Clay Minerals (1992) 27, 119-130

S Y N T H E S I S OF Z E O L I T E S F R O M T H E R M A L L Y A C T I V A T E D K A O L I N I T E . S O M E O B S E R V A T I O N S O N

N U C L E A T I O N A N D G R O W T H

M. M U R A T , A . A M O K R A N E , J. P. B A S T I D E AND L. M O N T A N A R O *

lnstitut National des Sciences Appliqudes (INSA) de Lyon, Groupe M(tallurgie Physique et Physique des Mat(riaux (UA CNRS N ~ 341), Laboratoire de Chimie-Physique Appliqude et Environnement, and Laboratoire de Thermochimie Mindrale, 20 Av. Albert Einstein, 69621 Villeurbanne, France, and * Politecnico di Torino,

Dipartimento dei Materiali ed Ingegneria Chimica, Corso Duca degli Abruzzi 24, 10129 Torino, Italia

(Received 9 October 1990: revised 12 July 1991)

ABSTRACT: Depending on the preparation temperature and the origin of the raw mineral, the products obtained by thermal activation of kaolinite lead, by hydrothermal treatment with sodium hydroxide at 102~ either to zeolite 4A, or to cubic zeolite P, or to a mixture of cubic and tetragonal zeolite P. The change in the type of zeolite obtained was explained on the basis of a radical change of the zeolite nucleation process due to either the presence of soluble potassium or high silica content resulting from dehydration of muscovite (an impurity in the raw-kaolinite), or from "segregation" associated with the first step of the metakaolinite to mullite transformation. These phenomena occur at a lower calcination temperature with poorly-crystallized kaolinite, and iron in such a sample promotes a sensitive decrease in the formation rate of zeolite crystals. Such results could be used to characterize the crystallinity of kaolinites.

The products obta ined by calcination or " thermal activation" of kaolinite can be used to synthesize zeolites or "molecular sieves" (Breck, 1974) by hydrothermal t rea tment (HT) in alkaline medium. The value of the molar Si/A1 ratio in metakaol in i te (a solid more reactive than kaolinite) is equal to unity, and then corresponds to the composi t ion of zeolite A. When metakol ini te recrystallizes at high tempera ture giving rise to spinel, mulli te or y-alumina structures, the side formation of reactive amorphous silica changed the Si/A1 ratio in the gel and theoret ical ly other zeolites should be obta ined, e.g. faujasite (Flank, 1970).

The aims of the present work were two-fold.

(i) to verify the existence of a thermal activation tempera ture (Tc) of kaolinite giving rise to a maximum of chemical reactivity versus sodium hydroxide solution in hydrothermal conditions (synthesis of zeolite A) . Effectively, in previous papers , it was shown that some parameters (X) directly connected with both structural disorganizat ion and chemical reactivity of metakaol ini te (Murat et al., 1989), e.g. "short range disorder" (Cd) measured by infrared spectroscopy (Bachiorrini & Murat , 1986), "disorganizat ion factor" (Fd)

measured from dissolution kinetics in dilute hydrofluoric acid (Murat & Driouche, 1988), and "dissolution enthalpy" (]AHd]) in concentrated H F (Murat et al., 1987), present a maximal value with metakaol ini te sample prepared by calcination of kaolinite at Tc ~- 720-750~ T h e same result was obta ined concerning the "pozzolanic reactivity" of metakaol ini te (same behaviour as "pozzolanas", e.g. amorphous Ital ian volcanic ashes,

�9 1992 The Mineralogical Society

120 M. Mura t et al.

TABLE 1. Values of molar ratios according to the type of zeolite required (Howell et al., 1965).

Na20/SiO2 SIO2/A1203 HzO/Na20

A 1.2-1.4 1.75-2-0 30-100 A(*) 1.3 2.0 50 X 1.4 3.0-5.0 35-50 Y 0-5 7.0-10.0 30-40

(*) Value chosen in the present work

which react with calcium hydroxide and water at ambient temperature and harden by formation of aluminosilicate hydrates, Murat & Bachiorrini, 1982; Murat et al., 1986).

(ii) To point out the role of both the crystallinity and mineralogical composition of the raw kaolinite in the kinetics of zeolite formation. Effectively, it is possible that thermal activation products of a poorly crystalline kaolinite have a chemical reactivity different from that observed for the thermal activated product of a well-crystallized one.

E X P E R I M E N T A L

Two kaolinite samples differing in their structural iron content and crystallinity (Herbillon et al., 1976; Li6tard, 1977; Cases et al., 1982), and both containing ferrimuscovite, were chosen as the raw material: well-crystallized English ANG kaolin (0-6% total iron as Fe203) and very poorly-crystallized French FON kaolin ( 1.3% total iron as Fe203).

Thermally activated samples (or dehydrated kaolin) were prepared by calcining kaolin for 3 h in air at 600-1100~ (Tc) . They were characterized by differential thermal analysis (DTA) using 600 mg of sample to observe the evolution of the exothermic peak at -950-980~ which corresponds to the formation of high-temperature phases (first step of the metakaolinite to mullite transformation).

Hydrothermal treatments were made at --100~ (true equilibrium temperature: 102~ on gels obtained by ageing for 12 h at ambient temperature, a mixture of dehydrated kaolin (265-280 g), water (2-430-2.450 1) and solid sodium hydroxide (220 g), which corresponds to the chemical composition leading to zeolite 4A (Table 1). The reactor used was similar to that previously developed by Thomas (1970) and Mange et al. (1976).

Samples were obtained at different reaction times (from 0-8 h) and the filtered and washed solid part was analysed by X-ray diffraction (XRD) using a Siemens D 500 diffractometer and Cu-Kol radiation, and observed by scanning electron microscopy (SEM) using a Hitachi S 2300 microscope, after drying at 120~ for 24 h.

R E S U L T S A N D D I S C U S S I O N

X R D analysis

XRD identification of the zeolites formed (Table 2) shows that zeolite 4A crystallizes only with dehydration products of ANG (well-crystallized kaolin) prepared at Tc up to 900~ and FON (poorly-crystallized kaolin) prepared up to Tc = 850~ With the ANG sample, only well-crystallized cubic zeolite P forms at Tc>900~ Cubic zeolite P forms with

Synthesis of zeolites from kaolinite

TABLE 2. Nature and wt% of zeolite formed versus hydrothermal treatment time.

121

Hydrothermal treatment time (h) Zeolite

Samples formed 0 1 2 4 8

ANG 600 4A 0 55.59 57.49 59.28 66.22 FON 600 4A 0 43.84 58.23 61.83 63.47 ANG 700 4A 0 65-89 69.21 70.57 72.04 FON 700 4A 0 60.77 63-19 64.40 66-37 ANG 750 4A 0 44.58 63-91 68.27 79.22 FON 750 4A 0 54.58 62.99 71.14 76.29 ANG 800 4A 0 27-47 75.53 78-03 82.02 FON 800 4A 0 63.19 67-42 70.79 74.76 ANG 850 4A 0 79.90 81-15 82.65 86.14 FON 850 4A 0 32.93 58.89 61.83 62.60 ANG 900 4A 0 0 84.59 88.86 91.19

FON 900 Pc 0 0 0 48-22 77-21 ANG 1000 Pc 0 0 25.66 78-35 93-47 FON 1000 Pc + Pt 0 14.32 43-76 63.23 79.95 ANG 1100 Pc 0 0 0 55.21 i00.00 FON 1100 Pc + Pt 0 0 32-07 54.41 79-84

Pc (cubic); Pt (tetragonal)

FON 900 sample, and a mixture of poorly-crystallized cubic (predominant) and tetragonal zeolite P forms with both FON 1000 and FON 1100 samples (Figs. 1 & 2) (differentiation between cubic and tetragonal forms is clearly shown for example by the modification of the 310 reflection as shown in Fig. 2).

Therefore for kaolinite thermally activated by calcining at Tc = 900~ the zeolite obtained by hydrothermal treatment in sodium hydroxide is quite different depending on the origin of the raw kaolinite, as pointed out in a previous paper (Murat et al., 1990). On the other hand, specific lines are also present on XRD patterns: lines of impurities in the kaolinite samples (anatase at 20 -~ 25.5 ~ for FON sample, and quartz at 20 = 26.8 ~ for both FON and A N G samples) and other low intensity lines that correspond to the presence of more or less amorphous non-reacted products.

SEM observations

There are three deductions which can be made from SEM observations of synthesized zeolites.

(i) Whatever the temperature of calcining (Tc) and the origin of the raw kaolinite, zeolite crystals do not appear separate but in the form of large agglomerates (Fig. 3a), indicating that the formation process is different from that when synthetic reactants (sodium silicate and sodium aluminate) are entirely dissolved in sodium hydroxide solution before hydrothermal treatment (Milton, 1956). This agglomeration process may start before zeolite crystallization, as shown in Fig. 3b which illustrates the microstructure of the dried reactant (dehydrated kaolinite) after hydrothermal treatment for 2 h at 102~ without any zeolite formation detectable by XRD. (ii) For a given value of Tc (e.g. 900~ Fig. 4) when zeolite 4A is formed, the evolution of

122 M. Murat et al.

332

420

321 311 3.B5

AN6 900 322

221 110 106

�9 , , A , . . . . . , . . . . .

4 2gin ~

310 FON 000

�9 . . 4 0 . . . . . . . . . 30 . . . . . . . . . 20 . . . . . . . . . 10 . . . . . . i w !

Flo, 1. XRD patterns of zeolite obtained from samples prepared at Tc = 900~ (HT for 8 h at 102~

zeolite 4A for ANG sample; cubic zeo[ite P for FON sample) showing characteristic reflections hkl, d in ~ (arrow), used for the kinetic approach and also lines for anatase (A) and quartz (Q).

the crystal habit of zeolite with increasing time of synthesis, shows the appearance of new crystals on the previously formed crystal (Fig. 4b). This phenomenon does not result from Ostwald ripening (there is no sensitive change in the crystal size), but rather from a secondary nucleation step (heterogeneous process) well known in zeolite synthesis (Plee, 1989, private communication). (iii) The microstructure of zeolite P formed with the FON sample from Tc = 900-1100~ (Fig. 5) is finer than that of zeolite P obtained with the ANG sample (Fig. 6). This observation cannot be interpreted in terms of crystallinity without referring to the XRD pattern. The value of the ratio "width at half height/height" for the line 211 for cubic zeolite P obtained from both FON 900 and ANG 1100 samples is 0-052 and 0.031, respectively. Therefore the cubic zeolite P obtained from the FON sample is effectively less well crystallized than cubic zeolite P obtained from the ANG 1100 sample (The comparison is not possible for Tc = 1100~ for both samples because the 211 line belongs to both cubic and tetragonal zeolite P.)

Kinetic approach

A semi-quantitative kinetic approach was carried out by studying the evolution, considering both the calcination temperature Tc and the reaction time for a given value of Tc, of the parameter W = 100(I/Io): I is the intensity (area under the XRD peak) of the characteristic reflections of the zeolite (line 311 at d = 3-65 A for zeolite A, and 211 at

Synthesis of zeolites from kaolinite 123

310 3.16

ANG 1100

321 211 t 4.10

2OO 110

2flinO /

310 t 1 F0N 1100

3.18 21 ~' 1 c ~-t 310 cl 4.10 3.15

321t 103"~J J 110 321 c 3.10 JR II

312 ,A I I I '

30 20 . . d l i . . ; . , . ~ . = A = ~ , . ; : . . . . . . . . 1 ( ] . . . = . .

Ft6.2. XRD patterns of zeolite obtained from samples prepared at Tc = 1100°C (HT for 8 h at 102°C; cubic zeolite P for ANG sample; mixture of cubic and tetragonal zeolite P for FON sample).

Additional lines for anatase (A) and quartz (Q).

d = 4.10/k for both cubic and tetragonal zeolite P, Fig. 1), and Io the intensity of the same reflections on standard samples (zeolite 4A from CECA Co., and zeolite P obtained with ANG 1100 sample after hydrothermal treatment for 8 h resulting in the maximum quantity of zeolite P).

The results obtained are listed in Table 2 and two examples of W = f(t) curves are given in Fig. 7 where it can be seen that the high quantity of zeolite is always obtained with the ANG sample. The results can be discussed in terms of nucleation, growth, and quantity of zeolite formed.

Nucleation. The value of the nucleation rate for formation of zeolite 4A is difficult to estimate because the zeolite is always present in samples taken after one hour of hydrothermal treatment and no samples were taken during the first hour. For formation of zeolite P, the nucleation rate, approximated by llto (to, induction period during which no zeolite is formed, see Table 2) assuming a homogeneous process at the beginning of the zeolite formation, is low with FON 900 sample and at a maximum with FON 100 sample. It is always lower with the ANG sample.

Growth. Comparison of growths can be approached from standardization of the kinetics curves, e.g. by drawing, vs. HT time, the parameter ~ = W/Wmax (W and Wmax are % zeolite formed at t = t and t = 8 h, respectively) and observation of the slope of the curves obtained.

124 M. Murat et al.

Ft6.3. (a) Aggregates of zeolite crystals (FON 750, HT for 8 h at 102~ (b) General microstructure of the gel before any XRD detectable crystalliza- tion of zeolite (FON 9(10~ HT for 2 h at

102~

FIG. 4. Crystal habit of zeolite 4A obtained from ANG 900 by HT at 102~

for (a) 2 h and (b) 8.h.

For formation of zeolite A, curves ~ = f(t) shows that growth rates are essentially the same for both F O N and A N G samples, with a sudden decrease at t = 2 h. This confirms that the greatest growth rate occurs at the beginning of the crystallization (Zhdanov, 1971). Curves are super imposed for Tc = 700~ When Tc = 800~ zeolite 4A begins to form earl ier with F O N sample (Fig. 8a) but the phenomenon is reversed for Tc = 850~

For the formation of zeolite P, the growth rate with the A N G sample is higher than with the F O N sample (Fig. 8b). This agrees with conventional kinetic results presented in Fig. 7. On the other hand, the decrease in the zeoli te reaction rate from kaolinite sample calcined at Tc = l l00~ could be due to the beginning of sintering of both FON and A N G samples; the sintering is more important with F O N sample (as in the investigation on compacted kaolinite powder by Murat et al., 1991) but this observat ion cannot explain the higher zeolite nucleation rate with this sample. Other parameters (e.g. content and kinetics of dissolution of amorphous silica) have to be taken into account and will be discussed later.

Quantity of zeolite formed. The greatest quantity of zeolite A formed at a reaction time >2 h (Fig. 7a) is not obta ined with the poorly crystallized F O N kaolin but with the bet ter crystallized A N G sample, and occurs at a higher calcining tempera ture of the raw mineral (Tc = 900~ Fig. 9). This result leads us to suppose that the paramete r which

Synthesis of zeolites from kaolinite 125

FIG. 5. Crystal habit of zeolite P obtained from FON 900 by HT for 8 h at 102~

FIG. 6. Crystal habit of zeolite P obtained fi-om ANG 1100 by HT for 8 h at 102~

100

i

W % 100

ANG 1/~ FON

I

2 4 6 a

t h

8

W%

ANG

FQN

t h

2 4 6 8

b

Flc. 7. Qoantity of zeolite formed vs. HT reaction time (in h): (a) zeolite 4A from ANG and FON samples with Tc = 800~ (b) zeolite P from ANG and FON sample with Tc = 1000~

126 M. M u r a t et al.

::: I #

#

t T

t h . I ] I i i I L

1 2 4 8 a

/I/: ,." ~ , lo>

/r t h o

i I i I �9

1 2 4 8 b

Fi6.8. Variation of ~ = W/Wmax vs. HT reaction time (in h): (a) zeolite 4A from ANG and FON sample with Tc = 700~ (full line, curves are superimposed) and 800~ (dotted line); (b) zeolite P for

Tc = 1000~ (full line) and ll00~ (dotted line) for ANG and FON samples.

W % 1 I

10C Z e o l i t e A ~ ' ~ " ~

f ~

l ' ?

Zeolite P

0 i I I I Tc ~ I .

600 700 800 900 1000 1100

FIG. 9. Quantity and nature of zeolite formed vs, Tc and time of HT. ANG sample: HT 8 h, curve 1: zeolite 4A; curve 1': zeolite P. HT 4 h, curve 2: zeolite 4A; curve 2': zeolite P. FON sample: HT 8 h,

curve 3: zeolite 4A; curve 3': zeolite P. HT 4 h, curve 4: zeolite 4A; curve 4': zeolite P.

controls the kinetics of zeoli te crystall ization is no t the s t ructural d isorganiza t ion state of metakaol in i te , which is at a m a x i m u m for samples p repared at = 720-750~ (Mura t & Bachiorr in i , 1982; Mura t & Dr iouche , 1988; Mu ra t et at . , 1989), bu t the con ten t of Fe ions passed into the so lu t ion dur ing the hydro the rmal t r ea tmen t , for Tc >750-780~ Effectively, Fe ions greatly affect the crystal l izat ion rate of zeolites, as po in ted ou t recent ly by Ball et al. (1986) and by Siffert (1990, pr ivate communica t i on ) for the synthesis of i ron silicate molecular sieves, a n d zeolites A and X f rom a smecti t ic type clay minera l , respectively.

Synthesis of zeolites from kaolinite 127

The same role of Fe ions was observed for the lowering of the crystallization rate of zeolite P synthesized from the FON sample (Fig. 7b).

From a pure kinetic point of view, the determination of both reaction order and activation energy was not possible as the experiments were performed at only one temperature.

Interpretation of the formation of zeolite P

Zeolite P (molar Si/A1 = 1.66) has been obtained from metakaolinite by HT for 3 days at 100~ after ageing of the gel for 3 days and subsequent addition of sodium silicate to the

metakaolinite-sodium hydroxide mix (p. 316 of Breck, 1974), but not in the experimental conditions adopted in the present research. On the other hand, some authors have prepared zeolite P by alkaline HT either of raw kaolinite at 400~ (p. 319 of Breck, 1974), or of a mixture of poorly crystallized halloysite and reactive silica at 100~ (p. 320 of Breck, 1974).

The formation of a P series zeolite by dissolution of heat-treated (980~ kaolinite was indicated only recently (Chakraborty & Ghosh, 1989) and reported by Jantzen (1990) after the present paper was first written.

The formation of zeolite P can be explained on the basis of different arguments.

(i) Its higher thermodynamic stability with respect to zeolites A, X and Y (p. 260-279 of Breck, 1974); (ii) The presence, in the gel, of K ions (Aiello & Barter, 1970). In the present case, K ions are derived from the dissolution of muscovite (or dehydrated muscovite) contained in both ANG and FON samples. Dehydroxylation of muscovite occurs between 700 and 1000~ and leads to formation of a spinel phase at 800~ (Eberhart, 1963) and a potassium feldspar (leucite) which probably dissolves more easily in the alkaline solution than muscovite, and brings K ions into the liquid synthesis medium. (iii) A higher silica content in the gel with respect to that necessary to obtain zeolite A or zeolite X (p. 340 of Breck, 1974), and after Barrer et al., 1959, if pure "free amorphous SiO2" or even higher SiO2-containing amorphous species were dissolving in alkaline medium, then P zeolites should form. This observation was recently verified by Jantzen (1990) for caustic dissolution of fibre glass in boiling NaOH. In the present work this increase in amorphous silica content can be correlated with the manifestation of the first step of the metakaolinite ~ mullite transformation which leads to extensive "segregation" of alumina and silica (LemMtre et al., 1975). The phenomenon is more marked with a poorly-crystalline raw kaolinite (Murat et al., 199t). This step occurs at a lower temperature (disappearance of the DTA normal exothermic peak for Tc = 900~ Table 3) with the poorly-crystallized FON sample, explaining the difference between the nature of zeolite formed from ANG and FON sample calcined at Tc = 900~ (for ANG sample, the normal DTA peak disappears only for the sample calcined at 1000~ Formation of leucite from breakdown of the structure of dehydrated muscovite could be promoted by the first step of the MK ~ M transformation.

As regards the crystallization of zeolite P only as a product of the reaction, let us suppose that either K ions or the high amorphous silica content in the calcined kaolin (or both) act on the nucleation stage during zeolite formation. Effectively, a mix of zeolite A and zeolite P was never obtained during the synthesis, and the partial transformation, at a given synthesis time, of zeolite A into zeolite P, such as illustrated on p. 386 of Breck (1974), was not observed by SEM in the present work.

128 M. Murat et al.

TABLE 3. Temperature (Tmax) of the maximum and intensity of the DTA exothermic peak for formation of high-temperature phases according to the temperature (Tc) of

preparation of samples.

Tmax (_+ 2~

Tc (~ FON sample ANG sample Intensity

600 956 974 Normal 700 954 976 Normal 750 954 976 Normal 800 956 978 Normal 850 958 980 Normal

900 956 / 984 Normal

1000 1024 1032 Very low 1100 1026 1040 Very low

On the other hand, the higher content of K (promoter of cubic zeolite P) in A N G kaolin (-~ 1.5% K20 against 0-08% in FON sample), which is compensated in the calcined FON sample by a higher silica content (also promoter of zeolite P), may also explain the high crystallinity of zeolite P.

Since the work of Breck (1974), no real relationship between synthesis variables (composition of the gel) and the symmetry of the resulting P phase (cubic or tetragonal) has been pointed out for formation of zeolite P. The presence of tetragonal zeolite P during synthesis from FON samples is probably due to a higher level of silica dissolved in the gel, resulting from a more pronounced segregation phenomenon during calcination of the kaolinite sample (formation of more silica than with A N G sample). This can be related to the particular characteristics of the FON kaolinite, as observed by Li6tard (1977). The product obtained by calcination at T/> 900~ has in particular a BET specific surface area larger than that of A N G sample obtained at the same temperatures (Murat et al., 1991), and this can facilitate the kinetics of dissolution of silica into the liquid alkaline medium, thereby explaining the higher zeolite nucleation rate. As for crystallization of zeolite 4A, the presence of Fe in the calcined sample decreases the quantity of zeolite formed, but this does not signify that Fe may not act as "promoter" for the nucleation of tetragonal zeolite P. An answer to this question needs complementary experiments on pure systems containing only Fe as the foreign cation.

Quartz and titanium dioxide (impurities contained in the raw kaolinite samples) do not seem to participate in the processes involved during zeolite crystallization, although the intensity of their XRD lines indicates a small decrease in zeolite obtained from the FON 1100 sample (Fig. 2).

C O N C L U S I O N S

The results presented show that the existence of a preparation temperature of metakaolinite for which the chemical reactivity is at a maximum in conditions of hydrothermal treatment in an aqueous sodium hydroxide medium, only applies when the raw kaolinite contains iron; then, it has a low crystallinity. On the other hand, and independent of the role of some foreign cations (K, Fe) on nucleation and growth of

Synthes i s o f zeol i tes f r o m kaol in i te 129

zeol i tes , t he resu l t s p r e s e n t e d m a y b e u s e d to def ine a s imple tes t for the r ap id

c h a r a c t e r i z a t i o n of kao l ins by syn thes i z ing a zeo l i t e f r o m the d e h y d r a t i o n p r o d u c t at

Tc = 900~ at this t e m p e r a t u r e the n a t u r e of t he zeo l i t e f o r m e d is qu i t e d i f f e ren t

d e p e n d i n g o n the crys ta l l in i ty of t he r aw m i n e r a l a n d its m u s c o v i t e c o n t e n t . A sys t ema t i c

s tudy is r e q u i r e d o n a large ser ies of kao l in i t e s of d i f f e ren t or ig in a n d crys ta l l in i ty to specify

s o m e aspec t s of t he se p r e l i m i n a r y resul ts .

A C K N O W L E D G M E N T S

We thank Dr. M. Mange, Maitre de Conference at the INSA of Lyon, for having placed the synthesis reactor at our disposal, and for discussion of the zeolite synthesis processes. We thank also the "Argiles et Mindraux Co." (Montguyon, Charentes, France) for partial financial support.

R E F E R E N C E S

AIELLO R. & BARRER R.M. (1970) Hydrothermal chemistry of silicates. Part XIV. Zeolite crystafiisation in presence of mixed bases. J. Chem. Soc., A, 1470-1478.

BALL W.J., DWYER J., GARFORTH A.A. & SMITH W.J. (1986) The synthesis and characterisation of iron silicate molecular sieves. Proc. 7th Int. Zeolite Conf. 137-144.

BACHIORRINI A. & MURAT M. (1986) Spectroscopie d'absorption infrarouge appliqude 5 la caract6risation de l'6tat d'amorphisation de la mdtakaolinite. C.R. Acad. Sci. 303, S6r. II, 1783-1786.

BARRER R.M., BAYNHAM J.W., BULnTUDE F.W. & MEIER W.M. (1959) Hydrothermal chemistry of silicates. Part VIII. Low-temperature growth of aluminosilicates, and some gallium and germanium analogues. J. Chem. Soc. I, 195-208.

BRECK D.W. (1974) Zeolite Molecular Sieves. Structure, Chemistry and Uses, pp. 313-320 and pp. 731-738. John Wiley & Sons, New York.

CASES J.M., LIF.TARD O., YVON J. & DELON J.F. (1982) Etude des propridtds cristallochimiques, morphologiques et superficielles des kaolinites ddsordonndes. Bull. Min. 105, 439445.

CHAKRABORTY A.K. & G~tOSH D.K. (1989) Comment on spinel phase formation during 980~ exothermic reaction in the kaolinite-to-mullite reaction series. J. Am. Ceram. Soc. 72, 1569-1570.

EBERHART J.P. (1963) Etude des transformations du mica muscovite par chauffage entre 700 et 1200~ Bull. Soc. Fr. Min. Cristallogr. 86, 213-251.

FLANK W.H. (1970) U.S. Patent No 3,515,511. HERB~LLON A.J., MESTDAGH M.M., VmLVOYE L. & DEROUANE E.G. (1976) Iron in kaolinite with special reference to

kaolinite from tropical soils. Clay Miner. l l , 201-220. HOWELL P.H., ACARA N.A. & TOWNE M.K. Jr. (1965) Production of molecular sieve adsorbents from kaolin

minerals. Brit. Patent No. 980,891. JANTZEN C.M. (1990) Formation of zeolite during caustic dissolution of fibreglass: implications for studies of the

kaolinite-to-mullite transformation. J. Am. Ceram. Soc. 73, 3708-3711. LEMA[TRE J., LEONARD A.J. & DELNON B. (1975) The sequence of phases in the 900-1050~ transformation of

metakaolinite. Proc. Int. Clay Conf., Mexico City, 545-552. LII~TARD O. (1977) Contribution ~ l'(tude des propridtgs physicochimiques, cristallographiques et morphologiques des

kaolins. Th~se de Doctorat es-Sciences, Univ. Nancy, France. MANGE M., MENTZEN B. & MURAT M. (1976) Proc6d6 pour la pr6paration de zdolites synthdtiques. Brevet Franfais

No. 2.359.073. MELTON R.M. (1956) Procdd6 de fabrication de mati6res absorbantes synthdtiques. Brevet Franfais No 1.117.776. MURAT M. & BACmORRINI A. (1982) Corrdlation entre l'6tat d'amorphisation et l'hydraulicit6 du mdtakaolin. Bull.

Min. 105, 543-555. MURAT M. & DmOUC~iE M. (1988) Conductimetric investigations of the dissolution of metakaolinite in dilute

hydrofluoric acid. Structural implications. Clay Miner. 23, 55~7. MURAT M., AMBRO1SE J. & PERA J. (1986) Les diff6rents proc6dds d'activation des min6raux argileux permettant

d'elaborer des liants pouzzolaniques ~ rdsistance optimale. Proc. 8th Congr. Chemistry of Cement, Rio de Janeiro, IV, 53-59.

MURAT M., AMOKRANE A. & BAST1DE J.P. (1990) Synth6se des z6olites/l partir des produits d'activation thermique

130 M. M u r a t e t al.

de la kaolinite. R61e des caract6ristiques structurales et min6ralogiques du min6ral de d6part. C.R. Acad. Sci. 310, S6r. II, 1725-1730.

MURAT M., MATHUmN D. & CHalHl EL M. (1987) Enthalpie de dissolution de diff6rentes kaolinites et m6takaol'inites dans l'acide fluorhydrique. Influence des caract6ristiques cristallochimiques. Thermoehim. Acta, 122, 79-85.

MURAT M., AMOKRANE A., MONTANARO L. & NE6RO A. (1991) Some experimental observations on the first step of the "Metakaolinite ~ Mullite" transformation. 3rd Europ. Interregion. Colloq. Ceramics, Lyon, 28-29.

MURAT M., MATHURIN D., DRIOUCHE M., BACHIORRINI A. & MONTANARO L. (1989) Investigations on some structural and physico chemical properties of metakaolinite. Abstracts 9th Int. Clay Conf., Strasbourg, 272.

THOMAS J.L. (1970) Influence de traitements thermiques prdalables sur les propridtds d'absorption des tamis moldculaires de type 4A et 5A, et sur l'dvolution structurale et texturale des tamis moldculaires de type 4A. Th~se de Doctorat es-Sciences Physiques, Univ. Lyon, France.

ZHDANOV S.P. (1971) Some problems of zeolite crystallization. Pp. 20-43 in: Molecular Sieve Zeolite~--l, Advances in Chemistry Series 101, American Chemical Society, Washington, DC.