Ind. Eng. Chem. Res. 1992,31,364-369
364
Registry No. H304P, 7664-38-2;muscovite, 1318-94-1.
Literature Cited Audrieth, L. F.; Kleinberg, J. Non-aqueous Solvents; John Wiley: New York, 1953; pp 260-273. Bailar, J. C., Jr.; Emeleus, H. J.; Nyholm, R.; Trotman-Dickensen, A. F. Comprehensive Inorganic Chemistry; Pergamon: New York, 1973;Vol. 2,pp 483. Black, C. A. Methods of Soil Analysis; American Society of Agronomy: Madison, WI, 1965;Part 2,pp 849-864, 1001-1007. BoullB, A.; Jary, R. Sur les phosphates de silicum. C. R. Hebd. Seances Acad. Sci. 1953,237,326330. Charlot, G.; Tremillon, B. Chemical Reactiom in Solvents and Melts; Pergamon: Braunschweig, Germany, 1969; pp 122, 442-444,500. Corbridge, D.E.C.; Lowe, E. J. Infrared Spectra of some Inorganic Phosphorus Compounds. J. Chem. SOC.1954,493-502. Fraser, D. G. Thermodynamic Properties of Silicate Melts. In Thermodynamics in Geology; Fraser, D. G., Ed.; D. Reidel: Boston, 1977;Chapter 15. Iler, R. K. The Chemistry of Silica; John Wiley: New York, 1979; PP 190. Jackson, M. L. Soil Chemical Analysis; Prentice Hall of India: New Delhi, 1973;pp 141-144. JCPDS. Powder Diffraction File; Joint Committee on Powder Diffraction Standards: Philadelphia, PA, 1971;card nos. 7-42, 3-319,4-0379, 5-490,10-423,11-500,15-364,18-1170,20-44. Jeffery, P. G. Chemical Methods of Rock Analysis; Pergamon: Oxford, U. K., 1970; pp 51,447-451. Kingery, W. D. Fundamental Study of Phosphate Bonding in Refractories. J. Am. Ceram. Soc. 1950,33, 239-250. Lyon, J. E.; Fox, T. V.; Lyons, J. W. An Inhibited Phosphoric Acid for Use in High Alumina Refractories. Am. Ceram. SOC.Bull. 1966,45,661-665. Mamykin, P. S.; Flyagin, F. G.; Ustyantsev, V. M. Interaction of Kaolinite with Phosphate Binders and the Effect of their Phase Transitions in C r y s W i e Quartzite. Refractories 1973,No. 9-10, 563-568.
Maason, C. R. An Approach to the Problem of Ionic Distribution in Liquid Silicates. Proc. R. SOC.London 1966,287A,201-221. Maxwell, J. A. Rock and Mineral Analysis; Interscience: New York, 1968; pp 8&101,323-425. Mellor, J. W. Comprehensive Treatise on Inorganic Chemistry; Longmans-Green: London, 1925;Vol. VI. p 990. Moellar, T. Inorganic Chemistry;Asia: Bombay, 1973;pp 321-334. Ohashi, S. Condensed Phosphates Containing Other Oxo-acid Anions. In Topics in Phosphorus Chemistry; Grayaon, M., GMith, E. J., Eds.; John Wiley: New York, 1964,Vol. Chapter V. Pinta, M. Atomic Absorption Spectrometry; Adam-Hilger: London, 1975;p 88. Ray, N. H. The Action of Phosphoric Acid on Glasses. J. Non-Cryst. Solids 1970,5,71-77. Roberta,G. J. Fe0-Kz0-P2O5 Glasses as a Source of Micronutrient in Soil. Am. Ceram. SOC.Bull. 1975,54,1069-1071. Robinson, P.; McCartney, E. R. Subsolidus Relations in the System 1964,47,587-593. Si02-Alz03-Pz06.J. Am. Ceram. SOC. Thilo, E. Condensed Phosphates and Arsenates. In Advances in Inorganic and Radiochemistry; Emeleus, H. J., Sharpe, A. G., Eds.; Academic: New York, 1962;Vol. 4,Chapter 1. Toop, G. W.; Samis, L. S. Activities of Ions in Silicate Melts. Trans. AIME 1962,224,878-887. Van Wazer, J. R. Phosphorus and Its Compounds; Interscience: New York, 1966; Vol. I, pp 717-800. Van Wazer, J. R.; Griffith, E. J.; McCullough, J. F. Analysis of Phosphorus Compounds. Anal. Chem. 1954,26,1755-1759. von b y , T.Ultrarot Absorption von AlP04und Si02in Abhangig keit von Fehlordnung und Temperatur. 2. Krist. 1966, 123, 263-314. Waltere, H.V. Corrosion of a Borosilicate Glasa by Orthophosphoric Acid. J. Am. Ceram. SOC.1983,66,572-574. Zamyatin, S. R.; Mamykin, P. S.; Knyazeva, T. P. Investigations on Clay-phosphate Bindings by Various Methods. J. Appl. Chem. (USSR) 1972,45,988-991.
Received for review November 19, 1990 Revised manuscript received July 18,1991 Accepted August 13, 1991
Wet Milling of Alumina and Preparation of Slurries for Monolithic Structures Impregnation Vasso Blachou, Diogenia Goula, and Constantine Philippopoulos* Laboratory of Chemical Process Engineering, Department of Chemical Engineering, National Technical University of Athens, GR-157 73 Athens, Greece
The wet grinding of alumina with HC1 for forming stable dispersions was investigated in a laboratory ball mill. Acid concentrations in the range 1-10 wt % of dry alumina do not influence the wet-milling product. The viscosity of the alumina slurries produced from the wet-milling process is strongly dependent on pH while minimum viscosity was achieved a t pH 3.5-3.8. Impregnation of ceramic monolithic substrates with alumina dispersions of varying particle size distributions affects the pore volume distribution of the alumina coating in the intermediate pore radius region.
Introduction Thin-walled honeycombed substrates find extensive use as catalyst supports in catalytic exhaust converters. To be useful in catalytic reactions, the catalyst supports require that a high surface area coating must be deposited upon its surface. The coating should provide adequate porosity 80 that the catalytic materials present can be more effective. A coating comprising these requirements is that of y-alumina. The alumina coating is applied usually by the immersion of the substrate in an alumina slurry. The support coated with the wet slurry is blown with air to remove excess slurry and subsequently dried and calcined (Stiles, 1983).
* To whom correspondence should be addressed.
To deposit the suitable amount of alumina in a single application, the solid content of the slurry has to be high. It is important to control the viscosity during the washcoating process, in order to achieve high-slurry alumina loadings, to aid the adhesion of the alumina to the substrate and to prevent the plugging of monolithic structure channels. Keith et al. (1967)prepared an alumina slurry by wet milling 970 g of alumina which had a composition of 86% y-alumina and 16% boehmite with the addition of 910 mL of deionized water and 20 mL of concentrated HC1. The mixture was milled for 18 h at about 80-112 rpm to obtain a thixotropic slip, which was used to impregnate an aalumina corrwated twe block. This slurrv has been found to be sufficiek in pioviding a uniform Eoating to the cyalumina block with dimensions 6 in. by 4 in. by 3 in.
0888-5885/92/2631-0364$03.00/00 1992 American Chemical Society
Ind. Eng. Chem. Res., Vol. 31, No. 1,1992 365 passage length and 7 corrugations/(linear in.) throughout the face of the 6-in.-length side. Alternatives to the previous preparation of alumina slurries have been used by Sowards and Stiles (1970) and Keith et al. (1971). A mixture of alumina and water was blended a t high speed in a Waring blender for 15 min to make an alumina slip, or pretreated alumina with H2PtCI,was used and HC1 was not added in the wet milling. A procedure for a slurry preparation with good adherence between the coating and the substrate has been proposed by Dwyer and Pesansky (1975). The slurry, which consisted of 45-70 wt % alumina trihydrate, 4-18 wt % polysiloxane resin, and 18-35 w t % toluene, was mixed by rolling for about 12 h. Then, the so-prepared slurry was used to coat a ceramic honeycombed substrate with 20 channels/in.2. Hoyer and Johnson (1977) proposed a coating process according to which a slurry with 25% alumina solids has been used to deposit a catalytic coating to a cordierite honeycomb structure with 300 square cells/in.2. Typical properties of a coating slurry used to impregnate a 400 cells/in.2 cordierite monolithic substrate by means of a specified process are given by Shimrock et al. (1985). They suggest that the solid loading of the slurry (i.e. alumina or any refractory solid) ranges between 35 and 52% and that the viscosity of the slurry is controlled in the range of 15-300 cP. With the rheological characteristics of the slurry to be adjusted in the range referred to above and using the coating process they have suggested, an impregnation of the support with a predetermined amount of alumina or catalyst could be achieved. An examination of the slurry preparation and substrate coating procedures referenced in the literature does not reveal the parameters influencing the wet-milling process for the preparation of slurries proposed for the monolithic structure impregnation. Recently, Tangsathitkulchai and Austin (1988) analyzed the influence of slurry density on the breakage parameters of quartz in a laboratory ball mill. They identified that one of the grinding characteristics associated with the change in slurry density was the slowing effect in fine grinding. The rheology of concentrated slurries of coal or quartz has also been investigated by Tangsathitkulchai and Austin (1989). The strong dependence of viscosity on the particle size distribution was confirmed from their experimental results. In the present work, the wet milling of alumina in a laboratory ball mill is investigated, in an effort to prepare stable colloidal dispersions. The influence of acid concentration on the wet-milling process and the kinetics of grinding are postulated. In addition, the effects of the solids loading, the particle size distribution, and pH upon the viscosity of alumina slurries are investigated. Finally the pore size distribution of monoliths impregnated with slurries prepared with solids of different particle size distributions were presented in this paper.
Experimental Section Raw Material. The raw material used in the present study was hydrated alumina provided by Aluminium of Greece, consisting mainly of gibbsite, according to X-ray diffraction analysis. The characteristic particle diameter L was equal to 77.6 nm, and the distribution factor n was 2.5 (Figure l),according to the Rosin-Rammler distribution. In order to feed the wet-milling process with a more finely divided raw material, a dry grinding of hydrated alumina was performed. A Bond ball mill was used,whose size and the grinding media loading characteristics are
ml
BOND MILL P R O D U C l c 25
0RAW
< I 1-1.51.5-22-3
MATERIAL
3-4
4-6
6-88-12
( 1 6 (24
DIFFERENTIAL PARTICLE
t32 (48 t64 t96 t 1 2 8 < 1 9 2
SIZE (pm)
Figure 1. Differential particle size distribution of raw material and material produced by dry milling. Table I. Size and Loadings of Bond and Pascal Ball Mills Bond Pascal ceramic ball mill ball mill volume/L 22.86 1.16 mill speed/rpm 120 grinding media (spheres) steel porcelain loading/ kg 19.352 0.487
Configuration of Grinding Media quantity diameter/mm Bond ball mill Pascal ceramic ball mill 41-38 5 31-30 115 26-22 36 121 21-20 19 32 13 54 10 15
presented in Table I. The mill operated for 15000 rounds, and the produced alumina (Al) had the differential particle size distribution illustrated in Figure 1,with a mean diameter of L = 14.4 pm and uniformity fador of n = 1.14. The producCwas then dried and calcined at 600 "C for 3 h to yield a mixture of y-alumina and boehmite with no trihydrates, according to X-ray diffraction analysis. Through mercury porosimetry measurements, the alumina produced (A2) was found to possess a specific surface area of 136 m2/g, a pore volume of 0.69 cm3/g, and a mean pore radious of 31 A In each of the experiments reported here, aluminas A1 and A2 were the raw materials. Wet-Milling Apparatus. A Pascal ceramic mill was used for the wet-milling batch process, whose size, grinding media weight distribution, and the speed of revolution are presented in Table I. The mill loading in our experiments was 150 g of alumina A1 or A2 with 150 g of acidified water. Samples of approximately 0.5 mL were taken every 10 h. The granulometric analysis of the samples was carried out using a SILAS 715 granulometer analyzer (1-200 pm), and the results represent the average of three repetitions. Viscosity Measurement Procedure. For viscosity measurements, a Brookfield viscometer was used with the LV spindle set. Because of the thixotropic nature of the slurries, the viscometer operated at the standard spindle speed of 12 rpm and the spindle LV3 was used in most of the measurements. Different spindle or spindle speed in viscosity determinations are specifically mentioned where applicable. Readings from the viscometer dial were taken after 2 min of operation. Impregnation of Monolithic Structures. The raw material used in the impregnation experiments was the product of the wet milling of alumina A2 with 3% HC1 for 40 h, which was dried for 24 h at 120 "C (alumina A3). Slurries were prepared by mixing alumina A3 with water in a high-speed blender for 3 min. This process was re-
366 Ind. Eng. Chem. Res., Vol. 31, No. 1, 1992 HYDRATED ALUMINA GAMMA ALUMINA ,
f
-0-
,
2 2 1
2 0 0
,
40
20 30 TIME (hours)
10
j
,
,
aw
=
o
10
1
0GAMMA ALUMINA
'
-
;5 -
DIF.PARTICLE
% 5-
4 1-16
'
-
w
--8.
4-
SIZE
C l
1.5-2
2-3
1 l
dWi/dt = -SiWi
+ j ZbijSjWj =O
for i = 0 to N - 1 (1)
where Si is the specific rate of breakage, Wi is the weight fraction of material size i, bij is the material fraction size i obtained by primary breakage of material size j , and N is the number of sizes. In our model the specific rate of breakage is supposed to be a function of fraction size Xj of the form
sj = al(xj)a
(2)
and the breakage function has the form (Austin et al., 1984) Bij = (1- exp(-(Xl-i/Xj)b))/(l
- exp(-l))
(3)
The term bij is combined with the breakage function by the equation b1 J. . = B1.J . - B l+lJ . . (4) The techniques for solving similar problems are examined by Reid (1965) and Austin et al. (1984). The solution of the differential equations involved in the proposed model, for all material sizes, was achieved by using the Adams-Moulton predictor corrector method (Lapidus, 1962), and the parameters a, al, and b are estimated from experimental data by a Marquardt-Levenberg algorithm (Beck and Arnold, 1977). The estimation of parameters gives value a = 1.457 [II = 4.19 X 10" b = 1.564
significance region (0.05 level) 1.31-1.75 (2.94-5.87) X loa 1.25-2.03
Figure 9. Plot of experimentalvs predicted values from the second model (1-6-pm range).
The experimental differential weight and the relevant model predictions are drawn in Figure 8 for the sake of comparison. It is evident from the latter figure that the model predicts adequately the results of the wet-grinding process. A simpler approach has been examined also. Focused on the rate of production of particles size of less than 1 pm, a model of comminution is considered in which the produced material coarser than size i depends on ita initial value only. Thus, if Pi,,denotes the material coarser than i at time t , the relationship describing the batch grinding is dPi,t/dt = -kP;,t
(5)
where k is generally assumed to be dependent on material size and c is a constant. In our case, where the model will be tested in a narrow band of material sizes, e.g. 1-4 pm, k is assumed to be independent of particle size and, on integration of eq 5 between 0 and t, yields p bt. (1-c) - p. (1-c) = -(I - c)kt (6) 1,O The values of Pi,,predicted by the latter model and the relevant experimental values, for 12 = 0.01 h-' and c = 0.41, are plotted in Figure 9. The values of constants k and c have been deduced by a nonlinear regression using the Marquardt method. This method gives a satisfactory fit within the experimental error of the data. From the previous analysis the conclusion which may be drawn is that both kinetic models proposed in this work describe satisfactorily the grinding process. The first model takes into account breakage of particles to any size, while the second deals with fine particle production only. Influence of pH on the Viscosity. Increasing acidity or basicity from the neutral point dramatically affects viscosity, as can be concluded from Figure 10. When the solids loading in the slurry was 24.8%, a maximum viscosity value appeared at pH = 8-10 and the alumina suspension showed a sharp viscosity increase at about pH = 6.5. The same behavior of alumina slurries was observed for different solids contents, but the sharp pronounced increase was observed near pH = 4 for the more dense suspension (Figure 10). The sensitivity of viscosity from pH is illustrated in Figure 11, where a slurry with a solids content of 36% is used and the pH has been changed from 2.3 to 4 (acidic side). An increase of viscosity by 30-60 times occurs for a pH change from 2.3 to 3.3. It is evident from these results that for a pH range of 3.5-3.8 slurries with high solids contents and minimum viscosities can be prepared.
368 Ind. Eng. Chem. Res., Vol. 31, No. 1, 1992
-
,-
Table 11. Mercury Penetration Porosimetry Rerults
+ 24.8% A1203
14
4
U
s
a 12
-g
10
2
6
+
33.7% A1203 40% A L 2 0 3 (LV2)
mean particle diameter in slurry/rm % solids gain (w/w) pore volume (cm3/g) surface area/(m2/g) % relative surface areaa
I-
a s
!z
a
( 0 4
0 $ 2
>
s1 52 53 1.24 2.62 10.54 26.8 29.5 24.3 0.256 0.240 0.226 28.80 29.44 25.60 20.28 14.03 9.94
64000-125 A macropores + intermediate pores. 0.25
0 2
6
4
8
10
12
0
PH
82
0.20
Figure 10. Influence of pH on the viscosity of alumina slurries for different solids contents.
A = -0.15
2000
2
Spindle speed
-+- "=30 r.p.m.
v
P
0
+ v-12 r.p.m.
50.10
B '0
0.05
0.00 2
2.2
2.4
2.6
2.0
3
3.2
3.4
3.6
3.8
4
PH
Figure 11. Viscosity of alumina slurries VB pH at acidic side for two different spindle speeds. The spindle LV2 wae used until pH = 2.8. Then spindle LV1 was used. Solids content = 36%. 1w.00
80.00 c
c
.-m
P
.-
60.00
0)
V
r
a"
40.00
a 20.00
OW
loo
Pore Rodius,r&)
to00
Figure 13. Pore size distribution of impregnated monolithic substrates.
With rheological characteristics of the slurry obtained by the adjustment of pH and solids content, as mentioned before, a weight gain of 1&22% (in dry basis) was achieved after two applications. To investigate the effect of the particles size distribution on the pore structure of the alumina coating, three slurries with different particle sizes were prepared. These slurries consisted of particles with mean diameters of 1.24, 2.62, respectively (denoted as S1, S2, and S3), and and 10.54 e, the differential particles size distribution is illustrated in Figure 12. The dispersions have solids loading and pH values as previously established, and also the impregnation and the drying processes were the same as before. From mercury penetration porosimetry results (Table 11) and the differential pore volume distribution diagram (Figure 131, a significant difference in the intermediate pore region (120-1000 A) can be observed. The microporea volume and pore volume distribution, which depended only on the nature of the alumina, and the macropore region, which depended on the ceramic substrate used, are the same for the three samples. Sample Sl, prepared by using the slurry with the lower mean particles diameter, indicates the greater value of pore volume equal to 0.26 mL/g. It is quite evident that if the controlling mechanism of a chemical process is the diffusion occurring within the intermediate pores, the catalytic activity is influenced by the preparation process of the coating slurry.
1
01
100
Pakicle size(&? Figure 12. Cumulative particle size distribution of slurries that have been used at monoliths impregnation.
Influence of Solids Loading on Viscosity. The influence of solids loading on the viscosity is demonstrated in Figure 12. For the pH range 3.5-3.8 the viscosity increases almost proportionally to the alumina A3 loading. The use of mixtures of alumina A2 and A3 leads to an interesting result. The viscosity is slightly dependent on the solids content when alumina A2 is added to a slurry which is containing 35% alumina A3. The suspensions prepared are stable up to 52% total solids loadings. In experiments of the mixturesof aluminas the mean particles diameter L ranged from 2.8 to 10.5 pm and particles with colloidal dimensions were at least 10.5 w t %. The Viscoeity measured was lower than 100 CPfor all of these cases. Impregnation of Monolithic Structures. According to the previous resulta, the fd viscosity of the dispersion will be dependent on the solids content chosen and the pH value. The pH of the slurries prepared for impregnation of monolithic substrates was adjusted at 3.7, and the solids loading was selected to be 42%.
Conclusions The impregnation of ceramic monolithic substrates with a certain alumina was examined and the following conclusions can be drawn. (1) Wet milling of y-alumina gives stable dispersions in water with 50% solids content when acid is added. The role of acid is to control the viscosity in the ball mill, and the wet grinding is independent from ita content in the range of 1-10 wt % dry alumina. Solids breakage does not proceed at either low or high acid concentrations. The latter observation can be attributed to the high slurry viscosity due to the prevailing pH value.
Ind. Eng. Chem. Res. 1992,31,369-373
(2) A power law model for the specific rate of breakage was found to describe the milling process. The power law constant was estimated to be equal to 1.457 when the constant of the breakage function was 1.564. The particles of size 1.0-1.5 pm appeared to have the more pronounced increase, and for the fine particles production a simple model was proposed. (3) Dispersions with high solids loading can be prepared if pH is controlled in the range 3.5-3.8. The dispersions are stable when the approximately 10.0% solids particles have diameters less than 1pm. (4) Impregnation of ceramic monolithic structures with different particle size distributions affects the alumina coatings pore structure in the intermediate pore radius region.
Acknowledgment C.P. acknowledges the Hellenic Cement Research Center for the X-ray diffraction and granulometric analyses of the samples. Thanks are also due to The Aluminium of Greece for the supply of hydrated alumina and Mr. D. Tsamatsoulis who helped in the kinetics models evaluation. Nomenclature a = constant, eq 2, dimensionless al = constant, eq 2, h-I b = constant in the breakage function, eq 3, dimensionless bij = material fraction of size i obtained by primary breakage of material size j , dimensionless Bij = breakage function, eq 3 c = constant of power law, eq 5 i , J’ = denote material size k = specific rate constant, eq 5, h-’ L = characteristic particle diameter, pm
369
n = uniformity or distribution factor, dimensionless
N = number of sizes
Pi,t = weight percent of material coarser than size i at time t
Si, Si= specific rate of breakage, h-’ t = grinding time, h X i , X, = weight of material size i or j , respectively u = spindle velocity, rpm Registry No. A1203, 1344-28-1; HC1, 7647-01-0.
Literature Cited Austin, L.; Kimpel, R.; Luckie, T. Process Engineering of Size Reduction: Ball Milling; Society of Mining Engineers: New York, 1984, pp 61-74. Beck, J.; Arnold, K. Parameter Estimation in Engineering and Science; Wiley: New York, 1977; pp 167-184. Dwyer, T.; Pesansky, D. U.S. Patent 3,873,350,1975. Hoyer, A.; Johnson, L. U S . Patent 4,039,482, 1977. Keith, C.; Kenah, P.; Bair, D. US.Patent 3,331,787, 1967. Keith, C.; Kenah, P.; Bair, D. U.S. Patent 3,565,830, 1971. Lapidus, L. Digital Computation for Chemical Engineers; McGraw-Hill: New York, 1962; p 98. Reid, K. A Solution to the Batch Grinding Equation. Chem. Eng. Sci. 1965,20,963-963. Shimrock, T.; Taylor, R. D.; Collins, J. Eur. Patent 0157651, 1985. Sowards, M. D.; Stilea, B. A. U S . Patent 3,518,206,1970. Stiles, B. A. Catalyst Manufacture, Laboratory and Commercial Preparations, Chemical Industries; Marcel Dekker, Inc: New York, 1983; Vol. 14, pp 86-99. Tangsathitkulchai,C.; Austin, L. G. Rhelogy of Concatrated Slurries of Particles of Natural Size Distribution Produced by Grinding. Powder Technol. 1988,56, 293-299. Tangsathitkulchai,C.; Austin, L. G. Slurry Density Effects on Ball Milling in a Laboratory Ball Mill. Powder Technol. 1989, 59, 285-293.
Received for review December 17,1990 Revised manuscript receiued July 17, 1991 Accepted August 13, 1991
Analysis of Zeolite Crystallizations Using the “CrystallizationCurve” C. J. J. den Ouden* KoninklijkelShell-Laboratorium,Amsterdam (Shell Research B. V.),Postbus 3003, 1003 A A Amsterdam, The Netherlands
R. W . Thompson Department of Chemical Engineering, Worcester Polytechnic Institute, Worcester, Massachusetts 01609
Zeolite crystallization experiments are quite commonly analyzed by means of a “crystallization curve“. These curves, collected for batch zeolite crystallizer operations, represent the evolution of zeolite mass in the crystallizer in the course of an experiment. The data are frequently presented as the zeolite mass, the zeolite yield, or the percentage of zeolite in the solid phase as a function of the crystallization time. One type of analysis of zeolite crystallizations involves the measurement of the induction time and the slope of the crystallization curve to quantify the nucleation and crystallization rates, respectively. It is shown here that these analyses of the crystallization curve, though commonly performed, are likely to give misleading or innacurate results, principally due to the insensitivity of measurements of the mass of zeolite by conventional methods. Although the analysis allows a reasonable comparison of similar systems, it cannot be used to reveal details regarding the crystallization kinetics or to compute activation energies from such an analysis.
Introduction Molecular sieve zeolites are crystalline aluminosilicates with regular pore structures suitable for use in several industrial processes. They are commonly formed by hydrothermal synthesis in caustic media in the presence of a precursor amorphous gel phase. Thus, during a typical zeolite crystallization, at least three phases are present:
amorphous gel, caustic solution, and crystalline product. Under normal circumstances, the aluminosilicates crystallize from the solution phase and are replenished there by dissolution of the gel phase. As a result of this process, if a sample of the ‘solid phase” is collected at some intermittent stage in the process, ita analysis by powder X-ray diffraction techniques
0888-5885/92/2631-0369$03.00/00 1992 American Chemical Society
