NEW TECHNIQUES FOR STUDYING THE INTERCALATION OF KAOLINITES FROM GEORGIA WITH FORMAMIDE

Transcription

1 Cloys and Clay Minerals, Vol. 47, No. 3, , NEW TECNIQUES FOR STUDYING TE INTERCALATION OF KAOLINITES FROM GEORGIA WIT FORMAMIDE RAY L. FROST, DANIEL A. LACK, GINA N. PAROZ, AND TU A T. TRAN Centre for Instrumental and Developmental Chemistry, Queensland University of Technology, 2 George Street, GPO Box 2434, Brisbane Queensland 4001, Australia Abstract A new technique utilizing Raman microscopy and Fourier transform infrared (FTIR) microscopy is described. This technique uses thin films of oriented clay aggregates on glass slides suitable also for X-ray diffraction (XRD). Raman microscopy proved the most useful technique providing both better resolution of the O-stretching bands and greater spectral resolution. Kaolinites from Washington County, Georgia, with varying defect structures involving layer stacking were intercalated with formamide and additional Raman bands were observed at 3610 and 3627 cm 1. A concomitant decrease in the innersurface O band intensities at 3695 and 3685 cm t occurred. These bands are attributed to the innersurface O hydrogen bonded to the formamide molecule through the C-O group. The 3627 cm 1 band is sharp with a half width of 7.5 cm i and comprises 11% of the total normalized band area. When two additional O bands are observed at 3610 and 3627 cm ~, two C=O bands at 1674 and 1658 cm 1 are observed also. The two additional Raman inner-surface O bands were not observed in the IR spectra. owever, a band of low intensity was observed at 3590 cm -l. Models for the intercalation of formamide in kaolinites are proposed. Key Words--Formamide, ydroxyls, Infrared Microscopy, Intercalation, Kaolinite, Roman Microprobe, X-ray Diffraction. INTRODUCTION The kaolin minerals are often classified as non-expandable clays. Wada (1961) introduced a new field of kaolin research when kaolinites were expanded using potassium acetate and other organic salts. Many organic molecules of the appropriate size for insertion between the 1:1 layers were found, including molecules such as dimethylsulphoxide, hydrazine, urea, and formamide (Olejnik et al., 1970; Kristof et al., 1997). The interlayer bonding between these molecules arises from the bonding between the inner-surface O groups of the octahedral gibbsite-like sheet and the oxygen atoms of the tetrahedral sheet of the adjacent layer. Also, van der Waals-type forces contribute to interlayer bonding (Cruz et al., 1973). For molecules such as formamide to penetrate between the 1:1 layers, sufficient energy must be provided to overcome these bonding forces. One model for the intercalation process is that the 1:1 layers are solvated by the organic molecules. IR spectroscopy shows this solvation. Where the inner-surface Os are solvated, the corresponding IR bands are displaced to lower frequencies. This clearly shows the existence of bonds between the surface hydroxyls and the intercalated molecules. Changes in the IR frequencies of the O-stretching modes were used to calculate the contribution of the network of bonds to the intercalation process. For the intercalation with formamide, for example, an additional IR band at 3590 cm -1 was observed (Cruz et al., 1969). The intercalation of the kaolinite was found also to be a function of the amount of water present and the p of the solution (Olejnik et al., 1970). Raman microscopy is useful for the study of the kaolin structures (Frost and van der Gaast, 1997) and also for the determination of stacking order-disorder relationships. In their work, kaolinites were classified according to the ratio of the two types of inner-surface Os at 3685 and 3695 cm 1. A hypothesis was proposed to explain the two inner-surface O bands at 3685 and 3695 cm -1. This hypothesis used models for bonding between the inner-surface Os of one layer and the siloxane units of the adjacent 1:1 layer. There is an apparent relationship between the defect structures and the intensity of the 3685 cm -1 band. Correlation between the intensities of the Raman spectrum of the O-stretching vibrations and the degree of disorder using the inckley index (inckley, 1963) showed a linear relationship between this index and the ratio of the intensities of the 3685 and 3695 cm 1 bands (Frost and van der Gaast, 1997). Thus, Raman spectroscopy provided a novel method for studying the defects in kaolinite. Additional insights using Raman microscopy of kaolinite and the other polytypes were forthcoming (Frost et al., 1996; Frost, 1996). In these studies, the spectra of the kaolin polytypes were found to show differences related to the structural order of the polytype. The application of Raman microscopy to the study of intercalated kaolin has also proven useful (Frost et al., 1997a; Frost and Kristof, 1997; Frost et al., 1997b). A Raman band, attributed to the inner-surface O groups strongly bonded to the acetate, was observed at 3605 cm 1 for the potassium acetate intercalate with the concomitant loss of intensity in bands Copyright , The Clay Minerals Society 297

2 298 Frost, Lack, Paroz, and Tran Clays and Clay Minerals Table 1. The origin, description, and inckley index of kaolinites from Georgia. Species, inckley index Location Description Kaolinite, 0.85 Kaolinite, 0.95 Kaolinite, 0.05 Kaolinite, 0.86 Kaolinite, 1.05 Kaolinite, 0.25 Avant (Theile Kaolinite Co.) Sparta Granite, Latterite Dukes (Theile Kaolinite Co.) Buffalo Creek (English China Clay International) Dry Branch Kaolin Co. Congo Boone (English China Clay International) soft kaolinite (low-defect structure) soft kaolinite with some metahalloysite hard kaolinite (high-defect kaolinite) soft kaolinite soft kaolinite with some montmorillonite hard kaolinite (high-defect kaolinite) at 3652, 3670, 3684, and 3693 cm 1. Intercalation of halloysite resulted in a Raman pattern similar to an intercalated ordered kaolinite, and at least on a molecular level, intercalation produced an increase in order for this kaolin (Frost and Kristof, 1997). They concluded that the intercalation process resulted in a decrease in the defects of the halloysite. Intercalation caused remarkable changes in intensity of the Raman spectral bands of the low frequency region. In particular, the 935 cm-' band increased in intensity possibly providing evidence for the acetate ion being bonded to the inner-surface O group such that the acetate- O unit approached an angle of -90 ~ to the (001) face. Furthermore, the O deformation region was shown to be very sensitive to not only the polytype stacking structure but also the degree of bonding between the 1:1 layers (Frost, 1998). In particular, where bonding is interrupted through intercalation, a new band at cm ' is observed. This band was attributed to the 'free' or non-hydrogen bonded innersurface O group. Upon intercalation of kaolinite with urea, remarkable intensity changes in the O-stretching bands occur. The relative intensity of the inner-surface O groups decreases because of the loss of bonding of the inner-surface hydroxyls between the layers. Urea disrupts the interlayer bonding between the 1:1 layers by the formation of bonds between the Si-O of the siloxane sheet and the of the amine group of urea (Frost et al., 1997a). This is supported by the formation of additional N- bands of increased intensity at 3392 and 3408 cm -1 caused by the symmetric and antisymmetric stretching modes of the N- of urea being bonded to the siloxane layer. Upon intercalation, additional bands were observed at 583 and 1009 cm-' and were attributed to urea vibrational modes. Pure urea has very weak Raman bands at 133 and 172 cm -1 but stronger bands at 544 and 1007 cm 1. The band at 544 cm 1, which shifted to 583 cm 1 upon intercalation, was attributed to the N- deformation mode. The change in peak position is attributed to the formation of strong bonds between N- and Si-O. The band at 1679 cm -1, attributed to the symmetric stretching vibration of the C=O group of urea, was not observed for the intercalate. A marked increase in intensity of the 1540 cm -1 band was noted and is attributed to an increase in symmetry of the - N- bending mode as the N- groups are bonded to the silicate sheet. The relative intensities of the bands at 1176 and 1640 cm -1 decreased. Formamide, like urea, possesses both C=O and N 2 groups but one N 2 group is replaced by. The formamide molecule may interact with the 1:1 surfaces through the N2 group. In this case, the interaction should be similar to that of urea. Thus no additional O bands should occur and existing O bands would have no concomitant decrease in the inner-surface O intensities. The formamide may also react with the 1: 1 surface through the C=O group, in which case the interaction should be similar to that of potassium acetate. Thus, additional Raman bands at cm -1 would be observed. In this paper we report alterations to the structure of Georgian kaolinites by intercalation with formamide using Raman microspectroscopy. EXPERIMENTAL Kaolinites intercalated with formamide Kaolinites were collected from Washington County, Georgia, during the F4 field trip of the International Clay Conference (Pickering et al., 1997). Table 1 shows the kaolinites, their origin, and description. The kaolinites occur in sedimentary rocks of Cretaceous and Tertiary age as lenses and horizontal beds, and are derived from kaolin-rich detritus weathered from feldspathic Piedmont crystalline rocks. These kaolinites may be divided into three types of clays: "hard", "soft", and "flint". The "soft" kaolinites are relatively coarse and fracture conchoidally. The "hard" kaolinites are finer grained with an earthy texture (urst and Pickering, 1997). The flint-type kaolinites often contain minor amounts of smectites. X-ray diffraction Five grams of kaolinite were mixed with 25 cm 3 of a 10% formamide in ethanol solution. Kaolinite suspensions were placed on glass plates that produce low X-ray background, and allowed to dry at 35-40~ overnight to produce an oriented clay aggregate. These samples were also used for Raman spectroscopic analysis. XRD experiments used a Philips PW 1050 X-ray diffractometer using CoKa radiation (normal focus, 6 ~ take-off angle) operating at 35 kv and 40 ma. A 1 ~ divergence and scatter slit and a 0.2 mm receiving slit

3 Vol. 47, No. 3, 1999 Georgian kaolinite 299 ~,~ v~ Vg20 V Wavenumber/cm -1 Figure 1. Raman spectra of the O-stretching region of a series of kaolinites from the Sparta-granite laterite kaolinite intercalated with formamide at (a) 3, (b) 7, (c) 11, and (d) 15%. were used. The samples were measured in stepscan mode from 2.0 ~ with steps of 0.05 to 40 ~ and a counting time of 2 s. The inckley index (Table 1) was determined as an indicator of the defects present (inckley, 1963). Spectroscopy After XRD analysis, the glass plates containing either the kaolinite or intercalated kaolinite were analyzed using an Olympus BSM microscope, equipped with 10, 20, and 50 objective lenses. The microscope is part of a Renishaw 1000 Raman microscope system, which also includes a monochromator, a filter system, and a charged-coupled device (CCD). Raman spectra were excited by a Spectra-Physics model 127 e/ne laser (633 nm) and recorded at a resolution of 2 cm and were acquired in sections, cm -~ for 633 nm excitation. Repeated acquisitions using the highest magnification were accumulated to improve the signalto-noise ratio. Spectra were calibrated using the cm -J line of a silicon wafer. See Frost and van der Gaast (1997), Frost and Shurvell (1997), and Frost et al. (1997b) for additional details. The IR spectra were obtained using an infrared Perkin Elmer System 2000 spectrometer, operating under the Perkin Elmer Infrared data manager. The Fourier transform infrared spectrometer (FTIR) was equipped with a Perkin Elmer infrared microscope incorporating a small element g-cd-te detector cooled with liquid N2, and a computer-controlled mapping stage. Reflectance spectra were referenced against a plane Au-surface mirror. Typically, a circular aperture was used to obtain a spatial resolution of microns. A spectral resolution of 8 cm-~ was used throughout and 64 scans with a mirror velocity of 2 cm s ~ with measurement times of --8 s were added. Spectral manipulation such as baseline adjustment, smoothing, and normalization used the Spectracalc Wavenumber/em "1 Figure 2. Band-component analysis of the Raman spectrum of the O-stretching region of Sparta-granite laterite kaolinite intercalated with formamide at 13%. software package GRAMS (Galactic Industries Corporation, Salem, New ampshire). Band-component analysis used the Jandel 'Peakfit' software package. Band fitting was done using a Lorentz-Gauss crossproduct function with the minimum number of component bands used for the fitting process. The Gauss- Lorentz ratio was maintained at values >0.7 and fitting was achieved when reproducible results were obtained with squared correlations (r 2) > RESULTS AND DISCUSSION Kaolinites show a complex set of overlapping bands (Figure 1) in the Raman spectra of the O-stretching region with five bands labeled vf to vs, The bands v~ through v4 are assigned to inner-surface O. The v 5 band is assigned to an inner-o stretching-vibration. The v I and v4 modes were attributed to two in-phase stretching vibrations with differing bond symmetry (Frost and van tier Gaast, 1997). The v2 and v3 modes are the out-of-phase vibrations of the two vl and 1)4 modes (Frost and van der Gaast, 1997). Other bands in the O stretching region are labeled Vx (where x = 1, 2... ); in the case of formamide intercalation two other bands are observed at 3610 and 3627 cm -~ and labeled v 6 and v7 respectively. Clay samples containing kaolinite are often intercalated with formamide or ethylene glycol for XRD analysis to reveal the presence of expandable clays. Glycolation of these samples shows the presence of smectite via expansion of d- values to 17 A. The intercalation of formamide is also used to reveal clays such as halloysites. The Raman spectra of intercalated kaolinites from the Sparta granite deposit (Keller, 1964) with differing amounts of formamide are shown in Figure 1. XRD shows that the sample contained traces of illite, metahalloysite, and quartz. The inckley index of this kaolinite is 0.94, indicating a kaolinite with a lowdefect structure. The percentage intercalation, using XRD, was 26%. The kaolinite is only partially intercalated. After intercalation, the Raman spectra show

4 300 Frost, Lack, Paroz, and Tran Clays and Clay Minerals Table 2. Band positions, half widths, and band areas of the various kaolinites. Band center Band widths Sample cm -~ cm ~ % area Raman structure of the O-stretching region of laterite kaolinite Va2o 3555 broad 5.6 1)s ) ) ) Vl laterite kaolinite intercalated with formamide 1)2o 3552 broad 5.5 v ) ) v v v ) Raman spectrum of the O-stretching region of Avant kaolinite 1)2o 3555 broad 9.5 1) ) v v v~ Avant kaolinite intercalated with formamide 1) broad 5.5 v ) v ) ) v ) Buffalo-Creek kaolinite v ) v ) v Buffalo-Creek kaolinite intercalated with formamide 1) v v ) ) ) vl Dry-Branch kaolinite Vzo 3555 broad 2.9 v ) ) ) v Table 2. Continued. Band center Band widths Sample cm ~ cm -t % area Dry-Branch kaolinite intercalated with formamide Vmo 3552 broad 4.4 v v v v v V Vl additional peaks in the O-stretching region at and 3627 cm -1 with bandwidths of 8.9 and 8.7 cm 1. A very weak band is also observed in some spectra at 3590 cm 1. The relative percentage area of these bands depends on the degree of intercalation and varies from 3 to 11%. The band-component analysis of one spectrum for the O-stretching region is shown in Figure 2. The spectroscopic data of the O-stretching region for the untreated and intercalated kaolinite are reported in Table 2. The presence of the 3610-cm 1 band appears related to the band at 3552 cm 1 and is attributed to water as the bands increase in tandem. With intercalation, there is a decrease in the relative intensities of the inner-surface O groups at 3690 and 3698 cm -l. The appearance of other bands, together with a decrease in the band intensities of the inner-surface O groups, suggests that the formamide molecule is bonded to the inner-surface O groups. The decrease in the relative intensities of these bands is equal to the increase in intensity of the new O bands. Figure 3 illustrates spectra from formamide-intercalated kaolinite samples from the Buffalo-Creek deposit. The Buffalo-Creek sample is a second example of a "soft" kaolinite and this sample has a inckley ee i ~ ~ ~ Wavenunlber/cm -l Figure 3. Raman spectra of the O-stretching region of a series of kaolinites from Buffalo-Creek kaolinite deposit, intercalated with formamide at (a) 5, (b) 7, (c) 12, and (d) 15%.

5 Vol. 47, No. 3, 1999 Georgian kaolinite A c v g (d),e Wavenulnber/cm -1 Figure 4. Band-component analysis of the Raman spectrum of the O-stretching region of Buffalo-Creek kaolinite intercalated with formamide at 12%. index of Again, two additional Raman peaks are observed at 3610 and 3627 cm 1, the latter peak being the more intense of the two bands. The band-component analysis of the Raman spectrum of the Ostretching region is shown in Figure 4. Note that only one band is observed at 3627 cm 1 with a bandwidth of 7.2 cm 1, and is relatively sharp, suggesting that the O groups in the intercalation complex are well defined. The spectral parameters for the Raman bands for these Georgian kaolinites appear to be at variance with other kaolinites (Frost and van der Gaast, 1997). In particular, the bands attributed to the inner-surface O groups at and cm l appear at shorter wavelengths to those reported previously. Usually, these bands occur at 3685 and 3695 cm -1. The relative intensity of the cm 1 band was used as a measure of the stacking order of the kaolinite. For these Georgian kaolinites, the intensity of the 3695 cm -1 band is significantly greater than the 3685 cm 1 band. As the intercalation of the kaolinite with formamide proceeds, the band at 3685 cm -1 appears to decrease at the expense of the 3695 cm 1 band. For the "hard" kaolinites such as the Dukes and Congo-Boone kaolinites, only one additional band is observed in the kaolinite O-stretching region, the band at 3627 cm 1. Again, this observation suggests that some directional bonding is involved. These kaolinites are disordered, and it is difficult to obtain quality spectra. Figure 5 shows the IR spectra of kaolinites using the technique of IR microscopy. A comparison of this figure with Figures 1 and 3 clearly demonstrates that the Raman microprobe technique is superior for the study of kaolin minerals, particularly for thin-film samples. Furthermore, the Raman microprobe has a spot size of approximately one micron compared with the 20-micron spot size of the IR microscope, which is limited by the optics of the technique. The bandcomponent analysis of the IR microscopic reflectance spectrum of the formamide-kaolinite complex is Wavenumber/cm -1 Figure 5. Infrared reflectance microscopy of kaolinites from (a) Sarah-Duke deposit, (b) Buffalo-Creek deposit, (c) Theile- Avant deposit, (d) Sparta granite, (e) Dry Branch, and (f) Congo-Boone. shown in Figure 6. Table 3 reports the analysis of the IR microscopic spectrum of Buffalo-Creek kaolinite involving both the stretching frequencies of N- and the O-stretching frequencies. The analysis is complex because of overlapping bands, and thus, the analysis is qualitative only. Nevertheless the Raman active bands at 3627 and 3610 cm 1 are not observed in the IR spectrum. In addition, no distinctively resolved band is found at 3590 cm -1 although a band in this position is required to fit the spectral profile. owever, a weak Raman band was observed in some of the spectra at this position. Such a band was observed by Cruz et al. (1969) and Ledoux and White (1966). Nstretching bands are observed at 3257, 3334, 3374, 3441, 3474, and 3514 cm 1. There are three sets of two bands corresponding to the three types of formamide molecules involved in the intercalation process. Such complexity of N-stretching bands suggests that the intercalation of kaolinite with formamide is not simple and several models for intercalation may be proposed. These include bonding through the lone pair _J Wavenumber/cln -1 Figure 6. Band-component analysis of the infrared reflectance microspectrum of the O-stretching region of Buffalo- Creek deposit kaolinite intercalated with formamide.

6 302 Frost, Lack, Paroz, and Tran Clays and Clay Minerals Table 3. Band positions, half-widths, and band areas of the infrared microscopic spectrum of the O-stretching region of Buffalo-Creek kaolinite. Band Band Suggested band center widths Sample assignment cm ] cm ~ % area VN )N Symmetric stretching bonded VNr ~ Symmetric stretching VN Symmetric stretching bonded VN Asymmetric stretching bonded VN Asymmetric stretching Vmo Adsorbed water v 6 Band formed upon intercalation v 5 Inner O v 3 Inner-surface O v2 Inner-surface O v 4 Inner-surface O v I Inner-surface O of electrons on the N, bonding through the N2 hydrogens, and bonding through the C=O of the amide group. The significant aspect of the Raman active/ir inactive vibrations is that they are highly symmetric. To model this vibration the C=O group is linear with the O group (Figure 7a). Alternatively the A1-O... O=C- may be nonlinear as in Figure 7b and 7c. Frost and van der Gaast (1997) proposed that the two inner-surface O-stretching frequencies at 3685 and 3695 cm -I are caused by linear and bent bonding. Values for these bands in this study are 3690 and 3696 cm 1, respectively. The 3610-cm -~ band is described by a linear -bond model as shown in Figure 7a. The 3623-cm J band is less strongly bonded and is described by Figure 7b and 7c. Another possibility is that bonding occurs through the lone pair of electrons on the N (Figure 7a). This is not considered likely because for urea, intercalation occurs through the bonding of the N 2 group to the 1:1 layer. Such bond- ing would not be indicated by the formation of additional O bands in the O-stretching region of kaolinite. The IR spectra using IR microscopy of the carboxyl region at cm ] show two stretching-frequency bands for the highly-ordered kaolinites at 1674 and 1658 cm -1, and only one band for the low-ordered kaolinites at cm-l The presence of two stretching frequencies suggests two types of C=O groups present in the intercalated kaolinite. The lower frequency at cm i is associated with the higher of the O-stretching frequencies. The higher frequency of the C=O group is observed only when the O band at 3610 cm -~ occurs. These results support the proposed model, i.e., there are two topologies in which the C=O group intercalates with the kaolinite innersurface hydroxyls. When two bands are observed at 3610 and 3627 cm ~, then two C=O bands are observed also, and if only the 3627 cm 1 band is observed, then only one C=O band is observed. A strong band is found at 1595 cm 1 and this is assigned to the N bending modes. A weak band is also found at 1630 cm 1 and this band may be assigned either to a water O-bending vibration or to a second N2-bending vibration. Odeformation vibrations were found at 888, 913, 931, and 952 cm -1 for the Dry-Branch kaolinite with the 931 cm -~ band comprising 56% of the relative band intensity. If the O-deformation modes of the innersurface hydroxyls are restricted by bonding to the C=O group, then an increase in the intensity of this vibration is expected. Note that the CN-stretching vibration increased in frequency from that of pure formamide at 1300 cm -1 to 1320 cm 1. This is also indicative of a more tightly bound structure. CONCLUSIONS A new technique for analyzing kaolinite and intercalated kaolinites, using oriented clay aggregates on low-background glass slides used for XRD, in combination with Raman microscopy and IR microscopy, \.c/n2 O L B O /1\ J D i //C--N 2 " O / /1\ /1\ i //C--N2 ~c~o I, O /1\ (a) (b) (c) (d) Figure 7. Molecular models for the intercalation of kaolinite with formamide.

7 Vol. 47, No. 3, 1999 Georgian kaolinite 303 is described. A comparison was made between the two spectroscopic techniques. The Raman technique gives superior-quality spectra with higher resolution. Upon intercalation of kaolinites from Washington County, Georgia, with formamide, additional bands are found at and 3627 cm -~. These bands were found to be Raman active and IR inactive, and are attributed to inner-surface O groups coordinated to the carboxyl group of the formamide. The two bands are described in terms of the strength of the bond formed and the direction of coordination between the inner-surface O group and the C=O group. When two additional bands were observed in the O-stretching region, then two bands were found for the C=O region. ACKNOWLEDGMENTS The financial support of the Queensland University of Technology Centre for Instrumental and Developmental Chemistry is gratefully acknowledged. L. Rintoul is thanked for assistance with the spectroscopy. Normandy Industrial Minerals (Aust) Pty Ltd is thanked for financial support through L. Barnes, Chief Geologist. REFERENCES Cruz, M., Laycock, A., and White, J.L. (1969) Perturbation of the hydroxyl groups in intercalated kaolinite donor-accepted complexes. In Proceedings of the International Clay Conference Tokyo. Volume 1, L. eller, ed., Israel University Press, Jerusalem, Cruz, M. Jacobs,., and Fripiat, J.J. (1973) Interlayer bonding in kaolin minerals. In Proceedings of the International Clay Conference, 1972, CSIC, Madrid: Division de Ciencias, Madrid, Spain, Frost, R.L. (1996) The Application of Raman microscopy to the Study of Minerals. Chemistry in Australia, 63, Frost, R.L. (1998) ydroxyl deformation in kaolins. Clays and Clay Minerals, 46, Frost, R.L. and Kristof, J. (1997) Intercalation of halloysite-- A Raman spectroscopic study. Clays and Clay Minerals, 45, Frost, R.L. and Shurvell,.E (1997) Raman microprobe spectroscopy of halloysite. Clays and Clay Minerals, 45, Frost, R.L. and van der Gaast, S.J. (1997) Kaolinite hydroxyls--a Raman microscopy study. Clay Minerals, 32, Frost, R.L., Fredericks, EM., and Shurvell,.E (1996) Raman microscopy of some kaolinite clay minerals. Canadian Journal of Applied Spectroscopy, 41, Frost, R.L., Tran, T.., and Kristof, J. (1997a) FT Raman spectroscopy of the lattice region of kaolinite and its intercalates. Vibrational Spectroscopy, 13, Frost, R.L., Tran, T.., and Kristof, J. (1997b) Intercalation of an ordered kaolinite--a Raman microscopy study. Clay Minerals, 32, inckley, D.N. (1963) Variability in "crystallinity" values among the kaolin deposits of the coastal plain of Georgia and South Carolina. Clays and Clay Minerals, 11, urst, V.J. and Picketing, S.M. (1997) Origin and classification of coastal plain kaolins, Southeastern USA and the role of ground water and microbial action. Clays and Clay Minerals, 45, Keller, W.D. (1964) Processes of origin and alteration of clay minerals. In Soil Clay Mineralogy, Rich, Kunze, eds., Univ. of North Carolina, Durham, North Carolina, Kristof, J., Frost, R. L., Felinger, A., and Mink, J. (1997) FT- IR spectroscopic studies of intercalated kaolinites. Journal of Molecular Structure, 41, Ledoux, R.L. and White, J.L. (1966) Infrared studies of hydrogen bonding interaction between kaolinite surfaces and intercalated potassium acetate, hydrazine, formamide and urea. Journal of Colloid and Interface Science, 21, Olejnik, S., Posner, A.M., and Quirk, J.R (1970) Intercalation of polar organic compounds into kaolinite. Clay Minerals, 8, Pickering, S.M., urst, V.J., and Elzea, J.M. (1997) Mineralogy, stratigraphy and origin of the Georgia kaolins. Field Excursion F4. Guidebook for the Kaolin Field Trip to the Macon District. 11 th International Clay Mineral Conference, June, 1997, Ottawa, Canada, Wada, K. (1961) Lattice expansion of the kaolin minerals. American Mineralogist, 46, (Received 26 November 1997; accepted 30 September 1998; Ms )