co KR from a long fine focus Co tube, operating at 40 kV and 40 mA. The samples were measured at 50% relative humidity in stepscan mode with steps of 0.02° 2θ and a counting time of 2 s. Spectroscopy. DRIFT Spectroscopy. Diffuse reflectance Infrared Fourier Transform spectroscopic (commonly known as DRIFT) analyses were undertaken using a Bio-Rad 60A spectrometer. Five hundred twelve scans were obtained at a resolution of 2 cm-1 with a mirror velocity of 0.3 cm/s. Spectra were coadded to improve the signal-to-noise ratio. Approximately 3 wt % kaolinite or intercalated kaolinite was dispersed in 100 mg of oven-dried spectroscopic grade KBr with a refractive index of 1.559 and a particle size of 5-20 µm. Reflected radiation was collected at ∼50% efficiency. Background KBr spectra were obtained and spectra ratioed to the background. The diffusereflectance accessory used was designed exclusively for Bio-Rad FTS spectrometers. It is of the so-called “praying monk” design and is mounted on a kinematic baseplate. It includes two fourposition sample slides and eight sample cups. The cup (3 mm deep, 6 mm in diameter) accommodates powdery samples mixed with KBr using an agate mortar and pestle in 1-3% concentration. The collection efficiency of this adaptor (part number 0990931) is approximately 50%. The reflectance spectra expressed as Kubelka-Mink unit versus frequency curves are very similar to absorbance spectra and can be evaluated accordingly. The advantage of using DRIFT measurements over the pellet technique is that in this case the likely interference of the mulling agent (intercalation of KBr in a liquid phase under pressure) can be avoided. Raman Spectroscopy. To obtain Raman spectra, very small amounts of the cesium acetate intercalated kaolinite were placed on a polished metal surface on the stage of an Olympus BHSM microscope, equipped with 10×, 20×, and 50× objectives. The microscope is part of a Renishaw 1000 Raman microscope system, which also includes a monochromator, a filter system, and a charge-coupled device (CCD). Raman spectra were excited by a Spectra-Physics model 127 He/Ne laser (633 nm) and recorded at a resolution of 2 cm-1 and were acquired in sections of approximately 1000 cm-1. Repeated acquisitions using the highest magnification were accumulated to improve the signalto-noise ratio. Spectra were calibrated using the 520.5 cm-1 line of a silicon wafer. The best method of placing the kaolinites on this metal surface was to take a very small amount on the end of the spatula and then tap the crystals on to the metal surface. Further details on the spectroscopy have been published elsewhere.10-12 Spectral manipulation such as baseline adjustment, smoothing, and normalization was performed using the Spectracalc software package GRAMS (Galactic Industries Corporation, NH). Band component analysis was undertaken using the Jandel “Peakfit” software package which enabled the type of fitting function to be selected and allows specific parameters to be fixed or varied accordingly. Band fitting was done using a LorentzGauss cross-product function with the minimum number of

Langmuir, Vol. 15, No. 26, 1999 8789

Figure 2. DRIFT spectra of the hydroxyl stretching region of (a) low-defect kaolinite, (b) low-defect kaolinite intercalated with cesium acetate at 25 °C, (c) high-defect kaolinite, and (d) high-defect kaolinite intercalated with cesium acetate at 25 °C. component bands used for the fitting process. The Gauss-Lorentz ratio was maintained at values greater than 0.7, and fitting was undertaken until reproducible results were obtained with r2 correlations greater than 0.995.

Results and Discussion X-ray diffraction shows that upon intercalation with cesium acetate at 25 °C, expansion of the kaolinite layers occurred from 7.20 to 13.89 Å along the c axis. This value is slightly less than that found for the expansion upon intercalation with potassium acetate at 25 °C where the d spacing of 14.05 Å was obtained. The reason for this difference is attributed to the slight differences in the hydrated ionic radii of the cesium and potassium cations. The radii of the cesium and potassium cations are 2.28 and 2.34 Å.16 No expansion occurs in any other direction. The extent of intercalation of the low defect kaolinite with cesium acetate was found to be 60% and for the high defect kaolinite 65%. Thus in any consideration of the phases present, two phases must be considered, that of the nonexpanded kaolinite and that of the expanded kaolinite. Thus in the vibrational spectrum of such a mixture, the spectrum will consist of the addition of the two spectra from each of the phases. When the kaolinite was treated with cesium acetate at 120 °C and 2 bar and at 220 °C and 20 bar, the 001 d spacing showed high asymmetry on the low-angle side. The significance of this asymmetry means that there is a wider variation in d spacings in the case of the hydrothermally treated cesium acetate intercalated kaolinites. This shows some expansion along the c axis but not to that of the kaolinite intercalated with cesium acetate at 25 °C. DRIFT Spectroscopy. An excellent method for studying the modification of kaolinite surfaces through the insertion of cesium acetate into the kaolinite interlayer space and the alteration of the kaolinite surfaces is to use infrared spectroscopy. Such a method explores the changes in the surface molecular structure of the kaolinite upon intercalation. Figure 2 shows the DRIFT spectra of lowand high-defect kaolinites and their 25 °C cesium acetate intercalates. It should be noted that while the spectra of the untreated kaolinites have a flat baseline, the baseline for the DRIFT spectra of the cesium acetate intercalated kaolinites is a sloping baseline, because these spectra are superimposed on the broad spectrum of adsorbed water. Table 1 reports the band component analysis of the DRIFT spectra of the hydroxyl stretching region of the low- and high-defect kaolinites together with their cesium acetate intercalates formed at 25 °C and 1 bar, 120 °C and 2 bar, (16) Rutgers, A. T.; Hendrikx, Y. Trans. Faraday Soc. 1962, 58, 2184.

8790

Langmuir, Vol. 15, No. 26, 1999

Frost et al.

Table 1. Band Component Analysis of the DRIFT Spectra of the Hydroxyl Stretching Region of (a) High-Defect Kaolinite, (b) High-Defect Kaolinite Intercalated with Cesium Acetate and Treated at 120 °C and 2 Bar, (c) High-Defect Kaolinite Intercalated with Cesium Acetate and Treated at 220 °C and 20 Bar, (d) Low-Defect Kaolinite, (e) Low-Defect Kaolinite Intercalated with Cesium Acetate and Treated at 120 °C and 2 Bar, and (f) Low-Defect Kaolinite Intercalated with Cesium Acetate and Treated at 220 °C and 20 Bar band 1 band 2 band 3 band 4 band 5 band 6 band 7 band 8 band 9 band 10 high-defect kaolinite

center/cm-1

band band area/% high-defect, intercalated with band center/cm-1 CsAc, 25 °C and 1 bar band area/% high-defect, intercalated with band center/cm-1 CsAc, 120 °C and 2 bar band area high-defect, intercalated with band center/cm-1 CsAc, 220 °C and 20 bar band area/% low-defect kaolinite band center/cm-1 band area/% low-defect, intercalated with band center/cm-1 CsAc, 25 °C and 1 bar band area low-defect, intercalated with band center/cm-1 CsAc, 120 °C and 2 bar band area/% low-defect, intercalated with band center/cm-1 CsAc, 220 °C and 20 bar band area/%

3692 25 3691 54.3 3693 24.2 3692 25.2 3692 28.0 3695 38.4 3691.5 25 3690 18.4

3687 0.5

3683 2.5 3685 8.4

and at 220 °C and 20 bar. For the untreated low-defect kaolinite, infrared bands are observed at 3692 (ν1), 3683 (ν4), 3669 (ν2), 3650 (ν3), and 3619 (ν5) cm-1. An additional infrared band is also observed at 3550 (νH2O) cm-1, which is attributed to water molecules adsorbed on the edges and surfaces of the kaolinite crystals. The high-defect kaolinite shows bands in similar positions with an additional band at 3600 cm-1 which is assigned to the hydroxyl stretching vibration of surface interlayer water molecules. An additional band is also observed at 3627 cm-1, which makes up some 4.7% of the total band intensity. This band is attributed to the inner hydroxyls of the curved kaolinite layers.11 The curvature caused a shift in the peak position of the inner hydroxyl group. Such an observation is supported by the X-ray diffraction results where a broadening of the d(001) spacing is observed. When the kaolinites are intercalated with cesium acetate, additional bands are observed at 3606 cm-1 for the low-defect kaolinite and at 3603 cm-1 for the highdefect kaolinite. These bands are assigned to the inner surface hydroxyls hydrogen bonded to the acetate ion. There is an apparent difference between the frequency of the band between the CsAc intercalated low- and highdefect kaolinites. Lower values of the hydroxyl stretching frequencies are indicative of stronger hydrogen bonds formed between the inner surface hydroxyls and the acetate ion. It is possible that the stronger interaction occurs because the cesium cation is weakly hydrated and the acetate can from a stronger hydrogen bond with the inner surface hydroxyls in the case of the low-defect kaolinite. The two bands at 3606 and 3603 cm-1 indicate the changes in the hydroxyl surface of the expanded kaolinite. A number of changes in intensity of the ν1 to ν4 bands occur upon intercalation with cesium acetate. For the low-defect kaolinite, the ν1 band at 3695 cm-1 increases in intensity from 28% for the untreated kaolinite to 38.4% for the 25 °C cesium acetate intercalated kaolinite. The bands at 3669 and 3650 cm-1 display a decrease in intensity upon intercalation. The intensity remaining in these bands is attributed to the nonexpanded kaolinite phase. Upon intercalation of the high-defect kaolinite with cesium acetate at 25 °C, an increase in intensity of the band at 3691 cm-1 is observed. It is not clearly understood why there is an increase in intensity of the ν1 band. One likely possibility is that the inner surface hydroxyls are hydrogen bonded to water and result in an increased dipole moment.

3671 5.6 3668 3.7 3667 15.2 3666 6.4 3669 12.2 3668 3.6 3667 10.9 3667 6.7

3651 12.9 3653 3.0 3647 9.1 3646 12.1 3650 26.2 3652 12.0 3649 8.0 3649.5 7.3

3627 4.7

3619 19.7 3618 12.3 3620 16.5 3619 14.9 3619 20.3 3619 14.0 3619 13.9 3618 12.6

3600 2.2 3602 24.6 3599 5.6 3598 3.0 3550 10.8 3605 23.4 3582 31.3 3582 33.9

3564 29.4 3583 2.2 3582 7.5 3582 5.5

3550 12.1 3550 16.7

3457 9.7 3471 16.0 3432 10.0

3441 20.2

It is also noteworthy that there is a significant decrease in bandwidth of the ν1 band upon intercalation. The bandwidth decreases from 21 to 16 cm-1. Such a decrease suggests that the inner surface hydroxyls are in a more well-defined structure. The position and bandwidth of the inner hydroxyl stretching frequency decreases upon intercalation with cesium acetate. The band at 3619 cm-1 for the untreated high-defect kaolinite is found at 3618 cm-1 for the intercalated kaolinite. The bandwidth decreases from 13.0 to 9.8 cm-1 for the high-defect kaolinite and from 9.2 to 7.9 cm-1 for the low-defect kaolinite upon intercalation. Such a decrease in bandwidth of the inner hydroxyl stretching frequency reflects also a better defined position for the inner hydroxyl group. It is possible that the cesium cation is sitting above the ditrigonal hole of the siloxane layer and is influencing the inner hydroxyl vibration. Often a low-intensity band at 3627 cm-1 is observed for high-defect kaolinites, and this band has been attributed to the OHin stretching vibration.11 The additional band occurs because of the folding of the kaolinite layers. This band is not observed in the spectra of the cesium acetate intercalated low-defect kaolinite. These intensity changes are not as remarkable as those that occur with the intercalation with potassium acetate.10,11 In the case of the intercalation with potassium acetate, all of the ν1 to ν4 bands are reduced to zero in intensity.13 This difference is because the kaolinites are only partially expanded upon intercalation with the cesium acetate. This assessment of the partial intercalation is in good agreement with the X-ray diffraction results. In an attempt to intercalate the kaolinites more fully with cesium acetate, hydrothermal conditions were used. Table 1 reports the band component analysis of the DRIFT spectra of the hydroxyl-stretching region of the low- and high-defect kaolinites thermally treated in the presence of cesium acetate. Figure 3 displays the DRIFT spectra of the hydroxyl stretching region for the low-defect and high-defect kaolinites intercalated with cesium acetate at 120 °C and 2 bar and at 220 °C and 20 bar. It is noteworthy that the infrared spectra closely resemble spectra a and c in Figure 2. The first conclusion that is made is that the hydrothermal treatment of the kaolinites with cesium acetate did not increase the degree of intercalation but rather the opposite occurred. X-ray diffraction showed no expansion to 13.9 Å. X-ray diffraction does show strong asymmetry of the original 001 d spacing. This suggests some expansion is occurring but only to a small extent. Intercalation with cesium acetate was

Modification of Kaolinite Surfaces

Figure 3. DRIFT spectra of the hydroxyl stretching region of (a) low-defect kaolinite intercalated with cesium acetate at 120 °C and 2 bar, (b) low-defect kaolinite intercalated with cesium acetate at 220 °C and 20 bar, (c) high-defect kaolinite intercalated with cesium acetate at 120 °C and 2 bar, and (d) highdefect kaolinite intercalated with cesium acetate at 220 °C and 20 bar.

incomplete. It is questionable whether any intercalation with cesium acetate is observed. Upon treatment of the high-defect kaolinite in the presence of aqueous cesium acetate, additional bands are observed in the hydroxyl-stretching region at 3599 cm-1 for both the 120 and 220 °C experiments. This band however is at a lower frequency than that (3603 cm-1) observed for the 25 °C experiment. For the low-defect kaolinite hydrothermally treated with cesium acetate, no additional band was observed at 3606 cm-1. However, an additional band was observed at 3582 cm-1. The relative intensity of this band is 31.3% for the 120 °C case and 33.9% for the 220 °C experiment. The 3582 cm-1 band is broad with a bandwidth of around 155 cm-1. Such broad bands are typical of water vibrations. The bandwidth of the hydroxyl stretching band at 3606 cm-1 for the 25 °C cesium acetate intercalated kaolinite is 13.4 cm-1. Thus the bands observed at 3582 and 3550 cm-1 are attributed to the hydroxyl stretching frequencies of water molecules. Some expansion of the kaolinite layers occurs to 8.45 Å. It is proposed that these water molecules are intercalating the kaolinite. Broad bands are also observed for the highdefect kaolinite intercalated with cesium acetate at 120 °C and 220 °C, at 3550 cm-1 and around 3457 cm-1. These two bands are attributed to surface-adsorbed water and to water hydrogen bonded to other water molecules. Important complementary information on the intercalation of kaolinite with cesium acetate may be obtained by the study of the DRIFT spectra of the hydroxyl deformation (libration) vibrations centered at 915 cm-1 (Figure 4). The hydroxyl deformation of the low-defect kaolinite is characterized by bands at 940, 923, 913, and 901 cm-1 (Table 2). The relative intensities of these bands are 30.8, 22.5, 40.9, and 5.7%. The band at 913 cm-1 is attributed to the deformation of the inner hydroxyl and the other three bands to the librations of the inner surface hydroxyl groups.17 The high-defect kaolinite shows three bands at 942, 913, and 877 cm-1 with band areas of 21.5, 72.7, and 5.8%. The low-frequency hydroxyl deformation band is attributed to free or non-hydrogen-bonded hydroxyl groups. When the low-defect kaolinite is intercalated with cesium acetate at 25 °C, additional bands at 910 and 898 cm-1 are observed. The intensity of the 940 and 923 cm-1 band decreases from 30.8 to 11.7% and from 22.5 to 15.8% respectively. The decrease in intensity is 38%, which corresponds well with the amount of nonintercalated kaolinite as determined by X-ray diffraction. The intensi(17) Frost, R. L. Clays Clay Miner. 1998, 46, 280.

Langmuir, Vol. 15, No. 26, 1999 8791

Figure 4. DRIFT spectra of the hydroxyl deformation region of (a) low-defect kaolinite, (b) low-defect kaolinite intercalated with cesium acetate at 25 °C, (c) high-defect kaolinite, and (d) high-defect kaolinite intercalated with cesium acetate at 25 °C.

ties of the 910 and 898 cm-1 bands are 11.4 and 33%. These bands are attributed to the hydroxyl groups hydrogen bonded to the acetate from the inserting molecule. The reason for two additional bands is not clear. However one possibility is that the band at 910 cm-1 is attributed to the hydroxyl groups hydrogen bonded to the water molecules and the band at 898 cm-1 to the kaolinite hydroxyls hydrogen bonded to the acetate ions. A weak infrared band from the acetate at 925 cm-1, which is attributed to the symmetric in-plane CH deformation, may interfere with band component analysis.18 However such a band is unlikely to effect the analysis in the spectral profile other than at 925 cm-1. The fact that the intensity at this point is low suggests that the band is not actually observed. Also this C-H deformation band has a bandwidth of 10 cm-1, which is a value considerably less than the bandwidths of the component bands in this analysis. When the high-defect kaolinite is intercalated with cesium acetate at 120 °C and 2 bar, bands are observed at 939, 912, 894, and 877 cm-1 with relative intensities of 21.1, 63.3, 2.3, and 13.2%. The bandwidths of these bands are 24.6, 27.8, 11.8, and 20.9 cm-1, respectively. The band at 940 cm-1 attributed to the hydroxyl deformation of inner surface hydroxyl groups hydrogen bonded to the next adjacent siloxane layer shows no change in intensity. This again suggests that the kaolinite was not intercalated. The band at 894 cm-1 with 2.4% relative intensity is the additional deformation vibration induced through intercalation. A large increase in intensity of the band at 877 cm-1 attributed to free or non-hydrogenbonded hydroxyl is observed. The value increases from 5.8% for the untreated kaolinite to 13.2% for the thermally treated kaolinite. When the high-defect kaolinite is thermally treated at 220 °C and 20 bar, similar results are observed. Some differences in the intensity of the 896 and 898 cm-1 bands are observed. X-ray diffraction shows strong asymmetry on the low-angle side after thermal treatment with cesium acetate. Such changes are in harmony with the changes in the hydroxyl stretching bands. Thus although intercalation did not occur, considerable changes in the hydroxyl deformation profile were observed. The conclusion is made that the intercalation of the high-defect kaolinite with cesium acetate has caused the realignment of adjacent kaolinite gibbsite-like and siloxane layers to try to accommodate the cesium acetate. When the low-defect kaolinite is intercalated with cesium acetate at 120 °C and 2 bar, bands are observed (18) Kaklhana, M.; Kotaka, M.; Okamoto, M. J. Phys Chem. 1983, 87, 2526.

8792

Langmuir, Vol. 15, No. 26, 1999

Frost et al.

Table 2. Band Component Analysis of the DRIFT Spectra of the Hydroxyl Deformation Region of (a) High-Defect Kaolinite, (b) High-Defect Kaolinite Intercalated with Cesium Acetate and Treated at 120 °C and 2 Bar, (c) High-Defect Kaolinite Intercalated with Cesium Acetate and Treated at 220 °C and 20 Bar, (d) Low-Defect Kaolinite, (e) Low-Defect Kaolinite Intercalated with Cesium Acetate and Treated at 120 °C and 2 Bar, and (f) Low-Defect Kaolinite Intercalated with Cesium Acetate and Treated at 220 °C and 20 Bar band 1 high-defect kaolinite high-defect, 25 °C and 1 bar high-defect, 120 °C and 2 bar high-defect, 220 °C and 20 bar low-defect kaolinite low-defect, 25 °C and 1 bar low-defect, 120 °C and 2 bar low-defect, 220 °C and 20 bar

center/cm-1

band band area/% band center/cm-1 band area/% band center/cm-1 band area/% band center/cm-1 band area/% band center/cm-1 band area/% band center/cm-1 band area/% band center/cm-1 band area band center/cm-1 band area/%

Figure 5. DRIFT spectra of the low-frequency region of (a) low-defect kaolinite, (b) low-defect kaolinite intercalated with cesium acetate at 25 °C, (c) high-defect kaolinite, and (d) highdefect kaolinite intercalated with cesium acetate at 25 °C.

at 940, 925, 913, and 896 cm-1 with relative intensities of 19.2, 31.8, 42.3, and 6.6%. When this kaolinite is intercalated with the cesium acetate at 220 °C and 20 bar, bands are observed at 940, 925, 913, and 898 cm-1. The relative intensities of these bands are 18.9, 25.7, 47.9, and 7.5%, respectively. Compared with the high-defect kaolinite intercalation experiment, a decrease in intensity of the inner surface hydroxyl deformation mode at 940 cm-1 is observed. The relative intensity for the untreated kaolinite is 30.8%, and that for the 120 °C treated kaolinite is 19.2%. This decrease in intensity suggests some reordering of the kaolinite layers is occurring. It is possible that with the incorporation of water into the kaolinite layers, some shuffling of the layers occurs with a consequential increase in the number of defect structures.13 The bandwidth of the 925 cm-1 band in this work is considerably broader. A low-intensity band is observed at 898 cm-1 with around 7% intensity. This band is attributed to the deformation modes of the inner surface hydroxyl hydrogen bonded to the acetate. The low intensity of this band indicates that in the hydrothermal treatment of the low-defect kaolinite with cesium acetate at 120 and 220 °C, very little intercalation takes place. Intercalation may also be followed in the low-frequency region, as well as in the hydroxyl stretching and deformation regions. Figure 5 shows the DRIFT spectra of the low-frequency region of the low- and high-defect kaolinites and their cesium acetate intercalates. Acetate bands are easily observed at 651 and 627 cm-1. When the kaolinites are treated with cesium acetate under hydrothermal conditions, these bands are not observed. Further very

942 21.5 940 2.4 939 21.1 937 18.0 940 30.8 940 11.7 940 19.2 940 18.9

band 2

923 21.6

923 22.5 923 15.8 925 31.8 925 25.7

band 3 913 72.7 915 37.0 912 63.3 913 68.6 913 40.9 915 27.9 913 42.3 913 47.9

band 4

899 36.1 894 2.3 896 5.0 901 5.7 910 11.4

band 5 877 5.8 867 2.9 877 13.2 878 8.4 898 33.0 898 6.6 898 7.5

Figure 6. Raman spectra of the hydroxyl stretching region of (a) low-defect kaolinite, (b) high-defect kaolinite, (c) low-defect kaolinite intercalated with cesium acetate at 25 °C, and (d) high-defect kaolinite intercalated with cesium acetate at 25 °C.

little differences in the spectra of the thermally treated kaolinites and the untreated kaolinites in the lowfrequency region are observed. When kaolinite was hydrothermally intercalated with potassium acetate, only partial intercalation was observed while complete intercalation with potassium acetate occurred at 25 °C. When kaolinites are intercalated with cesium acetate, the degree of intercalation is always less than that for the intercalation with potassium acetate. So in some ways it is not unexpected that the hydrothermal intercalation with cesium acetate did not occur. The conclusion is reached that little or no intercalation of the kaolinites with cesium acetate occurred at elevated temperatures and pressures. Such an observation is the opposite of what may have been predicted. It could be expected that if the pressure was increased, then increased intercalation might have been observed. This did not take place. Raman Spectroscopy. The Raman spectra of the lowand high-defect kaolinites together with the cesium acetate intercalated kaolinites formed at 25 °C are shown in Figure 6. The results of the band component analyses of these spectra together with the spectra of the cesium acetate intercalated kaolinites formed at 120 and 220 °C are reported in Table 4. The Raman spectra of the low-defect kaolinite shows bands at 3692, 3684, 3669, 3652, and 3620 cm-1. The relative intensities of these bands are 27.5, 22.0, 13.7, 15.9, and 20 0.5%. Low-defect kaolinites are characterized by strong bands at 3684 cm-1. The relative intensities of the 3684 and 3692 cm-1 bands can be used as a measure of the order of the kaolinite.12 No water

Modification of Kaolinite Surfaces

Langmuir, Vol. 15, No. 26, 1999 8793

Table 3. Band Component Analysis of the DRIFT Spectra of the Water Bending Region for (a) High-Defect Kaolinite, (b) High-Defect Kaolinite Intercalated with Cesium Acetate and Treated at 120 °C and 2 Bar, (c) High-Defect Kaolinite Intercalated with Cesium Acetate and Treated at 220 °C and 20 Bar, (d) Low-Defect Kaolinite, (e) Low-Defect Kaolinite Intercalated with Cesium Acetate and Treated at 120 °C and 2 Bar, and (f) Low-Defect Kaolinite Intercalated with Cesium Acetate and Treated at 220 °C and 20 Bar band 1 center/cm-1

high-defect kaolinite high-defect, 25 °C and 1 bar high-defect, 120 °C and 2 bar high-defect, 220 °C and 20 bar low-defect kaolinite low-defect, 25 °C and 1 bar low-defect, 120 °C and 2 bar low-defect, 220 °C and 20 bar

band band area/% band center/cm-1 band area/% band center/cm-1 band area/% band center/cm-1 band area/% band center/cm-1 band area/% band center/cm-1 band area/% band center/cm-1 band area/% band center/cm-1 band area/%

band 2

band 3

band 4

1613 7.2

1597 8.5 1601 26.4 1601 26.5

1570 84.3 1570 73.5 1570 73.5

1612 6.6

1595 8.1 1602 21.2 1595 20.2

1570 85.3 1575 77.3 1575 69.8

nil

nil

1684 1.5 1682 0.3

1613 9.7

Table 4. Band Component Analysis of the Raman Spectra of the Hydroxyl Stretching Region of (a) High-Defect Kaolinite, (b) High-Defect Kaolinite Intercalated with Cesium Acetate and Treated at 120 °C and 2 Bar, (c) High-Defect Kaolinite Intercalated with Cesium Acetate and Treated at 220 °C and 20 Bar, (d) Low-Defect Kaolinite, (e) Low-Defect Kaolinite Intercalated with Cesium Acetate and Treated at 120 °C and 2 Bar, and (f) Low-Defect Kaolinite Intercalated with Cesium Acetate and Treated at 220 °C and 20 Bar band 1 high-defect kaolinite high-defect, Intercalated with CsAc, 25 °C and 1 bar high-defect, Intercalated with CsAc, 120 °C and 2 bar high-defect, Intercalated with CsAc, 220 °C and 20 bar low-defect kaolinite low-defect, Intercalated with CsAc, 25 °C and 1 bar low-defect, Intercalated with CsAc, 120 °C and 2 bar low-defect, Intercalated with CsAc, 220 °C and 20 bar

band center/cm-1 band area band center/cm-1 band area band center/cm-1 band area band center/cm-1 band area band center/cm-1 band area band center/cm-1 band area band center/cm-1 band area band center/cm-1 band area

3696 39.7 3698 21.1 3698 34.3 3698 34.2 3692 27.5 3693 22.9 3693 22.3

bands were observed in the Raman spectra of the lowdefect kaolinite. In contrast, the Raman spectra of the high-defect kaolinite show bands at 3696, 3674, 3652, 3627, and 3620 cm-1 with relative intensities of 39.7, 17.5, 3.5, 198.8, and 15.3%, respectively. High-defect kaolinites are characterized by intense bands at 3696 cm-1 with no intensity at 3684 cm-1. Further high-defect or highly disordered kaolinites are characterized by a band at 3627 cm-1. This band has been attributed to the inner hydroxyl group of folded kaolinite surfaces.9 When the low-defect kaolinite is intercalated with cesium acetate at 25 °C and 1 bar, additional bands are observed at 3606 and 3598 cm-1. Only one additional infrared hydroxyl stretching frequency was observed. In the infrared spectra of the hydroxyl deformation modes, two bands were observed and these two bands may well correspond to the two Raman hydroxyl-stretching frequencies. The 3606 cm-1 band is both Raman and infrared active while the 3598 cm-1 band is Raman active only. The relative intensities of the 3606 and 3598 cm-1 bands are 37.4 and 15.7%. The intensity of the 3692 cm-1 band has diminished to zero and the 3684 cm-1 band reduced to 20%. The 3664 and 3648 cm-1 bands also show diminished intensity. The inner hydroxyl band 3620 cm-1 shifted to 3618 cm-1 and displays an increase in bandwidth from 4.9 to 5.2 cm-1. The band at 3606 cm-1 is broad with a bandwidth of 13.1 cm-1. The band at 3598 cm-1 is sharp

band 2

3683 5.4 3680 14.8 3680 16.0 3684 22.0 3684 20.0 3683 27.2 3683 25.6

band 3

band 4

band 5

band 6

3674 17.5 3664 0.6

3652 3.5 3650 4.7 3654 9.8 3652 8.6 3652 15.9 3648 5.7 3651 18.9 3651 17.8

3627 19.8

3620 15.3 3618 16.9 3620 18.0 3620 13.9 3620 20.5 3618 17.5 3620 23.3 3620 25.9

3669 13.7 3664 0.5 3668 6.9 3668 7.5

3625 17.4 3626 22.5

band 7

3606 39.1

3606 37.4

band 8 3599 4.2 3598 6.2 3598 5.6 3598 4.8 3598 15.7 3600 0.7 3600 0.9

with a bandwidth of 6.0 cm-1. A most interesting feature of the cesium acetate intercalate is the two bands observed 3606 and 3598 cm-1. One possibility is that there are two different hydroxyl groups accessible by the intercalating acetate ion. A second possibility is that the first band at 3606 cm-1 attributed to the inner surface hydroxyl group hydrogen bonded to the acetate ion while the second is attributed to the inner surface hydroxyl group hydrogen bonded to water in the intercalate. When the high-defect kaolinite is intercalated with cesium acetate, the two additional bands are again observed at 3606 and 3598 cm-1 with 39.1 and 12.0% relative intensity. These values are similar to those for the intercalated low-defect kaolinite. The band at 3696 cm-1 shifts to 3689 cm-1 upon intercalation with a decrease in intensity from 39.7 to 21.1%. A band at 3683 cm-1 is observed with 5.4% relative intensity. When the low-defect kaolinite is attempted to be intercalated with cesium acetate at 120 °C and 2 bar, no additional bands were observed at ∼3606 cm-1. An additional band was observed at 3600 cm-1, which was weak in intensity. The band at 3692 cm-1 showed a slight decrease in intensity from 27.5% to 22.9%. Some increase in intensity of the 3685 cm-1 band from 22.0 to 27.2% was observed. If the ratio of the two in-phase vibrations is used as a test for increased order, then upon intercalation with cesium acetate, this ratio has increased. This suggests

8794

Langmuir, Vol. 15, No. 26, 1999

that although intercalation did not occur, some rearrangement of the kaolinite layers on the molecular scale did take place. A slight increase in the bandwidth of the inner hydroxyl stretching frequency is observed from 4.9 to 5.2 cm-1. When the low-defect kaolinite is treated with cesium acetate at 220 °C and 20 bar, a set of results similar to those at 120 °C and 2 bar is obtained. The effect of increased temperature and pressure did not enhance the intercalation process. Such observations have been made before with the intercalation of the kaolinites with potassium acetate.2 When the high-defect kaolinite is treated with cesium acetate hydrothermally, an additional band is observed at 3598 cm-1. The relative intensity of the band is 5.6% for the 120 °C experiment and 4.8% for the 220 °C experiment. The band is broad with bandwidths of 15.4 and 14.2 cm-1. The band is similar to that observed for the nontreated kaolinites and therefore is assigned to interlayer water. Conclusions Upon intercalation of both low- and high-defect kaolinites with cesium acetate at 25 °C, additional infrared bands were identified at 3606 and 3603 cm-1 for the lowand high-defect kaolinites, respectively. Raman bands were observed at 3606 and 3598 cm-1. The 3606 cm-1 band was both Raman and infrared active, but the 3598 cm-1 band was Raman active only. These bands are attributed to the inner surface hydroxyl groups hydrogen bonded to the acetate ion. The reason two bands are observed is attributed to two molecular environments of the inner surface hydroxyl groups. It is proposed that the two infrared and the two Raman bands are not mutually inclusive. Indeed, the band at 3606 cm-1 is common to both the Raman and the infrared spectra and is assigned to the inner surface hydroxyl group hydrogen bonded to the acetate ion. The infrared band at 3603 cm-1 is also attributed to an inner surface hydroxyl group. However the Raman band at 3598 cm-1 is more likely assigned to an intercalated water molecule. When the kaolinites are hydrothermally treated at 120 °C and 2 bar and at 220 °C and 20 bar, no intercalation of the cesium acetate occurs. Upon intercalation of the high-defect kaolinite with cesium acetate, additional bands

Frost et al.

are observed in the hydroxyl-stretching region at 3599 cm-1 for both the 120 and 220 °C experiments. For the low-defect kaolinite hydrothermally treated with cesium acetate, no additional band was observed at 3606 cm-1. However additional bands were observed at 3582 cm-1. Both these bands at 3599 cm-1 for the high-defect kaolinite and at 3582 cm-1 for the low-defect kaolinite are attributed to water molecules intercalating the kaolinite layers. Changes in the hydroxyl deformation modes occur upon thermal treatment in the presence of cesium acetate. These changes infer that some changes in the interlayer spacings are occurring with consequential changes in the defect structures of the kaolinites. Clearly the kaolinites have not been fully intercalated under hydrothermal conditions. The reason for this is unclear when perhaps intercalation at elevated temperatures and pressures would be expected to be complete. The intercalation of kaolinites depends on the order of the kaolinite. The more ordered the kaolinite, the more complete is the intercalation. The hydrothermal treatment of the kaolinite causes the kaolinites to become more disordered, and therefore the cesium acetate will incompletely intercalate the kaolinites. The significance of this work rests with the industrial application of the use of cesium acetate for the delamination of kaolinite. The use of cesium acetate means that the kaolinites can be delaminated without the complete insertion of the molecule into the kaolinite interlayer region. Importantly, the use of vibrational spectroscopy has enabled the study of the modification of kaolinite surfaces to be explored. Acknowledgment. The financial and infrastructure support of the Queensland University of Technology Centre for Instrumental and Developmental Chemistry is gratefully acknowledged. Normandy Industrial Minerals Ltd is thanked for financial support through Mr. Lew Barnes, Chief Geologist. Financial support from the Hungarian Scientific Research Fund under Grant OTKA T25171 is also acknowledged. Mrs. M. L. Taylor of the School of Physical Sciences at the Queensland University of Technology is thanked for drawing Figure 1. LA981755A