morphology of kaolinite crystals synthesized under hydrothermal

Clays and Clay Minerals, Vol. 43, No. 3, 353-360, 1995.

MORPHOLOGY OF KAOLINITE CRYSTALS SYNTHESIZED U N D E R HYDROTHERMAL CONDITIONS SAVERIO FIORE,1 F. JAVIER HUERTAS,2 FRANCISCO HUERTAS,2 AND Josf/LINARES 2 Istituto di Ricerca suUe Argille, CNR, P.O. Box 27, 85050 Tito (PZ), Italy 2 Estaci6n Experimental del Zaidin, CSIC Profesor Albareda, 1, 18008 Granada, Spain Abstract--Scanningelectron microscopy has revealed the presence of spherical, lath and platy kaolinite in gels with SL/A1atomic ratio ranging from 1.84 to 0.76 that are hydrothermally treated under different temperature and time conditions. Hemispherical structures and excavated zones, at different stages of evolution, have been observed on the surface of the gel grains, indicating that spherical particles do not precipitate from the solution but are generated from the gels. The quantity of spherical particles depends on temperature, time and the chemical composition of the starting gel. Products from starting material with Si/A1 ~ 1 yield the highest quantity of these particles. Being metastable, sphere dissolution controls the chemistry of the solution and consequently the morphology of the precipitating crystals thus producing more elongated, curved and irregular outlines when gels with Si/AI ~ 1 are hydrotherrnally treated. Key Words--Crystal growth, Electron microscopy, Kaolinite, Laths, Morphology, Plates, Spherules, Synthesis.

INTRODUCTION A considerable body of literature is devoted to the experimental crystallization of kaolinite because of its frequency in geological environments and also because of its use in many industrial fields. Consequently, as stated in a detailed review by Van Oosterwyck-Gastuche and La Iglesia (1978), the problem of the genesis of the mineral has been examined from m a n y points of view. As it is expected, different physico-chemical conditions of the m e d i u m give rise to different morphological and textural appearances. Detailed direct observations by scanning electron microscopy by Keller (1976a, 1976b) showed that kaolinite may exhibit distinct size and morphology depending on its origin. Furthermore, throughout the last decade, research has contributed greatly to the morphological description of the mineral. In particular, besides the well known hexagonal outline, kaolinite of spherical morphology has been reported. This was synthesized for the first time by T o m u r a et al (1983) starting from non-crystalline aluminosilicate gels under hydrothermal conditions. The presence of spherical particles associated with elongated crystals and hexagonal plates, however, was previously reported in experiments of synthesis from gels under hydrothermal condition (Rayner 1962, De Kimpe et al 1964, Rodrique et al 1972), but they were not identified as kaolinite. The formation of such a peculiar morphology cannot be regarded as an unusual occurrence, because it has been reported several times (e.g., T o m u r a et a11985a, 1985b, 1993, Huertas 1991, Huertas et al 1993a). Regarding the mechanism of growth of the mineral, it was suggested (Tomura et al 1985b) that spherical particles precipitate from solution with a high degree of supersaturation at the beCopyright 9 1995, The Clay Minerals Society

ginning of the synthesis process, whereas platy kaolinite grows at relatively low degrees of supersaturation. As far as we know, conditions governing the growth of lath kaolinite in a hydrothermal system have not been fully discussed. The possibility that spherical and platy kaolinite growth occurs at different times is supported by results of a kinetic study. Huertas (1991) considered that the process of crystallization of kaolinite may not be a single step process because the expected quantity of kaolinite production based on theoretical calculations was higher than the real experimental values obtained. However, a two stage kinetic model (Huertas et al 1993b) is consistent with both experimental data and theory. It is not clear whether a relationship between spherical, lath and platy kaolinite and mechanisms of growth can be established or not. In our opinion, a possible way to contribute to the question is to embark upon a meticulous observation of synthetic products. In the present paper, the results of microscopic research card e d out on several dozens of samples are reported and possible explanation for the mechanism of growth of the mineral have been proposed. MATERIALS A N D M E T H O D S

Kaolinite samples were obtained, under hydrothermal conditions, starting from non-crystalline aluminosilicate gels. Details about the synthesis method are given in Huertas et al (1993a) and are reported in Table 1. Six sets of runs were carried out using six gel compositions and 176 samples were synthesized. The products of the runs from gels with Si/A1 from 1.84 to 0.99 were already characterized by Huertas 353

354

Clays and Clay Minerals

Fiore et al

Table 1. Synthesis condition. Gel composition (Si/A1) Solution Solid/solution Temperature (*C) Pressure Time (days)

1.84, 1.56, 1.26, 0.99, 0.84, 0.76 0.1 M KOH 2.5 g/10 ml 150, 175, 200, 225, 250 Equilibrium water pressure 0.5, 1, 2, 4, 8, 15, 30, 60, 120

(1991), using X-ray diffraction, infrared spectroscopy, and thermal analysis, and they contained up to 70 wt. % kaolinite, the only detectable crystalline phase. The products from the other two gels (Si/A1 = 0.84 and 0.76) were characterized for the present research using the same techniques. Kaolinite was present at up to 80 wt. %. It should be remarked that bohemite never occurred, even if gels o f a very high aluminium content were treated. Morphological study was carried out by scanning electron microscopy (SEM, Cambridge Stereoscan $360) coupled with an energy dispersive X-ray spectrometer (EDS, L I N K A N 10000) and by transmission electron microscopy (TEM, Zeiss 10MC). A drop o f the material, dispersed in distilled water by an ultrasonic probe, was placed on the holder and air-dried at r o o m temperature. For SEM studies, a carbon stub was used to avoid interference from the holder during EDS analyses. Samples were coated with a carbon film. For T E M examination, a microgrid covered with a Formvar film stabilized with carbon was used as a holder; the working voltage was 80 kV. RESULTS Use o f SEM gave an important contribution to understanding the mechanism o f formation and evolution o f the three morphologies observed in all the products o f the synthesis: spheroidal particles, lath or elongated crystals, and hexagonal plates. Relicts o f gel were also detected in the samples.

e--

Figure 1. Scanning electron micrograph of aggregates of spherical kaolinite. Product synthesized from gel with Si/AI 0.99, at 225~ for 721 hours. Figure 2. Transmission electron micrograph of "crushed" particles. Product synthesized from gel with Si/A1 = 1.26, at 225~ for 48 hours. Figure 3. Scanning electron micrograph of clusters of "lath" kaolinite (see text). Product synthesized from gel with Si/A1 = 1.26, at 225~ for 736 hours. Figure 4. Scanning electron micrograph of rose-shaped aggregates showing the aciculate part of the crystals connected by very thin membranes. Product synthesized from gel with Si/A1 = 1.84, at 225~ for 792 hours.

Vol. 43, No. 3, 1995

Morphology of Synthetic Kaolinite

The spheroidal particles had a well defined outline and their dimension was rather constant in all the products of the runs, ranging from 0.25 to 0.35 #m (Figure 1) in diameter. Particles exhibiting a partial spherical morphology, resembling crushed spheres, were extremely rare. Figure 2 shows the best documentable example of such an occurrence observed by TEM. The spheroidal particles were usually associated in groups composed of a very great n u m b e r of individuals. But it was much more frequent to find them as either isolated spheres or near to lath/platy crystals forming groups with them or around gel grains. Microscopic observations on the spherical particles as a whole indicated the absence of void space. Their chemical composition does not depend on the composition of the starting gel; as showed by EDS analyses, the Si/AI atomic ratio is always near 1. The lath crystals appeared in large clusters or in small rose-shaped aggregates (Figure 3), which were associated with the other two morphologies without a clear preference. The greater part of elements forming aggregates cannot be defined as lath s e n s u stricto; in fact, they have complex shapes composed of a flat central area from which rigid laths depart with no defined geometry. Sometimes the lath portions of such particles were found to be connected by thin delicate m e m branes, which gave the aggregate the flake-like appearance (Figure 4). Elongated crystals with well defined outlines were also present (Figure 5). Platy crystals exhibited pseudohexagonal outlines and formed little stacks (Figure 6), of up to 0.05 #m in thickness, or rose-shaped clusters, whose dimensions could reach 5 #m in diameter. TEM investigations also permitted us to identify the c o m m o n presence of platy crystals with well defined hexagonal outlines (Figure 7). Before the hydrothermal treatment, the starting material exhibited a morphology resembling a stack of curling leaves or a stretched sponge, although grains having a compact appearance were also present (Figures 8, 9). They were also characterized as being smaller in size and higher in aluminium content. With regard to this,

---'4

Figure 5. Transmission electron micrograph of elongate crystals of kaolinite. Product synthesized from gel with Si/A1 = 1.84, at 225"C for 48 hours. Figure 6. Scanning electron micrograph of stacks of platy kaolinite. Product synthesized from gel with Si/A1 = 0.84, at 200"C for 192 hours. Figure 7. Transmission electron micrograph of pseudohexagonal crystals of kaolinite with well defined outlines. Product synthesized from gel with Si/A1= 1.26, at 150~ for 763 hours. Figure 8. Scanning electron micrograph of gel grains with compact texture. Si/Ai = 1.84.

355

356

Fiore et al

it can be remarked that the chemical composition o f grains from the same gel may change considerably, as can be seen in Figure 9, where grains having Si/AI ratio from 1:1 up to 10:3 are pointed out. The richer in silica the gels are, the more compact the grains are. As previously stated, the gel had not been completely converted into kaolinite and the microscopic observation produced surprising results. First o f all, the grains became less spongy and as the temperature o f the runs increased they became more and more compact. Secondly, their surface was not planar but pimply, due to the presence o f many hemispherical structures which were integral parts o f the gel (Figure 10). It can be seen that there are analogies with the swelling humps documented in non-crystalline Fe-Si-Al-oxyhydroxides (Eggleton 1987) a n d in volcanic glass (Eggleton and Keller 1982, Tazaki et al 1992, Fiore 1993). The abundance o f the three morphological types depends on growing conditions. Several attempts to quantify them by image analysis were carried out, but the results were not encouraging. The ratio between spheres, laths and plates, estimated in different micrographs o f the same product m a y produce different resuits. However, it was possible to make a rough estimation and the "evolution" o f the different populations was detected in the products from the runs. The relationship between the three p a r a m e t e r s - - t e m p e r a ture, time and gel c o m p o s i t i o n - - a n d the morphologies m a y be summarized as follows. Temperature

Temperature had a direct influence on the abundance o f the three shapes o f kaolinite occurring in the synthesis products. At low temperatures, the gels were only partially transformed, with the leafy structure still remaining, but they were found to be more compact in comparison with the initial products. Surfaces frequently showed hemispherical humps or were covered by spheres. As temperature increased, the amount o f laths and plates increased and rosette aggregates and isolated hexagonal plates also appeared. It was not unc o m m o n to find spheres, rose-shaped aggregates and plates interrelated with each other. As temperature increased the number o f spheres decreased and their size did not change appreciably. ~me

As the aUocation o f time was increased, crystals formed at a given temperature grew and became less fibrous, their arrangement became more orderly and the amount o f laths and plates increased. At first, the population o f spherical particles increased but later decreased; the rate o f such a transformation seemed to depend directly on temperature and therefore at higher temperatures the same number o f particles were formed in less time.


Figure 9. Scanning electron micrograph of a gel (Si/AI = 1.54) showing grains with distinct Si/AI ratios. Gel composition

The Si/AI ratio o f the starting material exerted a control on both the abundance o f the three morphologies and on their shape. Hexagonal plates were pred o m i n a n t when Si/AI in the products from the gels are richer in A1 or richer in Si. Otherwise, lath crystals are predominant in the product with Si/AI near 1. With regard to spherical particles, it appeared that products having a Si/A1 near 1 underwent a more extensive change than Si-richer and Al-richer gels, since a greater number o f particles were formed. Regarding the lath crystals, they exhibited a tendency to bend and to form more disordered aggregations. It seems as though there were a great number o f crystallization nuclei producing a massive and disordered growth. DISCUSSION Microscopic observations carried out in the present study suggested that spherical particles o f kaolinite m a y grow directly from these gels. They started to grow as small vesicles, subsequently their size increased until they detached from the surface. This would explain why spherical kaolinite is not formed under hydrothermal conditions using amorphous calcium silicate and aluminum chloride (Tomura et al 1993) and why spheres frequently occurred around gels, as was also reported by previous authors (e.g., Rodrique et al 1972, T o m u r a et al 1983, 1985b). A n obvious question then arises: H o w are spheres formed from the gel? It is likely that the growth o f such structures cannot be seen as a mechanism o f exsolution from a m e d i u m with a relatively high density (the gel) i m m e r s e d into another o f a lower density (the solution), as a consequence o f a rise in temperature. In fact, spherical particles always have a Si/A1 ratio close to 1, irrespective o f the chemical composition o f the starting gel, while an exsolution mechanism m a y give origin to products with a distinct Si/A1 ratio depending on gel composition. Likewise, it

Vol. 43, No. 3, 1995


357

\l

ll

t= 1 ~

~

t= 2

-- ~.C~ "~'-'-...Si:A1 ~- 1 "Si:A1 ~ 1

t=3

~

~

.

Figure 10. Scanning electron micrograph of gel grains exhibiting spherical humps on the surface. Product synthesized from gel with Si/Al = 0.99, at 225~ for 721 hours.

Figure 11. Schematic drawing showing the hypothesized mechanism of formation of spherical kaolinite (see text).

is not possible to hypothesize on a mechanism as suggested by Eggleton (1987) to explain the origin of hollow packed spheres in non-crystalline Fe-Si-Al oxyhydroxide gels, because the spheres synthesized during the hydrothermal experiments do not contain void spaces. It is possible, however, that the mechanism could be similar to this (see Figure 11). During the early stage of runs a local and partial arrangement of rows of O-SiO-AI-OH chains occur and kaolinite domains are formed. The formation of such domains might be favoured by the contraction of the gel, as it is documented by the change of leafy grains into compact grains (see particularly Figures 8 and 10). It is unlikely that the growing spherical particles were non-crystalline, or that the formation of a kaolinite structure occurred later, because in such a case a large variation in the chemical composition of the spherical particles should be expected. O n the other hand, the formation of spherical structures in non-crystallinematerials, such as volcanic glasses, is not u n c o m m o n and their presence has been documented by various authors, whatever the origin of the glass (Eggleton and Keller 1982, Tazaki et al 1989, 1992). The reason why the kaolinite takes on a spherical shape is still unclear. It was suggested (Tomura et al 1983, 1985a) that the spherical morphology was the result o f a spherulitic aggregation of crystals. TEM observations (Figure 12) showed that the spherical particles have an undulated surface and light areas near the surface can be observed which might indicate the existence of round, less dense zones within the particle. It suggests that the mechanism might consist in the growth of domains with a kaolinite-like structure. The arrangement of groups of such domains might tend to be spherical, because this is the geometry corresponding to the m i n i m u m surface and the m a x i m u m volume. Unfortunately, we do not have data to confirm or reject

such a speculation and further investigations should be carried out to deal with this topic. As seen previously, the a m o u n t of spheres formed during the experiments was a function of the chemical composition of the gel; in particular m a n y more particles were observed when the starting material had a Si/A1 ratio near 1. This observation apparently contrasts with the result of Rodrique et al (1972) who affirmed that " . . . genesis of kaolin minerals is the easiest when the starting gels are characterized by a high silica content," but it should be taken into account that this affirmation was related to kaolinite with a well defined shape and not to spherical kaolinite. In any case, the reasons why the spherical particles a r e specially favoured in gels with Si/A1 ~ 1 are unclear. Recently Satokawa et al (1994) found that the structure of the gel influenced the morphology of kaolinite. Since gels with distinct Si/A1 ratios have different structures (Cloos et a l 1969), a role should be played by the chemical composition of the gel. It must be emphasized,

Figure 12. Transmissionelectron micrograph of some spheres showing their undulated surface (see text). Product synthesized from gel with Si/A1 = 0.99, at 225~ for 721 hours.

Fiore et al

358


elS .

"%

. . . . . . , , . - .

_

.............-

-

9

higher supersaturation

..:. :;::/:.

precipitation of lath~platy kaolinite

:iii~i~:iii

. . . . . . . .

._g ~

~k~

ere

release of spherical kaolinite :~

d,

Gel is dissolved, became more compact as temperature and/or time increases. Hemispherical humps appear on the surface.

Platy and/or lath kaolinite precipitate from solution, dependant on the degree of supersaturation. Spherical kaolinite is released from gel; its quantity is controlled by the gel composition.

:

))

lower supersaturation

Different quantity of dissolved spherical kaolinite determine Si-AI concentration in the solution. Growth of lath kaolinite is favoured from a relatively higher concentration whereas platy kaolinite precipitates when concentration is lower.

Figure 13. Scheme of formation of spherical, lath and platy kaolinite from Si-A1 gel under hydrothermal conditions. however, that the difference in behavior o f gel as a function o f its chemical composition m a y explain the results o f the kinetic study by Huertas et al (1993b), who calculated that the activation energy o f the first stage o f the process is not constant, but varies from 86.2 kJmol -~ for Si/A1 = 0.99 up to 117.5 kJmol - t for Si/A1 = 1.84. In a previous study on the growth o f kaolinite starting from gels with Si/AI = 1 (Tomura et al 1985b), the influence o f the solid/water ratio (S/W) on the precipitation o f spherical and platy kaolinite was suggested. Although in the present study S/W ratio was constant, in the light o f the data now acquired, speculative suggestions could be possible. During the earlier stage o f the runs, Si and AI are released from the gel to the solution and platy kaolinite m a y precipitate; at the same time, spherical kaolinite grows from the gel. I f S / W is high, the dissolution rate o f the gels is high (Lasaga 1981), supersaturation o f the solution is reached in a short time and only a few platy crystals can precipitate because o f the unfavourable chemical condi-

tions in the medium. In fact, thermodynamic considerations (Van Oosterwyck-Gastuche and La Iglesia 1978) and laboratory data (Espiau and Pedro 1984) indicated that precipitation o f the hexagonal platy kaolinite is favoured from solution having a low concentration o f Si and AI. At the same time, a great number o f spherical particles occur in the products o f the runs because their quantity does not depend on the chemical composition o f the solution. On the contrary, i f S / W is low, the degree o f supersaturation unfavourable for precipitation o f platy kaolinite, is reached over a larger time period and more platy crystals m a y precipitate. The proposed mechanism is also consistent with the suggestion by T o m u r a et al (1985b) that the presence o f bohemite in the product o f the run is at a very low S/W ratio. In fact, when S/W ratio is very low the saturation o f the solution is low enough to allow the precipitation o f kaolinite, whereas bohemite crystallizes instead o f kaolinite (Tsuzuki 1976, Walter and Hegelson 1977). According to T o m u r a et al (1985b), the evolution o f

Vol. 43, No. 3, 1995


/.

/,"

decreases

~

/~

xx

pLaitS

-.

j

~

spheres

9 /

// /'x "%.

1

~.

inca~eases

Si/AI atomic ratio

Figure 14. Scheme of the distribution of morphological types as a function of the gel composition: at Si/A1 ratio near 1 more spherical particles and lath crystals are formed; at Si/ A1 far from 1 the quantity of the platy crystals is highest.

the sphere population with time and temperature is an indicator of the instability of the spherical kaolinite. It probably dissolves as runs progress because there is no evidence of the transformation into laths and/or plates through a solid state process. The formation and the subsequent dissolution of the spherical particles is followed by crystallization of lath and/or platy crystals from the solution. This affirmation is consistent with Huertas et a! (1993b), who suggested that the formation of kaolinite from aluminosilicate gels is a two stage process. The first stage might correspond to the formation of spherical domains whereas the second two crystals having platy or lath morphology. Microscopic observations showed that the habit is not casual but is related to the chemical composition of the starting material: lath kaolinite is more frequent in products from gels with kaolinitic composition (Si/AI ~ 1), while platy kaolinite is preferentially formed in products from gels having a composition far from kaolinite. Since crystal morphology is controlled by chemistry of the solution (see Sunagawa 1987, and references therein), it can be deduced that the solution in contact with the gel with kaolinitic composition have a higher chemical potential difference than the solution in contact with both Si-rich and Al-rich gels. Gel composition controls the chemistry of the solution but it is likely that also the quantity of dissolved spherical particles, which are more a b u n d a n t in the earlier products from gel with Si/A1 ~ 1, plays a role in controlling the chemistry of the solution. CONCLUSION Mechanisms governing the transformation of aluminosilicate gels into kaolinite are summarized in Figures 13 and 14 and the main features may be summarized as follows: - - T h e overall reaction is the result of two different stages. During the first, the arrangement of the gel takes place and kaolinite-like materials are formed. The sec-

359

ond stage corresponds to the precipitation of kaolinite crystals from the solution. --According to other authors, during the first stage of the transformation of gel under hydrothermal conditions, spherical kaolinite is formed. It is likely that spherical domains are formed and grow within the gel and later they are detached by dissolution of the gel. - - T h e a m o u n t of the spherulitic particles occurring in the product of the runs depends on the composition of the starting gel: the highest a m o u n t of spheres occurs in the products from gel with Si/AI near 1. - - T h e spherical particles are unstable and dissolve at a rate dependent on temperature and time. Their dissolution influences the chemistry of the solution which, in turn, controls the morphology of the mineral: more elongated, curved and irregular shaped crystals precipitate from solutions with a high chemical potential difference. As the highest a m o u n t of spheres are formed from gels with Si/A1 ~ l, in these runs lath crystals are more frequent than hexagonal crystals. ACKNOWLEDGMENTS This study has been supported by D G C Y T (Project PB 88-093) and CNR. The C N R is also acknowledged for financing the SEM laboratory at the Dipartimento Geomineralogico (University of Bari, Italy), whose facilities were used in the present work. TEM observations were done at Centro de Instrumentaci6n Cientifica of the University of Granada (Spain). The authors are grateful to R. E. Ferrel Jr. and the two referees, J. Elzea and K. Tomita, for the revision of the m a n u script. R. Gautier is acknowledged for the revision of the English. REFERENCES Cloos, P., A. J. L6onard, J. P. Moreau, A. Herbillon, and J. J. Fripiat. 1969. Structural organization in amorphous silico-aluminas. Clays & Clay Miner. 17: 279-287. De Kimpe, C., M. C. (3astuche, and G. W. Brindley. 1964. Low temperature synthesis of clay minerals. Am. Mineral. 49: 1-16. Eggleton, R.A. 1987. Non crystalline Fe-Si-Al-oxyhydroxides. Clays & Clay Miner. 35: 29-37. Eggleton, R. A., and J. Keller. 1982. The palagonitization of limburgite glass--A TEM study. N. Jb. Miner Mh. Jg. 1982: 321-336. Espiau, P., and (3. Pedro. 1984. Comportement des ions aluminiqueset de la silice en solution: etude de ha formation de la kaolinite. Clay Miner. 19: 615-627. Fiore, S. 1993. The occurrences of smectite and illite in a pyroclastic deposit prior to weathering: Implications on the genesis of 2:1 clay minerals in volcanic soils. Appl. Clay Sci. 8: 249-259. Huertas, F.J. 1991. Sintesis hidrotermal de caolinita. Estudio cinetico. Ph.D. thesis. University of Granada, 211 PP. Huertas, F. J., F. Huertas, and J. Linares. 1993a. Hydrothermal synthesis of kaolinite: method and characterization of synthetic materials. AppL Clay Sci. 7: 345-356. Huertas, F. J., F. Huertas, and J. Linares. 1993b. A new approach to kinetics of kaolinite synthesis. Proc. 4th Int.

360

Fiore et al

Symposium on Hydrothermal Reaction. Nancy 1993.8790. Keller, W . D . 1976a. Scan electron micrograph of kaolins collected from diverse environments of origin--I. Clays & Clay Miner. 24:107-113. Keller, W . D . 1976b. Scan electron micrograph of kaolins collected from diverse environments oforigin--II. Clays & Clay Miner. 24: 114-117. Lasaga, A . C . 1981. Rate laws of chemical reaction. In Kinetic of GeochemicalProcesses. Reviews in Mineralogy, vol. 8, A. C. Lasaga and R. J. Kirkpatrick, eds. Washington, D.C.: Mineral. Soc. Amer., 1-68. Rayner, J.H. 1962. An examination ofthe rate offormation of kaolinite from a co-precipitated silica gel. Colloque International C.N.R.S. sur "Gen~se et synth6se des argiles. Paris, 123-127. Rodrique, L., G. Poncelet, and A. Herbillon. 1972. Importance of silica subtraction process during the hydrothermal kaolinitization of amorphous silico-aluminas. Proc. Int. Clay Conf., Madrid 1972, 187-198. Satokawa, S., Y. Osaki, S. Samejima, R. Miyawaki, S. Tomura, I. Shibasaki, and Y. Sugahara. 1994. Effects of the structure of silica-alumina gel on the hydrothermal synthesis of kaolinite. Clays & Clay Miner. 42: 288-297. Sunagawa, I. 1987. Morphology of minerals. In Morphology of Crystals. I. Sunagawa, ed. Terra Sci. Pub. Co., Tokyo, 509-587. Tazaki, IC, W. S. Fyfe, and S. J. van der Gaast. 1989. Growth of clay minerals in natural and synthetic glasses. Clays & Clay Miner. 37: 348-354. Tazaki, K., T. Tiba, M. Aratani, and M. Miyachi. 1992. Structural water in volcanic glass. Clays & Clay Miner. 40: 122-127.


Tomura, S., R. Miyawaki, K. Inukai, Y. Shibasaki, M. Okazaki, S. Samejima, S. Satokawa, and M. Kamori. 1993. Formation process of kaolinite from amorphous phases. Int. Clay Conf., Adelaide 1993, P-151 (abs). Tomura, S., Y. Shibasald, H. Mizuta, and M. Kitamura. 1983. Spherical kaolinite: synthesis and mineralogical properties. Clays & Clay Miner. 31: 413--421. Tomura, S., Y. Shibasaki, and H. Mizuta. 1985a. Origin of the morphology of spherical kaolinite. Clay Sci. 6: 159166. Tomura, S., Y. Shibasaki, H. Mizuta, and M. Kitamura. 1985b. Growth conditions and genesis of spherical and platy kaolinite. Clays & Clay Miner. 33: 200-206. Tsuzuki, Y. 1976. Solubility diagrams for explaining zone sequences in bauxite, kaolin and pyrophyllite-diaspore deposits. Clays & Clay Miner. 24: 297-302. Trichet, J. 1969. Study of the structure of volcanic glass and its relation to the alteration of volcanic rocks. Proc. Int. Clay Conf., Tokyo 1969, vol. 1, L. HeUer, ed. Jerusalem: Israel University Press, 443--453. Van Oosterwyck-Gastuche, M. C., and A. La Iglesia. 1978. Kaolinite synthesis. II. A review and discussion of the factors influencing the rate process. Clays & Clay Miner. 26: 409--417. Walter, J. V., and H. C. Hegelson. 1977. Calculation of the thermodynamic properties of aqueous silica and the solubility of quartz and its polimorphs at high pressures and temperatures. Am. J. Sci. 277: 1315-1351.

(Received 5 July 1994; accepted 7 November 1994; MS. 2434)