Verified Syntheses of Zeolitic Materials

2nd Revised Edition

The pH-value and its importance for the crystallization of zeolites

Hans Lechert
Institute of Physical Chemistry of the University of Hamburg
20146 Hamburg, Germany

1. Introduction

The alkalinity in a synthesis batch is one of the most important parameters for the control of the crystallization of zeolites. It determines their composition and is to a great extent responsible for the type of the crystallizing product. [1-13]

Generally, the crystallization proceeds via the solution phase so that the species of silicate, aluminate, and aluminosilicate in the solution are important for the crystallization mechanism. The pH-value of the solution is determined by the total alkali content and complicated buffering equilibria of the mentioned species. [14, 15] Zeolites are usually synthesized in the presence of an amorphous gel phase. The solubiity of this gel phase also depends on the alkalinity. It assures the supersaturation for nucleation and growth processes.

The composition of synthesis batches can be described by the formula: MAlO2n[MmH4-mSiO4] pH2O (1)The ╬excess alkalinity" (m) in this formula is the difference between total alkalinity (MOH) and the alkali aluminate (MAlO2) per mol of SiO2. m = (MOH - MAlO2)/SiO2 (2)In systematic studies of the influence of the alkalinity on the product Si/Al-ratio, m is generally used as a critical parameter. [1-4]

A detailed discussion of the formation of zeolites has been given recently by Jansen [5] and by Feijen, et al. [6] In [7-12] thorough studies of the Si/Al ratio and the rate constant of linear growth in their dependence on m and n have been carried out for a series of zeolites.

The pH-value has been discussed in only a few papers in connection with the parameters of the crystallization. Generally, the different zeolite types crystallize within rather narrow ranges of pH. For faujasites, values between 12.3 and about 13.8 are observed. Robson has discussed procedures of a synthesis of NaY with Si/Al = 5 at pH = 11. [12] Between 11.3 and 12.7 mordenite usually crystallizes; at lower values ZSM-5 is obtained. [1,12] Donahoe and Liou have found a linear dependence of the Si/Al ratio of phillipsite and merlinoite in the pH range 13.3-13.7 in crystallization experiments from clear solutions of systems with an extremely high silica content. [13] In a series of papers, single values of the pH or pH changes before and after crystallization are reported. These differences are usually 1-2 units on the pH-scale.

Much work has been done in the analysis of the species in silicate and silico-aluminate solutions. Important results have been obtained from NMR. [14-16] A summary of this and other work can be found in the book of Engelhardt and Michel [16]. A thermodynamic analysis of these species has been undertaken by Guth, et al. [17, 18]

Generally, it was found that in the pH range of the crystallization of zeolites, the silicate is most probably present as [SiO2(OH)2]2- or [SiO(OH)3]-. Only comparatively low concentrations of higher condensed species are present at pH values above about 12.0. At lower pH the concentration of dimers and four-membered ring species increases. This range can be roughly identified with the crystallization region of the more siliceous zeolites with five-membered rings in their structure.

The aluminate is generally present in very low concentrations and is often described to be present as [Al(OH)3OSi(OH)3]-. [16,18] A general survey of the pH-dependency of the hydrolysisequilibria of cations has been given by Livage. [19] These data are important for syntheses with other anions present beside aluminate and silicate or aluminate and phosphate. [20]

For a more detailed investigation of the pH in zeolite synthesis, we have done thorough experiments on the connection of the pH with m and n and the Si/Al ratio of faujasites within a broad range of compositions.

2. Remarks on the Measurement of pH

Direct control of the pH during the crystallization would be desirable especially for industrial processes. However, this is difficult because zeolites crystallize at rather high pH-values. For a measurement of these values the Pt|H2-electrode or the glass-electrode can be used.

With the Pt|H2-electrode, in principle, very exact results can be obtained as has been shown by Lagerström [21] and Ingri [22] in very careful studies of the pH-dependency of the hydrolysis equilibria in silicate solutions. However, the Pt|H2-electrode is not applicable for routine measurements for automatic control of a crystallizing batch in an industrial reactor.

For measurements with the glass electrode, it must be taken into account that this electrode responds to the concentration of the alkali ions as well as the H3O+ present in the solution. This so-called ╬alkaline error" becomes important at pH > 11.0, depending on the sensitivity of the electrode material. Data for different electrode materials are demonstrated in [23]. The alkaline error increases with temperature. It is advisable to follow the suggestions of the manufacturer of the electrode carefully. At high pH the properties of the glass membrane of the electrode may change with time. An obvious alternative is to take samples from the crystallizing batch and to carry out the measurement of the pH at ambient temperature. This procedure is thoroughly described in the literature.

3. Preparation of the Reaction Mixtures and pH-Measurements

Generally, the batches for zeolite crystallization were prepared by mixing a silicate source and aluminate source and adjusting the excess alkalinity m with a suitable quantity of alkali hydroxide. For the present study the following sources were used:

Alumina source: 2.5 mole Al(OH)3 (Merck, reinst) and 5 mole NaOH (Merck, reinst) in 1000g solution.

Silica source: Water glass [Merck, sp. gr. = 1.37) or Silicic acid (Fällungskiesel säure, Merck) and NaOH to give 273.5 g SiO2 in 1000 g solution.

The solutions were mixed at ambient temperature adding the silicate to the aluminate and the additional NaOH. Most of the experiments (p = 400) were carried out with a concentration of 50 g (AlO2- + n SiO2) / 1000 g H2O.

Regarding the general batch composition, MAlO2n(MmH4-mSiO4) pH2O, the concentrations were adjusted to values of n = 2.0 - 14.0, m = 0.3 - 2.5 and p = 400. Some of the batches had lower water contents of p = 195 and p = 260.

After 2 hours homogenization an appropriate amount of nucleation gel was added to assure the crystallization of Y-zeolite, again stirred for half an hour and then the pH measured. The nucleation gel had a composition of NaAlO2 7.5 (Na2.0H2.0SiO4) :155 H2O. [11] The batch was then heated to the crystallization temperature of 88║C.

After crystallization the samples were recovered and characterized as described in [7-11]. The Si/Al-ratios of the products were determined by the EDAX method in a Phillips SEM 515 and the EXAX-Analyzer PV 9900. The reaction mixtures had pH values between 12.4 and 14.1. From all mixtures faujasite crystallized. The pH measurements were done with WIWS 12 and a pH electrode belonging to this apparatus.

Reproducible results could be obtained according to the following procedure:
- Calibrate at pH = 7.0 using a commercial buffer solution (Riedel de Haen).
- Calibrate at pH = 13.0 using a commercial buffer solution (Riedel de Haen) and adjust the slope ofthe instrument.
- Read the pH of the sample. Usually a time of about 5 minutes was sufficient to obtain a constant reading.

Between the different steps the electrode was cleaned with distilled water.

4. Discussion

In Fig. 1 the correlation of m and the measured pH values is demonstrated for the different Si/Al rations n in the batch. The relation of both quantities can be fitted by a logarithmic relation which is equivalent to a titration curve of a weak acid with a strong base at the alkaline end. For n <5, the curves for the different n are fairly close together. For higher n, deviations are observed which are not demonstrated in Fig. 1. These n are often applied for crystallization of NaY

Click here for Figures 1 and 2

with high Si/Al ratios and also for crystallization of the more siliceous zeolites as, e.g., mordenite or ZSM-5. m is a more sensitive measure for the alkalinity than the pH value, apart from the advantage that it is given directly by the batch composition.

Fig. 2 shows the dependency of the Si/Al ratio of the products on the measured pH for different n in the batches. The crosses demonstrate results of crystallization experiment leading to phillipsite and merlinoite carried out by Donahue and Liou. [12]

The product Si/Al ratio decreases almost linearly with the pH if n is held constant. For increasing n the slope of these straight lines increases. The values for n = 6.0, 7.5 and 8.0 lie near the curve for n = 5.0. Formally, the curves for different n meet at Si/Al = 1.5 and about pH = 13.5.

Extrapolating ihe curve of n = 5 and 6.0, 7.5 and 8.0 in Fig. 2 to pH = 11.0 as it was suggested for the faujasite synthesis, a Si/Al ratio of about 4.0 can be obtained. [12]

Summarizing, Fig. 2 shows that for fixed n the Si/Al ratio of faujasites is a unique function of the pH value in the solution phase. However, for a prediction of the Si/Al ratio, the pH is not very suitable because exact pH values in the batches can be expected only after a time-consuming aging procedure.

Therefore, we have tried to find a direct relation between the parameters n and m of the batch composition and the final Si/Al ratio. [11] From extensive experiments, including studies of the concentrations of OH- ions and the silicate in the solution phase [7-10], a simple relation could be derived which holds with great accuracy Si/Al =1+b|SiO2|sol/|OH-|sol (3)using b = 2. Eq. 3 can be explained by a simple model of the formation of zeolites. [24]

Like the pH values, the concentrations are not very suitable for practical use in the prediction of the Si/Al ratios from the initial composition of the gel. With some obvious assumptions, Eq. 3 has been changed to an empirical relation containing only n and m as parameter with the constant b =2. Si/Al = (b+m)n/(b+mn) (4)In Fig. 3 a large number Si/Al product ratios from batches for different m with 1.4<n <14.0 are compared with values which have been calculated using Eq. 4.

Click here for Figure 3

It can be seen that the Si/Al ratio of the crystallizing product can be predicted by Eq. 4 over the whole range of n much more reliably than from the pH values of the solution phase.

From extensive studies of literature data, especially from [1-4] and the first edition of ╬Verified Syntheses," it could be seen that Eq. 4 can be used successfully for all zeolites with four-and six-membered rings in their structure. [24] As examples, ZSM-3, LTL, Rho and Offretite shall be mentioned. Data for high silica zeolites like mordenite, ZSM-5 and Beta can be described by Eq. 4 using larger values of b. [25]

Systematic arguments for a choice of b for a special zeolite are, however, missing until now. Therefore, Eq. 4 must be regarded as an empirical relationship at the present stage of discussion.

The general role of the OH- ion as a mineralizer can be partially replaced by fluoride leading to a variety of new syntheses at lower pH values. These syntheses are usually called ╬low alkaline syntheses" or ╬fluoride route." Most probably the F- increases the solubility of the aluminate as could be shown, for example, in kinetic experiments of the crystallization of Y zeolites. [26]

5. Conclusions

The composition and the ranges of stability of the zeolitic products depend on the pH value, the alkalinity and the Si/Al ratio present in the batch. For practical use in the planning of syntheses, the pH value is rather complicated to handle. The empirical Eq. 4 proved to be suitable for a prediction of the product Si/Al ratio from the batch composition with the restrictions discussed for high silica zeolites. Eq. 4 includes the influence of the alkalinity m as well as the Si/Al ration n of the batch.

Predictions of the kind of the crystallizing zeolite and also of the effect of templates have not been possible thus far and remain a matter of experience.

6. References

[1] R. M. Barrer, Hydrothermal Chemistry of Zeolites, Acad. Press, London, 1981
[2] D. W. Breck, Zeolite Molecular Sieves, John Wiley & Sons, New York, London, Sydney, Toronto,1974
[3] Zeolite Synthesis, ACS Symposium Series 398, M. L Occelli, H. Robson (eds.), Amer. Chem. Soc., Washington DC, 1989
[4] Molecular Sieves, Vol. 1, M. L Occelli, H. Robson (eds.), Van Nostrand Reinhold, New York, 1992
[5] J. C. Jansen, in Introduction to Zeolite Science and Practice, H. van Bekkum, E. M. Flanigen, J. C. Jansen (eds.), Elsevier, Amsterdam, 1989, p. 77
[6] E. J. P. Feijen, J. A. Martens, P. A. Jacobs, in Stud. Surf. Sd. Catal. 84A, J. Weitkamp, H. G. Karge, H. Pfeifer, W. Holderich (eds.), Elsevier, Amsterdam, 1994, p. 3
[7] H. Kacirek, H. Lechert, J. Phys. Chem. 79 (1975) 1589
[8] H. Kacirek, H. Lechert, J. Phys. Chem. 79 (1975) 1291
[9] H. Lechert, H. Kacirek, Zeolites 11(1991) 720
[10] H. Lechert, H. Kacirek, H. Weyda, in Molecular Sieves, M. L Occelli, H. Robson, (eds.), Van -Nostrand Reinhold, New York, 1992, p. 494
[11] H. Lechert, P. Staelin, Ch. Kuntz, Zeolites 16 (1996)149
[12] H. E Robson, in Zeolite Synthesis, ACS Symposium Series 398 M. L Occelli, H. Robson (eds.), Amer. Chem. Soc., Washington DC, 1989, p. 436
[13] R J. Donahoe, J. G. Liou, Geochim. Cosmochim. Acta, 49 (1985) 2349
[14] G. Harvey, L S. Dent Glasser, in Zeolite Synthesis, ACS Symposium Series 398, M. L Occelli, H. Robson (eds.), Amer. Chem. Soc., Washington DC, 1989, p. 49
[15] A. T. Bell in Zeolite Synthesis, ACS Symposium Series 398 M. L Occelli, H. Robson (eds.), Amer. Chem. Soc., Washington DC, 1989, p. 66
[16] G. Engelhardt, D. Michel, High Resolution Solid State NMR of Zeolites, Wiley Sd. Pubi., Chichester (1987)
[17] J. L Guth, P. Caullet, A. Seive, J. Patarin, F. Delprato, in Guidelines for Mastering the Properties of Molecular Sieves, D. Barthomeuf, et al (eds.), Plenum Press, New York, 1990, p. 69
[18] P. Caullet, J. L Guth, Zeolite Synthesis, ACS Symposium Series 398, M. L Occelli, H. Robson (eds.), Amer. Chem. Soc., Washington DC, 1989, p. 83
[19] J. Livage, in Stud. Surf. Sci. Catal. 85, J. C. Jansen, et al. (eds.), Elsevier, Amsterdam, 1994, p. 1
[20] J. L Guth, H. Kessler, R Wey, in New Development in Zeolite Science and Technology, Y.Murakami, A. Lima, J. W. Ward (eds.), Kodansha, Elsevier, Amsterdam, 1986, p. 121
[21] G.Lagerström.Acta Chim. Scand. 13 (1959), 722
[22] N. Ingri, Acta Chim. Scand. 13 (1959) 775
[23] D. A. Skoog, J. J. Leary, Principles of Instrumental Analysis, Saunders College Publishing, Harcourt Brace College Publishers (1992) 498
[24] H. E. Robson, Verified Syntheses of Zeolitic Materials, Micropor. Mesopor. Materials 22, H. Robson (ed.), Elsevier, Amsterdam, 1998, p. 495
[25] H. Lechert, Micropor. Mesopor. Mat., submitted for publication
[26] T. Lindner, H. Lechert, Zeolites 14 (1994) 582