Infrared Study of the Intercalation of Kaolinite by Caesium Bromide and Caesium Iodide

JOURNAL OF COLLOID AND INTERFACE SCIENCE ARTICLE NO.

204, 389–393 (1998)

CS985577

Infrared Study of the Intercalation of Kaolinite by Caesium Bromide and Caesium Iodide K. H. Michaelian,*,1 I. Lapides,† N. Lahav,† S. Yariv,† and I. Brodsky‡ *Natural Resources Canada, CANMET Western Research Centre, 1 Oil Patch Drive, Suite A202, Devon, Alberta, Canada T9G 1A8; †Department of Inorganic and Analytical Chemistry, The Hebrew University of Jerusalem, 91904 Jerusalem, Israel; and ‡Department of Applied Science, The Hebrew University of Jerusalem, 91904 Jerusalem, Israel Received March 9, 1998; accepted April 7, 1998

CsBr- and CsI-kaolinite intercalation complexes were synthesized by gradually heating caesium halide disks of the DMSO– kaolinite intermediate up to 3307C. Infrared spectroscopy revealed two types of complexes with the caesium salts: almost nonhydrous, obtained during thermal treatment of the DMSO complex, and hydrated, produced by regrinding the disk in air. Comparison of band positions for CsBr–kaolinite and CsI–kaolinite with those for the CsCl complex (observed in a previous study) shows that the strength of the hydrogen bond between the intercalated halide and the inner surface hydroxyl decreases on the order CsCl ú CsBr ú CsI. The nonreactivity of CsI in mechanochemical intercalation may arise from weak interaction between I 0 and inner surface hydroxyl groups, resulting from the fact that caesium is a very soft acid and iodide is a very soft base. Consequently, the very strong interaction between the two ions in the crystal is not disrupted during mechanochemical treatment. q 1998 Academic Press and Minister of Natural Resources, Canada

Key Words: kaolinite; caesium bromide; caesium iodide; alkali halide–kaolinite intercalation complex; DMSO–kaolinite intercalation complex.

INTRODUCTION

In previous studies we showed that the grinding of kaolinite with caesium chloride (1–4) and caesium bromide (1– 3) brings about the intercalation of these salts into this TO dioctahedral clay mineral. To intercalate CsCl, one may either mix the halide with kaolinite or grind this mixture in air or in the presence of water, the degree of intercalation increasing in the same sequence. For CsBr to intercalate kaolinite, it is necessary to add a few drops of water to the mixture of the two solids and to keep it wet during grinding, but the extent of intercalation is small. CsCl and CsBr intercalation complexes with kaolinite can also be prepared by evaporating aqueous clay–salt suspensions and aging the samples in an atmosphere saturated with water vapor (5). In contrast, attempts to obtain a CsI–kaolinite complex by mechanochemical treatments are not successful. 1

To whom correspondence should be addressed.

The intercalation of CsCl or CsBr together with water into kaolinite is proven by substantial modifications to the infrared spectrum (1–4) and DTA curve (1, 2, 4, 6, 7) of this important layer silicate. The most significant changes in the infrared spectrum that accompany intercalation are the diminution of bands resulting from inner surface hydroxyls (bands A, B, and C) and the appearance of a new band (A * ), which can be assigned to perturbed inner surface hydroxyl vibrations (1–4). The location of the latter band depends on the halide, whereas its relative intensity varies with thermal treatment of the complexes. Moreover, curve-fitting analysis of diffuse reflectance infrared spectra of kaolinite and CsCl– and CsBr–kaolinite intercalation complexes (3) confirms the presence of water bands whose locations depend on the halide anion, proving that these complexes contain intercalated water molecules. The existence of intercalated water is also verified by an infrared study of the replacement of this water by D2O (8). The water bands are located at higher frequencies in the bromide complex than in CsCl–kaolinite, consistent with the expected weaker hydrogen bond between the halide and the water molecule in the former case. Recently we showed by thermo-IR-spectroscopy and thermal analysis that the hydrated CsCl–kaolinite complex obtained by grinding loses some water during thermal treatment (4). With thermal evolution of intercalated water, band A * becomes weak whereas an additional band (denoted A 9 ) intensifies. It is reasonable to assume that when water is evolved, hydrogen bonds between inner surface hydroxyls and intercalated water molecules are partially replaced by bonds between these hydroxyls and intercalated chlorides. This leads to the attribution of bands A * and A 9 to perturbed inner surface hydroxyls hydrogen bonded to water and chloride, respectively. It should also be noted that band A 9 is observed in the spectrum of CsF–kaolinite (9). Interpretation of the OH stretching region is complicated by the similarity in frequencies of modified inner surface hydroxyl vibrations and those of intercalated water (4). From these observations, it is obvious that the intercalated water molecules are coordinated to the halide anion and that

389

AID

JCIS 5577

/

6g46$$$361

07-23-98 18:07:12

0021-9797/98 $25.00 Copyright q 1998 by Academic Press and Minister of Natural Resources, Canada All rights of reproduction in any form reserved.

coida

390

MICHAELIAN ET AL.

inner surface hydroxyls form hydrogen bonds with intercalated water molecules and halides. Bands resulting from Si– O stretching in kaolinite are shifted to lower frequencies, which also indicates that the inner surface oxygens form hydrogen bonds with intercalated water molecules. Infrared spectra of kaolinites, either mixed or ground (air or wet) with CsI, are similar to those of kaolinite ground without additives. The same spectrum is obtained for CsI– kaolinite mixtures even after an aging period of 3 months in the ambient atmosphere. We previously suggested (1) two possible reasons for this phenomenon: (1) I 0 is too large and cannot penetrate the interlayer space of the kaolinite, and (2) hydrogen bonds between inner surface hydroxyls and sorbed water molecules or intercalated anions, essential for intercalation, are very weak with I 0 (10). Weiss et al. (11) intercalated CsBr among other alkali halides into kaolinite indirectly by treating a hydrazine– kaolinite complex with aqueous solutions of different salts. The intercalation of the alkali halide was verified by X-ray determination of the basal spacing. Similarly, we recently used X-ray diffraction to show (12) that kaolinite intercalation complexes with alkali halides can be obtained by heating disks of the appropriate alkali halides containing kaolinite previously intercalated with dimethylsulfoxide (DMSO). During this thermal treatment, DMSO is evolved, and intercalation complexes of alkali halides are obtained. CsBr– and CsI–kaolinite complexes obtained under these conditions showed c-spacings of 1.09 and 1.17 nm, respectively. The objective of the present investigation is to analyze infrared spectra of CsBr– and CsI–kaolinite complexes so as to gain better insight into the intercalation of caesium halides in kaolinite and the structure of these specific complexes. In addition, a thermo-IR-spectroscopy study of CsBr and CsI disks of DMSO-kaolinite described here gives information on the mechanism of formation of the complexes. The CsBr complex is included because our knowledge of the infrared spectrum of dehydrated CsBr–kaolinite (unlike CsCl–kaolinite) is not complete. EXPERIMENTAL

Materials Well-crystallized Georgia kaolinite (KGa-1), supplied by Ward’s, was gently ground to 80 mesh. Suprapure grade caesium halides were supplied by Merck. DMSO was supplied by Riedel de Hahn. Preparation of DMSO– and Caesium–Halide Kaolinite Intercalation Complexes The DMSO–kaolinite complex was prepared by heating 3.0 g kaolinite in 20 ml 90/10 wt% DMSO/HOH solution for 5 h at 607C. The suspension was allowed to stand 48 h at room temperature, after which the excess solvent was

AID

JCIS 5577

/

6g46$$$362

07-23-98 18:07:12

evaporated. The air-dried solid showed a c-spacing of 1.12 nm, which is characteristic for the DMSO–kaolinite intercalation complex (13). DMSO-kaolinite samples of 50 mg were manually ground with 100 mg of CsBr or CsI (previously dried at 1057C) and pressed into 12.5-mm disks at 8 bar. The disks were heated at 50 and 1007C for 24 h and at 150, 200, 250 and 3307C for 2 h at each temperature. After each thermal treatment, the disks were repressed and their infrared spectra recorded. A parallel thermo-IR-spectroscopy study was carried out with dilute CsBr and CsI disks which contained 1 mg DMSO–kaolinite and 150-mg alkali halide. Infrared Spectra Infrared absorption spectra of both freshly prepared and thermally treated disks were recorded in Jerusalem using a Bruker IFS 113v FT-IR spectrometer. After 2 months, the treated disks were reground; diffuse reflectance and photoacoustic spectra were then obtained using a similar spectrometer in Devon. Frequencies of the water bands were determined using the curve-fitting program in Bruker OPUS software. RESULTS

Figures 1 and 2 show the OH stretching region of IR spectra of DMSO–kaolinite in CsBr and CsI disks, respectively, at room temperature; heated at 100, 200, and 2507C; and after rehydration. In the spectrum of untreated kaolinite, two strong bands appear at 3693 and 3620 cm01 (denoted A and D, respectively), with weaker bands at 3667 and 3654 cm01 (B and C). Bands A, B, and C are attributed to inner surface hydroxyls, whereas band D arises from inner hydroxyls (1–4). In spectra of DMSO–kaolinite, band A is very weak, and bands B and C are not observed. The sharp band at 3663 cm01 , which appears in the spectrum of the DMSO– kaolinite disk at room temperature or after heating at 1007C, is a perturbed inner surface hydroxyl. This modification of band A is an indication that the inner surface hydroxyls are involved in hydrogen bonds by donating protons to the SO group of intercalated DMSO (14). Bands appear at 3505 and 3540 cm01 in spectra of DMSO–kaolinite in CsBr or CsI disks at room temperature, or after heating at 1007C. These bands were also attributed to perturbed inner surface hydroxyls hydrogen-bonded to DMSO, but in a different manner (15–17). Very weak bands at 3595 and 3592 cm01 appear in the spectra of CsBr or CsI disks heated at 1007C, indicating that at this temperature replacement of DMSO by the alkali halide occurs to a very small extent. It should be noted that the reaction with CsCl was almost complete at this temperature. Band D is not expected to be affected by intercalation of DMSO; indeed, this band appears in the spectrum of DMSO–kaolinite at 3620 cm01 and is very sharp and intense.

coida

KAOLINITE INTERCALATION

FIG. 1. Infrared spectra in the 3300–3800 cm01 region for DMSO– kaolinite in CsBr disks (a) at room temperature; after heating at (b) 1007C, (c) 2007C, and (d) 2507C; and (e) after rehydration.

When the CsI disk was heated at 2007C, the bands at 3663, 3539, and 3505 cm01 disappeared, indicating that the intercalated DMSO was evolved. However, the intensity of band A increased only very slightly and was much less than that in the spectrum of untreated kaolinite. Simultaneously, two new bands appeared at 3591 and 3552 cm01 . According to our previous assignment for thermally treated CsCl-kaolinite (4), these peaks are attributable to perturbed inner surface hydroxyl vibrations and should be labeled A * and A 9. These results are characteristic for a kaolinite intercalation complex (18). One can therefore conclude that caesium iodide intercalated the kaolinite during the evolution of DMSO. Bands A * and A 9 are accompanied by shoulders at 3600, 3576, 3540, and 3517 cm01 , which can be attributed to water. At 2507C, these features become weaker. Band A 9 intensifies relative to A * and shifts to 3552 cm01 . Thus conclusions similar to those based on the thermal behavior of CsCl–kaolinite also pertain to CsI–kaolinite. During thermal treatment, water is evolved, and hydrogen bonds between inner surface hydroxyls and intercalated water molecules are partially replaced by bonds between these hydroxyls and intercalated halides. Concomitantly, band A * becomes weaker, and band A 9 intensifies. These results affirm our previous assignment of bands A * and A 9 to perturbed inner surface hydroxyls hydrogen-bonded to water and halide, respectively.

AID

JCIS 5577

/

6g46$$$362

07-23-98 18:07:12

391

FIG. 2. Infrared spectra in the 3300–3800 cm01 region for DMSO– kaolinite in CsI disks (a) at room temperature; after heating at (b) 1007C, (c) 2007C, and (d) 2507C; and (e) after rehydration.

Spectra of CsBr disks showed that DMSO was evolved only at temperatures above 2007C. The spectrum recorded after heating the disk at 2507C is characteristic for a CsBr– kaolinite complex that is almost completely dehydrated. Bands A, A *, and A 9 are located at 3693, 3592, and 3541 cm01 , respectively. The former, arising from that fraction of the clay which does not form an intercalation complex, is very weak. The second band, due to inner surface hydroxyls bound to water, is of medium intensity. The latter, arising

FIG. 3. Infrared spectra in the 400–1200 cm01 region for intercalated kaolinite after rehydration in (a) CsBr and (b) CsI.

coida

392

MICHAELIAN ET AL.

TABLE 1 Infrared Spectra of Caesium Halide–Kaolinite Complexes Complexes CsCl (4) Assignment

Symbol

Natural clay

CsBr

CsI

a

b

a

b

a

b

Al–OH deformation

E P F G H I J K L

3692 3686* 3668 3653 3623 — — — — — — — 1117 1099 1040 1013 939 918 — 756 694

3692 — — — — 3600 3576 3502 — — — — 1117 — 1032 1005 987 903 780 759 690

— 3672 — — 3616 3599 — — 3582 — 3508 3466 1111 — 1030 1009 — 903 779 760 690

3693 — — — 3619 3592 — 3541 — — — — 1112 — 1023 1002 968 903 — 759 707

3696 — — — 3618 3591 — 3552 — — — — 1113 — 1019 995 959 902 — 759 705

3697 3690** 3670** 3646** 3618** 3591 — 3560 3600** 3576** 3545** 3517** 1113 — 1020 1000 963 905 — 760 698

Al–O deformation

M

552

Si–O deformation

N O

476 434

570 — 474 443

565 510 471 440

— — — —

3694 3683** 3663** 3645** 3618 3594 — 3545 3596** 3568** 3530** 3500** 1113 — 1021 1004 977 905 — 759 703 673 559 510 469 439

— — — —

556 510? 468 440

Inner surface OH Inner surface OH Inner surface OH Inner surface OH Inner OH Inner surface OH Inner OH Inner surface OH Intercalated water Adsorbed water Intercalated water Adsorbed water Si–O stretching

A Z B C D A* D* A9

a: almost dehydrated (freshly prepared); b: rehydrated. * From micro-Raman experiment. ** Band position determined by curve fitting.

from inner surface hydroxyls bound to bromide, is very intense. Band D is located at 3619 cm01 , which is very similar to the frequency observed in the spectrum of untreated kaolinite. Positions of all the bands are summarized in Table 1. For comparison, data for CsCl–kaolinite are also included; these results are not discussed further in this paper. The CsBr and CsI disks were stored in air for 2 months and then ground. The IR spectra (Figs. 1e and 2e) showed that the water bands increased substantially relative to all others. The positions of these bands (determined by curve fitting) are summarized in Table 1. Moreover, with the adsorption of water, band A 9 becomes weaker and A * intensifies, which supports the assignment of A * and A 9 described previously. Frequencies and assignments of IR absorption bands below 1200 cm01 in spectra of freshly prepared CsBr– and CsI–kaolinite intercalation complexes obtained by thermal treatment of disks of DMSO–kaolinite, and the same samples after they had been reground and hydrated (Fig. 3), are

AID

JCIS 5577

/

6g46$$$362

07-23-98 18:07:12

summarized in Table 1 together with the corresponding bands in the spectrum of untreated kaolinite. The Si–O stretching and deformation vibrations are modified in spectra of the intercalation complexes. In every case, the bands are shifted from their positions in the spectrum of untreated kaolinite, and their shapes are changed. The data suggest that hydrogen bond formation between intercalated water molecules and atoms of the oxygen plane affect the perturbation. The two AlO–H deformation bands (designated H and I) and the Al–O deformation band (M) are also perturbed in the complexes with respect to their frequencies in the spectrum of the original kaolinite. The locations of both the former and the latter depend on the halogen. Bands H and M occur at higher wavenumbers for the bromide salt (977 and 559 cm01 ) than the iodide (963 and 556 cm01 ); consistent with this trend, band H is not detected for CsCl, where it presumably is masked by the Si–O stretching bands, and band M is located at 560 cm01 . Band I shifts to lower wave-

coida

393

KAOLINITE INTERCALATION

numbers (902–905 cm01 ) and is identifiable in spectra of both complexes.

DISCUSSION

ACKNOWLEDGMENT This study was completed when one of the authors (S.Y.) was on sabbatical at the CANMET Western Research Centre in Devon, Alberta, Canada. He is grateful to Syncrude Canada Ltd. for partial support during his stay in Canada.

REFERENCES

It has been shown previously that DMSO–kaolinite intercalation complexes can be decomposed by thermal treatment (19, 20). The present study shows that intercalated DMSO can be replaced by caesium halides by thermal treatment of a DMSO–kaolinite disk prepared from the appropriate halide. During this treatment, DMSO is evolved, and the salt diffuses inside the interlayer space. Caesium halide intercalation complexes are obtained together with water that is initially present in the disk. At elevated temperatures some of the water is removed and almost non-hydrous intercalation complexes of caesium halides are obtained. CsI complexes, although not obtained by mechanochemical techniques, can be prepared by this indirect thermal technique; the assumption that iodide is too large and cannot intercalate kaolinite is thereby disproven. Instead, the primary reason for the inertness of CsI during mechanochemical treatment must be the weak hydrogen bonds formed between intercalated I 0 and inner surface hydroxyls, and between the anion and intercalated water, as is shown by the frequencies of band A 9 and the HOH bands, respectively. Stronger interactions are obtained with Br 0 and the strongest are obtained with Cl 0 . Another reason for the inertness of CsI in mechanochemical treatment may be the fact that the cation and anion in this salt are a soft acid and base, respectively; consequently, the interaction between the two is strong (21). Similar nonreactivity of CsI was observed during a mechanochemical study of caesium and sodium halides (22). CsCl and CsBr displayed high mechanochemical reactivity, probably due to shearing of crystallographic planes, forming hydrated solid solutions and double salts with sodium halides. Analogous complexes were not obtained with CsI, and this was related to the fact that this crystal was not disrupted during the grinding process (23– 25). In conclusion, the mechanochemical reactivity of the caesium halides and their intercalation ability decreases with increasing basic softness of the anion. On the other hand, the thermal diffusion of all caesium halides enables their intercalation at temperatures above 1507C.

AID

JCIS 5577

/

6g46$$$362

07-23-98 18:07:12

1. Michaelian, K. H., Yariv, S., and Nasser, A., Can. J. Chem. 69, 749 (1991). 2. Yariv, S., Nasser, A., Deutsch, Y., and Michaelian, K. H., J. Thermal Anal. 37, 1373 (1991). 3. Michaelian, K. H., Friesen, W. I., Yariv, S., and Nasser, A., Can. J. Chem. 69, 1786 (1991). 4. Yariv, S., Nasser, A., Michaelian, K. H., Lapides, I., Deutsch, Y., and Lahav, N., Thermochim. Acta 234, 275 (1994). 5. Lapides, I., Yariv, S., and Lahav, N., Intern. J. Mechanochem. Mechanical Alloying 1, 79 (1994). 6. Yariv, S., Mendelovici, E., and Villalba, R., ‘‘Thermal Analysis,’’ Proc. 7th Intern. Conf. Therm. Anal. Vol. 1, pp. 533–540. John Wiley, Chichester, 1982. 7. Mendelovici, E., Villalba, R., and Yariv, S., Isr. J. Chem. 22, 247 (1982). 8. Yariv, S., Int. J. Tropic. Agric. 4, 310 (1986). 9. Lapides, I., Yariv, S., and Lahav, N., Clay Miner. 30, 287 (1995). 10. Greenwood, N. N., and Earnshaw, A., ‘‘Chemistry of the Elements,’’ pp. 57–69. Pergamon Press, Oxford, 1984. 11. Weiss, A., Thielepape, W., and Orth, H., Proc. Int. Clay Conf., Jerusalem 1, 277 (1966). 12. Lapides, I., Lahav, N., Michaelian, K. H., and Yariv, S., J. Thermal Anal. 49, 1423 (1997). 13. Costanzo, P. M., and Giese, R. F., Jr., Clays Clay Miner. 34, 105 (1986). 14. Olejnik, S., Aylmore, L. A. G., Posner, A. M., and Quirk, J. P., J. Phys. Chem. 72, 241 (1968). 15. Johnston, C. T., Sposito, G., Bocian, D. F., and Birge, R. R., J. Phys. Chem. 88, 5959 (1984). 16. Lipsicas, M., Raythatha, R., Giese, R. F., and Costanzo, P. M., Clays Clay Miner. 34, 635 (1986). 17. Raupach, M., Barron, P. F., and Thompson, J. G., Clays Clay Miner. 35, 208 (1987). 18. Ledoux, R. L., and White, J. L., J. Colloid Interface Sci. 21, 127 (1966). 19. Adams, J. M., and Waltl, G., Clays Clay Miner. 28, 130 (1980). 20. Breen, C., and Lynch, S., Clays Clay Miner. 36, 19 (1988). 21. Huheey, J. E., Keiter, E. A., and Keiter, R. L., ‘‘Inorganic Chemistry,’’ 4th ed., pp. 344–355. Harper Collins College Publishers, New York, 1993. 22. Yariv, S., and Shoval, S., Appl. Spectrosc. 39, 599 (1985). 23. Severin, I., Seifert, H. J., and Yariv, S., J. Solid State Chem. 88, 401 (1990). 24. Yariv, S., Seifert, H. J., Uebach, J., and Shoval, S., J. Chem. Eng. Data 37, 219 (1992). 25. Shoval, S., Seifert, H. J., Azoury, R., and Yariv, S., J. Chem. Eng. Data 37, 224 (1992).

coida