Abstract
A series of experiments of zeolite X synthesis with variations of compositions and temperatures had been performed. The crystallization time would be affected by both the values of Na2O/Al2O3 and SiO2/Al2O3. The crystallization time would de-crease with Na2O/Al2O3, but increase with SiO2/Al2O3. These might be caused by the solubility of the gel in the solution. Based on the assumption that the hypotheti-cal structure unit and nuclei were Na11[(AlO2)11(SiO2)13]·33H2O and Na88[(AlO2)88(SiO2)104]·264H2O, the reaction order was determined to be nearly 0.5. The activation energy of the crystal growth was determined by the Arrehenius plot to be 18.00 Kcal/mole. In addition, the mechanism of the crystal growth was suggested that the solution and gel contained the hypothetical structure units. The hypothetical structure unit would dissolve from the gel into the solution and migrate to the surfaces of the nuclei or crystals. Then the crystal surface reaction involved three elementary steps. The structure unit would rearrange and compose into two equivalent species X, and the two X would react with the nuclei or crystals sequentially to form the new surfaces of the crystal. The repetition of these elementary steps would contribute to the macroscopic growth of the crystal. The modified model of zeolite X synthesis in the batch system consisting of homogeneous nucleation and crystal growth mecha-nisms was proposed. The interfacial energy and the size of the stable nucleus were calculated from the parameter A to be nearly 37 erg/cm2 and 40 Angstroms, respec-tively. The model predictions of the crystallization curves and crystal size distribu-tions were in qualitative agreement with the experimental results. Only at the larger size zone, the model prediction of crystal size distribution exhibited a dramatically steeper rise than the experimental data did. This difference might be caused by two reasons. One was that the assumption of the isothermal condition could not be achieved practically. However, the temperature effect on the crystal size distribution of the product, which needed longer time to synthesis was not significant due to the small ratio of the time required for temperature change to the crystallization time. The other reason was that all species in the reaction mixtures were postulated to be the hypothetical structure units. This assumption set all species as the active species and would overestimate the concentration of the active species at the beginning of the synthesis. Therefore, the nucleation rate would increase gradually instead of rapidly. The attempt to simulate the CSTR performance of zeolite X synthesis was also made. A modified model for prediction of the CSTR performance of the zeolite synthesis, consisting of homogeneous nucleation and crystal growth mechanisms, was proposed. The CSTR performance of zeolite A synthesis [8], was modeled and the correlation between the CSTR and batch performance of zeolite A synthesis was studied. Only the precipitation and reverse rate would be significantly influenced by the flow pattern. Finally, the CSTR performances of zeolite X synthesis were simu-lated with the parameters of the model of the batch system and the postulated precipi-tation rate constant. The predicted evolution of the conversion with time would ex-hibit the S-shaped type. The conversions would approach approximately to 90%. The temperature effect on the predicted crystal size distribution of 50% conversion operation was not significant due to the opposed effects of the temperature on the crystal growth rate and the retention time of 50% conversion. It should be noted that no experimental data or simulated results present in the literatures could be employed to verify the prediction of the CSTR performance of zeolite X synthesis. Thus the results proposed herein would be helpful to the design of the CSTR for zeolite X synthesis.