J . Chem. SOC., Faraday Trans. I , 1982, 78, 1741-1753 Exchange of Ammonium and Sodium Ions in Synthetic Faujasites BY PHILIP FLETCHER AND RODNEY P. TOWNSEND* Department of Chemistry, The City University, Northampton Square, London EClV OHB Received 19th May, 1981 Ion-exchange isotherms for the Na $ NH, equilibrium in zeolites X and Y were constructed at 298 K and a total solution normality of 0.1 g-equiv . dmP3. Values for the standard free energy of exchange were calculated and shown to be consistent with the affinity sequence predicted using dielectric theory. Comparisons are made between previously reported studies on the NaGNH, exchange in X and Y and this current work, and the differences in experimental data are discussed. Statistical thermodynamic considerations provide a greater understanding of the Na s NH, exchange equilibrium in X and Y, and collectively the evidence suggests that high framework charge and ion-sieving factors in X are the dominant properties determining exchange selectivity and maximum exchange levels in this zeolite.In Y, the ion-sieve effect and repulsion between the entering ammonium ions are the dominant factors. Studies of the exchange of ammonium ions into the sodium forms of the synthetic zeolites X and Y have been undertaken by various workers in the past, ever since Barrer et a!.' reported a 92 % replacement of sodium by ammonium ion in X using 2 mol dm-3 ammonium chloride solution. Sherry2 first published an isotherm for the Na-NH,/Y equilibrium in 1966, and Theng et al. obtained isotherms for this exchange in both X and Y two years later.3 In his review of ion exchange in zeolites Sherry4 included a Na-NH,/X isotherm which was not found in his earlier paper,2 and since then, as part of a wider study of Szilard-Chalmers recoil studies in X and Y, Lai and Rees5 published a further isotherm for the NH, Na equilibrium in Y.Finally, Herman and Bulko include some data on the effect of temperature on maximal levels of exchange of ammonium ion in Y in a recent study on copper zeolites.6 Apart from the work of Theng et al., none of the above studies was intended to be concerned primarily with the sodium-ammonium exchange equilibrium in X and Y. Most of the studies were, however, carried out under similar conditions, and it is therefore surprising that the results are not in good agreement, especially in the case of X.As part of an ongoing study of transition metal exchange in ammonium and sodium zeolite^,^ l 1 it became necessary to re-examine the sodium-ammonium equilibrium in the faujasite-type zeolites, and thence to compare and contrast these data with previously reported information. The results of this examination are reported in this paper. EXPERIMENTAL MATERIALS Synthetic sodium zeolites were supplied by Union Carbide. Chemicals used, whether for analysis or exchange purposes, were of AnalaR grade. 17411742 ION EXCHANGE IN SYNTHETIC FAUJASITES ANALYSES (i) Zeolite phase: samples were analysed for Na,O, SiO,, Al,03, Fe,O,, (NH,),O and H,O content by methods outlined previously in the literature.’.(ii) Solution phase: sodium was determined in solution using either atomic absorption spectroscopy or flame photometry. Ammonia was determined by a modified Kjeldahl method described previo~sly.~. * PREPARATION OF EXCHANGED FORMS OF ZEOLITES Initially all zeolite samples were treated with solutions containing 0.5 mol dm-3 sodium chloride in order to ensure that starting materials were in the homoionic sodium forms. Extensive washing of these samples thereafter was avoided in order to minimise hydrolysis of the zeolites, especially X. For the isotherms (which were constructed at a total solution normality of 0.1 g-equiv . dmP3) maximum levels of exchange of ammonium for sodium in X and Y were obtained by exhaust- ively exchanging 0.2 g samples of the sodium zeolite with 50 cm3 aliquots of 0.1 mol dm-3 ammonium nitrate solution at 298 K.The resulting maximum exchange levels agreed with those indicated by the general isotherm shapes (see Results). Other experiments on the maximum exchange level of ammonium in X and Y were also conducted usingsolutions ofammonium chloride ofdifferent concentrations at room temperature or with 0.5 mol dm-3 ammonium chloride solution over a range of temperatures. In agreement with Herman and Bulko,6 the maximum exchange level in Y was found to be sensitive to temperature. The temperature sensitivity of X was much less marked. These results are recorded in this paper but also discussed el~ewhere.~. l2 EQUILIBRIUM STUDIES All equilibria were constructed at a constant total normality of 0.1 g-equiv .dm-3 using ammonium and sodium nitrate solutions containing different ratios of the two cations, and at a temperature of 298 K. Forward isotherm points were constructed by conventional methods described previ~usly.~ Reverse isotherm points were constructed by a modified method recently outlined9-l1 in order to avoid problems arising from ion-site redistributions on drying the zeolite. THERMODYNAMIC TREATMENT OF DATA Values of the standard free energy of exchange were calculated by the usual methods, described in detail re~ent1y.I~ Important aspects, summarised here, are the exchange equation and thermodynamic equilibrium constant K,, uiz. NH:(s)+Na+(c)sNH,+(c)+Na+(s) (1) and K a = ( Q N H ~ ( ~ ) ~ ~ a ( s ) / a ~ ~ ~ ( s ) aNa(c)) (2) where a is the activity, and (c) and (s) refer to crystal and solution phases, respectively.Values of K, were determined using the procedure of Gaines and Thomas’, from (3) where (4) and ion r is (NH,): and (Na): are the n~rmalised~~’ l5 equivalent fractions of ammonium and sodium in the zeolites. mNa and mNHl are the concentrations (mol dm-3) of the ions in solution. the ratio of corresponding single-ion activity coefficients in solution, which may be evaluated using Glueckauf’s method.16 In practice, for the Na e NH, system, the ratio is close to unity, and it is also (for uni-univalent systems) independent of crystal phase composition,1° so this correction is near negligible.P. FLETCHER AND R. P. TOWNSEND 1743 Finally, application of the Gibbs-Duhem equation enables the calculation of the crystal phase activity coefficients fNH4 and fNa using the relations14 In fNH4 = - In KF + (NH4); In KF + { In KF d(NH,): ( 5 ) ( N H ~ ) ? RESULTS Analytical data for samples of X and Y before and after exhaustive exchange of the sodium forms with 0.5 mol dm-3 ammonium chloride solution at 298 K are shown in table 1.It is apparent that 100% exchange of sodium by ammonium ion was not TABLE 1 .-CHEMICAL ANALYSES OF ZEOLITES sample unit-cell composition Na-X Na84.6(A102)84. 7(si02)107.3 .25 .7H20 Na-Y Na6~.3(A102)61.4(Si02)130.6242 * 5H20 NH,/Na-X (NH4)61.7Na22.9(A102)84.8(si02)107.2 ' 236H20 NH,/Na-Y (NH4)43Nai8.~(A102)6i.4(si02)*~~.6 * 219.4H2O sample oxide formula Na-X Na-Y NH4/Na-X NH,/Na-Y Na20. A1203 - 2.53SiO2. 5.94H20 Na20. A120, * 4.25SiO2 0.002Fe203 * 7.89H20 ((NH,)20)o,73(Na20)o~27 * Al,O,. 2.53Si0, * 5.57H20 ((NH,)20)o~7,(Na20)o~30 * A12034 25Si0, * 0.002Fe203 - 7. 14H20 achieved under these conditions, the maximum levels of exchange being 73 and 70%, respectively.The effects of both temperature and concentration of ammonium salt on the maximal level of exchange in X and Y are shown in table 2. Ion-exchange isotherms for the Na c NH, equilibria in X and Y are shown in fig. 1. Both systems were reversible. Plots of In K , against (NH,), are shown in fig. 2 and 3, and in fig. 4 are the corresponding crystal-phase activity coefficients, which were calculated using eqn ( 5 ) and (6). Values for the standard free energies of exchange and the thermodynamic equilibrium constant Ka are given in table 3.DISCUSSION GENERAL COMMENTS The exchange isotherm for the Na + NH, exchange in Y obtained in this work may be compared with others in the 3 7 The isotherm shape in fig. 1 conforms to type d of Breck's classification,17 and agrees closely with those found by Sherry2 and Theng et aZ.3 The AG* values are similar (table 3); indeed, the small differences observed are probably the result of the different silica to alumina ratios found in the1744 ION EXCHANGE I N SYNTHETIC FAUJASITES TABLE 2.-EFTECT OF TEMPERATURE AND SALT SOLUTION CONCENTRATION ON DEGREE OF AMMONIUM EXCHANGE percentage sodium removeda /K /mol dmP3 Na-X Na-Y temperature “H4ClI 298 298 313 333 353 298 298 298 298 298 298 298 0.1 0.5 0.5 0.5 0.5 1 .o 1.5 2.0 3.0 4.0 5.0 saturated solution 70.0 72.7 73.4 73.4 73.3 73.1 93.0 93.1 92.9 93.2 93.0 93.0 70.1 70.1 76.3 85.4 93.1 70.1 70.3 70.3 69.9 69.8 70.0 70.1 a For all experiments 0.2 g of zeolite were exchanged five times for 24 h with fresh 50 cm3 aliquots of the appropriate ammonium chloride solution.(NH,), FIG. 1 .-Ion-exchange isotherms (not normalised) for the Na $ NH, exchange in (a) X (b) Y: 0, forward points; x , reverse points; 0, direct analysis of exhaustively exchanged samples.P. FLETCHER AND R. P. TOWNSEND FIG. 2.-Normalised Kielland plot of N a e N H , exchange in X. 1745 I 0.0 0.2 0.4 0.6 0.8 1.0 (NH.9): FIG. 3.-Normalised Kielland plot for Na e NH, exchange in Y. Y samples used by the different workers. (This point is discussed further below.) The observed maximum levels of exchange are also similar, uiz.68,2 703 and 70% (this work). In contrast, the isotherm given by Lai and Rees5 is different, appearing sigmoidal in shape (i.e. type b in clas~ification),~~ and the maximum exchange level appears higher. However, their Y zeolite5 had a substantially higher aluminium content (68 aluminium atoms per unit cell) than those used here or in the other 57 FAR I1746 ION EXCHANGE IN SYNTHETIC FAUJASITES 1 C C f 0 0 0 0.2 0.4 0.6 0.8 (NH4),N FIG. 4.-Plots of the phenomenenological activity coefficients f for the ions NH: and Na+ in X and Y. Data are plotted in terms of the normalised equivalent fraction of ammonium ion in the crystal phase (NH,)?. TABLE 3.-cOMPARISON OF THERMODYNAMIC DATA FOR THE NH,,Na EXCHANGE IN DIFFERENT ZEOLITES SiO, : A1,0, zeolite ratio AGe/kJ rno1-I ref.temp./K ref. X X Y Y Y mordenite chabazite clinoptilolite clinoptilolite 2.52 2.53 5.33 4.98 4.25 10.53 5.13 10.00 10.00 - 2.801 - 0.534 - 2.759 - 2.592 - 2.005 - 3.760 -4.138 - 5.392 - 5.727 293 298 298 293 298 298 298 303 333P. FLETCHER A N D R. P. TOWNSEND 1747 studies,2? which may account for the different isotherm shape (see comments below on X). Regarding the maximum exchange level, Lai and Rees extrapolated their isotherm from < 80 to loo%, and in general used different conditions for their experiments to the other studies* since, as the authors themselves state, ‘thermo- dynamic analysis. . .was not the prime purpose of the exchange measurements, no attempt was made to cover the complete range of A , with the thoroughness such an analysis would warrant’.5 Finally, further data on the maximum exchange level for ammonium in Y are found in a recent paper by Herman and Bulko.6 Using a Y sample containing 56 aluminium atoms per unit cell, they found a maximum level of exchange of 75 % after three equilibrations at 298 K, and an exchange level between 68 and 7 1 % after one equilibration.These data are in good agreement with this work and other In contrast to Y, agreement over the sodium-ammonium exchange in X between different workers is very poor. There are less data on X in the literature; the only studies with which this work can be compared are those of Sherry4 and Theng et aL3 Sherry’s isotherm4 is type b, in common with fig. 1. However, in this present study a maximum exchange level of 70% was obtained at a solution concentration of 0.1 mol dme3, yet Sherry extrapolates his isotherm from ca.85 to 100%. In the absence of comment in the paper4 it is not possible to ascertain if any markedly different experimental conditions were used which may explain this discrepancy. Sherry does not publish a value for the free energy of exchange of ammonium into sodium X. Theng et aL3 calculated AG* for this exchange and obtained a value very much higher than that found here (see table 3). In addition, their isotherm is type d, and differs very markedly from both Sherry’s4 and this work, with a maximum exchange level of only 63%. Although Theng et aL3 determined their isotherm at a total solution molarity of 0.05 mol dmP3, whereas the solution concentration in both Sherry’s study4 and this present work was double this, this difference cannot be the explanation for the observed3 different isotherm shape, as Barrer and Klinowski have shown that for uni-univalent exchanges the isotherm shape is independent of the external solution concentration.1 9 9 2o CONSIDERATIONS OF ION SIZES A N D DISTRIBUTIONS The potassium ion (Pauling radius 0.133 nm) can completely replace sodium in both X4 and Y.2 In contrast,2v4 neither rubidium nor caesium (Pauling radii 0.148 and 0.169 nm, respectively) exchange to 100% in either zeolite. The effective diameter of the six-oxygen windows (which lead into the sodalite units) is quoted3 as being between 0.266 and 0.288 nm. It is apparent on comparing the relative sizes of the potassium, rubidium and caesium ions with the six-oxygen window diameters in X and Y that the above observations are readily explained.The ionic radius of the ammonium ionz1 is 0.143 nm, which lies between those of potassium and rubidium; its diameter is in fact almost identical to the window size. It is therefore probable that this accounts at least partially for the variability in maximum exchange level that is observed when replacing sodium by ammonium ions in X. It is not obvious however why this variability should be much greater in X than in Y if ion sieving is the only important consideration. The structure of X is only marginally more open than Y,22 and in fact there is now much evidence that in the case of Y an exchange level which is > 70% does not necessarily imply that the ingoing ions have even removed all the original ions from the super cage^.^^^^^ Thus, for example, a potassium Y sample exchanged with ammonium ions to a level of 72% * Samples were pretreated with saturated ammonium chloride solution in order to remove sodium.1s 57-21748 ION EXCHANGE IN SYNTHETIC FAUJASITES still contained seven potassium ions per unit cell in the type I1 More recently, CremersZ5 has given collected evidence for ion redistributions taking place during exchange, and Vansant and Uytterh~even~~ agree with Theng et aL3 in emphasising that ‘the maximum limit to exchange was determined by the same factors as the exchange ~electivity’.~~ They also note, in common with experimental observations made here, that the maximum exchange level obtained varied more with temperature in the faujasites with the higher silica: alumina ratios.Thus, in this present work (table 2) it was found that the maximum level of exchange of ammonium for sodium in Y was ca. 90% at 353 K using similar solution concentrations to Herman and Bulko.6 In contrast, raising the temperature of exchange with X had no effect on the observed maximum exchange level of 73% (table 2). It is apparent therefore that differences in framework charge density between X and Y zeolites are a primary factor affecting the Na and NH, exchange levels and selectivity in these zeolites. Framework charge densities can be considered using simple dielectric theory, or from a statistical thermodynamic viewpoint. APPLICATION OF DIELECTRIC THEORY Application of simple dielectric theory leads to the expressionl19 2 6 t 27 where rNHl and rNa are the ionic radii of ions NH: and Na+, E,, E, are the permittivities of the crystal and solution phases, respectively, e is the charge on the electron and N is Avogadro’s constant. Since26 E , < E,, and rNH4 > rNa, eqn (8) leads to the prediction that the standard free energy of exchange for the exchange reaction given in eqn (1) should be negative for all zeolites.This prediction is borne out by experimental data in the literature for a range of zeolites (table 3). In addition, it has been shownll that simple dielectric theory leads to the conclusion that if the same exchange is observed in two zeolites of differing framework charge where the subscripts ‘ hc’ and ‘lc’ refer to ‘high charge’ and ‘low charge’, respectively.Thus, if (as is the case here) dielectric theory predicts that AG* should be negative for the exchange of ammonium in sodium zeolites, then the theory also implies that AG* should be less negative for the zeolite of higher framework charge. This prediction is also confirmed from a plot of AG* values for the Na + NH, exchange against (Al), (fig. 5), where NA1 = Nsi + NA1 and NA1 and Nsi are the number of aluminium and silicon atoms, respectively, per unit cell of zeolite. The trend observed in fig. 5 involves data from zeolites of very different structures, and exemplifies the comments of Barrer and Davies2’ on simple dielectric theory, viz. ‘this model is simple in principle although possibly only semi-quantitative in application.It is however the most useful because no information is required about the geometric arrangement of cations. ’ The broad agreement between the predictions of simple dielectric theory and the plot in fig. 5 thus emphasises the significant contribution that the framework charge makes to the magnitude of AG*, irrespective of the different framework structures. Nevertheless, the overall significance of the trend observed in fig. 5 should be assessed in the light of the following points.P. FLETCHER AND R. P. TOWNSEND 1749 X X \ \V\ \ \ \ \ '.A 0 ? \ \ \ \ 0 \ FIG. 5.- Values of AG* for the Na S NH, exchange reaction in different zeolites plotted as a function of the equivalent fraction of aluminium in the framework (Al)c: x , clinoptilolite [ref.(30) and (31)]; V, synthetic mordenite [ref. (28)]; a, chabazite [ref. (29)]; A, zeolite Y [ref. (2), (3) and this work]; 0, zeolite X [ref. (3)]; 8, zeolite X (this work); (-------) best fit of results for all the synthetic zeolites excluding the X data. (i) Except for the value of A G e obtained for X by Theng et al., all the data for synthetic species follow a very clear trend. AG* decreases almost linearly with increasing aluminium content in the framework from synthetic mordenite (-AG* = 3.76 kJ mol-l) to X(-AG* = 0.53 kJ mol-l). (ii) Although all the measured AG* values for the natural zeolites are negative in agreement with prediction [see inequality (9)], the data lie well off the line followed by the synthetic species. Allowance can be made for the fact that the A G e values for c l i n ~ p t i l o l i t e ~ ~ ~ ~ ~ were measured at 303 and 333 K using the AH* value given by Howery and but the correction does not alter significantly the magnitude of AGe.The problems involved in purifying natural zeolite^,^ the care necessary in order to obtain homoionic forms of (for example) ~linoptilolite~~ and especially the ammonium form7133 make data for natural zeolites difficult to compare with those of the synthetic materials. In the context of these reservations, and considering therefore only the free energy values obtained for synthetic species in 5g. 5, it does appear that the value obtained for the N a e N H , exchange in X by Theng et aIq3 is rather high. STATISTICAL-THERMODYNAMIC CONSIDERATIONS Using statistical thermodynamics, Barrer and Klinowski3, have shown that the activity coefficients of the exchanging ions in the crystal phase are related to the framework charge density :1750 ION EXCHANGE IN SYNTHETIC FAUJASITES COAA is an excess interaction energy which arises when pairing of the entering A': ions takes place.q is related to the framework charge density by where N is the number of ion sites in the crystal and No the number of charges (i.e. equivalent to the number of framework aluminium atoms). Eqn (1 1) and (12) were derived assuming a random siting of cations.34 For uni-univalent exchange, such as the ammonium-sodium exchanges considered here, eqn (1 1) and (1 2) reduce to and Barrer and Klinow~ki~~ demonstrated that for this very simple case, eqn (14) give a theoretical basis to the original semi-empirical approach of Kielland.35 If eqn (14) are applicable to the Na + NH, exchanges in X and Y, inspection of eqn (14) shows that fNa and fNH, [which were phenomenologically determined using eqn (5) and (6)] should vary with (NH,), and (Na), in a symmetrical manner, a criterion which is met well with Y (fig.4). Applying eqn (14) to the Na$NH4 exchange in (for example) X gives P(NH,),N - 11 (15) ~ A A In (fNH,(X)/fNa(X)) = ___ VXkT when the subscript X refers to zeolite X. A similar expression holds for Y. Thus plots of ln(fNH,/fNa) against (NH,), should yield plots with intercepts equal to -(wAA./qkT), and gradients twice these values and of opposite sign. Using the data shown in fig. 4, plots of eqn (15) for both X and Y were constructed, and are shown in fig.6. The data conform well with the simple statistical-thermodynamic model3, in the case of Y, but the fit for X is much poorer. Correlation coefficients on the linear best fits to the data are found to be = 0.997 and R(x) .= 0.945, respectively, where (as usual) R = o,m/oy, with ox and oy the standard deviations ( N weighting) of the dependent and independent variables, respectively, and rn is the gradient of best fit. Comparing either the best fit gradients or the intercepts for X and Y in fig. 6 yields (16) a very similar result, uiz. Since X and Y are isostructural, it follows from eqn (13) and the analytical data in qXwAA(X) = o*4qY0AA(Y). table 1 that qy/rx = Nx/Ny = 1.435 so that wAA(X) = o*279wAA(Y) (18) where COAA(X) and wAA(y) are the interaction energies involved when the entering ammonium ions pair in X and Y, respectively.Examination of fig. 6 shows that both COAA(X) and wAA(y) are positive. To evaluate these, it is necessary to know qx and qy, which requires a knowledge of N [eqn (1 3)]. N is difficult to estimate, but since many of the ions in the supercages are unsited2*24 it seems likely that N c No for both zeolites. Values of wAA(X) and w A A ( ~ ) determined on this assumption are shown in table 4.P. FLETCHER AND R. P. TOWNSEND 1751 0.0 0.2 0.4 0.6 0.8 (NH,): FIG. 6.-Plots of the crystal-phase activity correction factor as a function of (NH,),N for zeolites X and Y. TABLE 4.-vALUES OF WAA, THE INTERACTION ENERGY BETWEEN AMMONIUM IONS, CALCULATED FOR DIFFERENT VALUES 0.25 0.100 248 0.358 887 0.50 0.200 496 0.717 1777 0.75 0.300 743 1.075 2664 1 .oo 0.400 99 1 1.434 35531752 ION EXCHANGE IN SYNTHETIC FAUJASITES Several comments are made on these data.(i) Barrer and Klinowski calculated34 theoretical isotherm plots for values of K, > 1 when o,,/kT was small and positive. These compare well with the experimental data found here {compare normalised plots of X and Y in fig. 1 (this work) with fig. 3(c) [l-41 for X and fig. 3(a) [2] for Y in ref. (34)). (ii) When coAA/kT > 0, the entering cations (ammonium in this case) avoid each The higher the value of oAA/kT, the stronger is this repulsion. Thus in Y, the ammonium ions avoid each other more strongly than in X, [Note that wAA for a given ion pair is not an intrinsic property of that ion pair, and therefore may vary from zeolite to zeolite. COAA is an excess energy function and refers to prescribed reference (in this case Na-X and Na-Y, respectively).It is therefore quite consistent that COAA for the same ion pair can be markedly different in different zeolites.] (iii) From the comments under point (ii) it is clear that the COAA values are consistent with the view that in X the framework charge is a dominant factor determining ammonium ion selectivity and sitings, whereas in Y the interactions between the ammonium ions themselves are far more important. This is consistent with Sherry’s view.2 (iv) It follows from point (iii) that ions in the supercages of Y are more randomly distributed than in X, since the important factor in Y is not ion binding to the framework2 but rather, as shown above, the tendency of the ammonium ions to avoid one another.Since eqn (1 1) and (12) were derived assuming a random distribution of cations, it follgws that the data for Y should fit eqn (15) much better than those for X. This is indeed the case (fig. 6 ) . CONCLUSIONS Experimental data and theoretical considerations combine to show that the ammonium ion is reluctant to enter the sodalite units in X and Y, and that the ammonium ions are more mobile and less strongly bound in Y than in X. It is probable therefore that the observed increase in maximum exchange level of ammonium ion with temperature in Y occurs by an initial migration of sodium ions out of both sites I and the sodalite units into the supercages, which is then followed by exchange, the ammonium ions then remaining in the supercages.In X, the sodium ions are less mobile and are bound much more strongly to the zeolite framework. This decrease in mobility, together with the reluctance of the ammonium ions to enter the sodalite units, makes the maximum exchange level for ammonium ion in X less sensitive to temperature changes. Dielectric theory provides a rationale for the observation that Y is more selective than X for the ammonium ion, and statistical-thermodynamic theory confirms that the ammonium ions bind less strongly to the framework in Y than in X. P. F. gratefully acknowledges a Scholarship from the British Gas Corporation, and subsequently a Research Fellowship from the City University.R. M. Barrer, W. Buser and W. F. Grutter, Helv. Chim. Acta, 1956, 29, 518. H. S. Sherry, J . Phys. Chem., 1966, 70, 1158. B. K. G. Theng, E. Vansant and J. B. Uytterhoeven, Trans. Faraday Soc., 1968, 64, 3370. H. S. Sherry, in Zon Exchange-A Series of Advances, ed. J. Marinsky (Marcel Dekker, New York, 1969), vol. 2, p. 89. P. P. Lai and L. V. C. Rees, J . Chem. Soc., Faraday Trans. I , 1976, 72, 1809.P. FLETCHER AND R. P. TOWNSEND I753 fi R. G . Herman and J. B. Bulko, in Adsorption and Ion Exchange with Synthetic Zeolites, ed. W. H. Flank, (ACS, Washington, D.C., 1980), vol. 135, p. 177. R. M. Barrer and R. P. Townsend, J. Chem. Soc., Faraday Trans. 1, 1976, 72, 661, 2650. R. M. Barrer and R. P. Townsend, J.Chem. Soc., Faraday Trans. 1, 1978, 74, 745. P. Fletcher and R. P. Townsend, Proc. 5th Int. Con$ Molecular Sieves, Naples, 1980, ed. L. V. C. Rees, (Heyden, London, 1980), p. 3 1 1. P. Fletcher and R. P. Townsend, J . Chromatogr.. 1980, 201, 93. * l P. Fletcher and R. P. Townsend, J. Chem. Soc., Faraday Trans. 1. 1981, 77, 497. l 2 P. Fletcher and R. P. Townsend, ‘Transition Metal Ion Exchange in Zeolites. Part 5’, paper in preparation. A. Dyer, H. Enamy and R. P. Townsend, Sep. Sci. Technol., 1981, 16, 173. R. M. Barrer, J. Klinowski and H. S. Sherry, J . Chem. Soc., Faraday Trans. 2, 1973, 69. 1669. D. W. Breck, Zeolite Molecular Sieres (Wiley-Interscience, London, 1974), p. 532. l 4 G. L. Gaines and H. C. Thomas, J . Chem. Phys., 1953, 21, 714. Ifi E. Glueckauf, Nature (London), 1949, 163, 414. IH L. V. C. Rees, personal communication. IY R. M. Barrer and J. Klinowski, J . Chem. Soc., Faraday Trans. I , 1974, 70, 2080. J. Klinowski, in The Properties and Applications of Zeolites, ed. R. P. Townsend (Special Publication, The Chemical Society, London, 1980), vol. 33, p. 288. 2 1 F. A. Cotton and G. Wilkinson, Adcanced Inorganic Chemistry (Interscience, New York, 1966), Lz D. W. Breck, Zeolite Molecular Sieces (Wiley-Interscience, London, 1974), pp. 176 and 177. 2:1 E. F. Vansant and J. B. Uytterhoeven, Adv. Chem. Ser., 1971, 101, 726. 24 W. J. Mortier, M. L. Costenoble and J. B. Uytterhoeven, J. Phys. Chem., 1973, 77, 2880. 25 A. Cremers ACS Symp. Ser., 1977, 40, 185. 2fi R. M. Barrer, L. V. C. Rees and M. Shamsuzzoha, J . Inorg. Nucl. Chem., 1966, 28, 629. 2 i R. M. Barrer and J. A. Davies, J . Phys. Chem. Solids, 1969, 30, 1921. 2x R. M. Barrer and J. Klinowski, J. Chem. Soc., Faraday Trans. I , 1974, 70, 2362. ey R. M. Barrer, J. A. Davies and L. V. C. Rees, J . Inorg. Nucl. Chem., 1969, 31, 219. :H’ D. G. Howery and H. C. Thomas, J . Phys. Chem., 1965, 69, 531. 31 R. M. Barrer, R. Papadopoulos and L. V. C. Rees, J . Inorg. Nucl. Chem., 1967, 29, 2047. 32 A. Ardya and A. Dyer, J . Inorg. Nucl. Chem., 1981, 43, 589. 3 3 M. Loizidou and R. P. Townsend, paper in preparation. 3 4 R. M. Barrer and J. Klinowski, Philos. Trans. R . SOC. London, 1977, 285, 637. p. 334. J. Kielland, J. Soc. Chem. Ind.. 1935, 54, 232T. (PAPER 1 /807)