Site Group Interaction Effects in Zeolite-Y Part 1.-Structural Examination of the First Stages of the Ag Ion Exchange BY MARTIN COSTENOBLE AND ANDRE MAES* Katholieke Universiteit Leuven, Centrum voor Oppervlaktescheikunde en Colloidale Scheikunde, De Croylaan 42, B-3030 Heverlee, Belgium Received 13th October, 1976 The localization of Na+ and Ag+ ions in hydrated zeolites NaAg-Y containing 2,7.25 and 14 Ag+ ions/u.c. is studied by X-ray diffraction. A high preference of Ag for site I is observed. On increasing the silver content the Ag occupancy of site I reaches a maximum value of %4 Ag/u.c. at the expense of Na+ ions in site I’. The Ag to Na preference pattern derived is; site I > site I1 > site I’ W unlocalized. The majority of ion exchange data on zeolites-X and -Y have been explained in terms of a static model of neutralization in small and large cavities, combined with hydration energy effects and the crystallographic radius of the exchanging ions.The number of possible neutralization points being in excess of the number of exchangeable cations, repositioning can occur during ion exchange.2 9 For example, the distribution in hydrated Na-Y is different from that of the K analog~e,~ pointing to the flexibility in network neutralization. However, two main constraints seem to limit the cation positioning. First, Mortier et aL4 state that each framework oxygen should within small limits bear t of the charge. Secondly there are exclusion rules which must be obeyed. Both the characteristics of the ions and the ion exchanger determine the selectivity of an exchanger for a certain cation.Recently, Maes and Cremers ti argued that cation positioning in small cages influences the selectivity in the large cages and vice versa. In Part 2 7 more evidence is presented showing the influence of the different nature of the ions in the large cavities on the Na+-Ag+ selectivity in the small cages. Studies of cation location in hydrated zeolites are still relatively scarce and are mostly restricted to the homoionic forms. It is the purpose of the present study to investigate cation positioning in hydrated zeolite-Y containing two different cations. In particular, three compositions, corresponding to the initial stages of the Ag+-Na+ equilibrium are investigated by powder X-ray diffraction techniques.The distribu- tion of the exchangeable cations over the different crystallographic sites is discussed in terms of the crystallographic radius, hydration energies and electronegativities of the two competing ions. EXPERIMENTAL SAMPLE PREPARATION A Linde Y-zeolite saturated with Na+ ions by repeated exchange with 1 mol dm-3 NaCI solution, subsequently washed until Cl- free (liquid/solid ratio on a weight basis ~ 3 5 ) ~ and air dried is called Na-Y and contains 55 Na+ ions/u.c. From the ion exchange data in the Part 2,’ zeolite-Y is shown to be initially very selective for Ag+. Three samples were 131132 SITE GROUP INTERACTION IN ZEOLITES prepared by adding a specific amount of a AgN03 solution to air-dried zeolite (liquid/ solid ratio -4). Overnight equilibration was followed by centrifugation and decantation of the supernatant liquid.The samples were air dried without previous washing, as the Ag+ content in the equilibrium solution was respectively 0.05,l and 5 % of the total solution ion content, resulting in negligible Ag in the interstitial solution of the sediment. The air-dried samples were analysed for Ag by atomic absorption methods after acid breakdown and contain respectively 2, 7.25 and 14 Ag+ ions/u.c. The sample table specifies the synthetic faujasite (Y), the ions (Na+/Ag+) and the amount of Ag+ ions/u.c. (e.g. Na/7.25 Ag-Y). The site notation of Smith is used. X-RAY DIFFRACTION AND CALCULATIONS The recording of the X-ray diffraction spectra, the data sampling and the calculation procedures are given elsewhere.lO The scattering due to unlocalised cations and water molecules was accounted for by applying a liquid scattering function.The electron densities of the ions and water molecules were summed and compared with those expected for water molecules. The unlocalised contribution was smeared out in spheres of radii 0.25 and 0.57 nm, corresponding to the sodalite units and large cavities respectively. An additional temperature factor of 15 x nm2 was applied. The residual of the intensities was defined as RI = z ~ ( k Io- IJ/& k Io. A least-squares refinement was carried out with the full-matrix program POWOW,ll based on intensities. The following weighting scheme was introduced The program ORFFE l2 was used to calculate the interatomic distances and angles.RESULTS AND DISCUSSION Table 1 lists the extra framework located material and other relevant information, such as unit cell dimensions, RI values and interatomic distances within the structures Na-Y,I0 Na/2Ag-Y, Na/7.25Ag-Y and Na/14Ag-Y. The structure parameters, framework interatomic distances and bond angles and the values of I. and Ic are not essential to the discussion and are not included in the results. These data are available upon request. Since three kinds of electron density (Ag+, Na+, HzO) can overlap on the same site the following guidelines were used to interpret the observed electron densities at a specific distance from the framework oxygen. The theoretical ionic distances are Na+-02- = 0.235 nm, Ag+-02- = 0.2266 nm and H202--02- = 0.280 nm.According to the short distances observed (see table 1) on site I (site 1-03) and site I1 (site 11-02) these electron densities are attributed solely to Ag+ and Na+ ions. The long site 11'--02 distance suggests that only water occupies site 11', as is generally accepted.13-l The framework of the relaxed hydrated faujasite structures is neutralized according to a specific pattern.4* lo Nevertheless the number of charges coordinated around the O3 oxygens (site I+site 1') is approximately constant: 18.5 in Na-Fj,lG 16 in Na,Ca-Fj l4 and 17.3 in Na-Y.I0 Since even in the case of the Na/l4Ag-Y sample the overall exchange of Na+ by Ag+ ions had only proceeded to -25 %, it would be expected that no dramatic change in the neutralization pattern of the studied hydrated structures will occur.The unit cell dimension is constant throughout the sequence of samples, which indicates the absence of any distortive influence of the Agf ions.M . COSTENOBLE AND A . MAES 133 The interpretation of the electron densities around the O3 oxygens is based on 17.5 charges, as found in the Na-Y parent material. The differentiation between Ag+ and Na+ is obtained according to the constancy of the number of charges around each oxygen4 and taking into account that the electron density of 1 Ag+ ion is similar to 4.5 Na+ ions. The Ag+ to Na+ conversion factor arises from the interpretation of the respective atomic scattering factors in the sin O/h range scanned in the experiments. The electron density observed at site I in the Na/2Ag-Y structure can be assigned to 1.9 Ag+ ions or 8.6 Na+ ions.However it is improbable that the introduction of 2 Ag+ ions results in a migration of 8.6 Na+ ions to site I, since in hydrated Na-Y lo no electron density was observed in the hexagonal prism (see table 1). TABLE AS ASSIGNMENT OF ELECTRON DENSITIES AT THE DIFFERENT SITES AND INTERATOMIC DISTANCES IN NM a site I site I’ site 11‘ site I1 total localized Ag chem. determined Ag unlocalized Ag total charge on sites I +I’ + I1 Tnteratomic distances site I’-Os - 0 2 -site I’ -site 11’ site 11’-O2 - 0 4 -site I‘ -site 11’ site 11-02 4 3 4 -site 11’ site I-O3 Rr a0 Na-Y - 1 7.3( 1 .3)Na+ 13.4( 1.1)HzO 10.2( 1.6)Na+ - - I 27.5 0.25 3 (9) 0.307(10) 0.3 8 8( 1 8) 0.259( 16) 0.283(13) 0.372(13) 0.259( 16) 0.343(25) 0.279( 13) 0.3 19( 13) 0.346( 18) 0.278(3) 0.1658 2.4700(3) Na /2Ag-Y 1.9(2)Agf 1 5.7( 1 .O)Na+ 15.7(1.3)HzO 9.7(1.9)Na+ 1.9 1.9 27.3 - 0.255(6) 0.309(6) 0.3 89( 12) 0.247( 12) 0.296( 11) 0.388(11) 0.247( 12) 0.298 (22) 0.253(20) 0.307(20) 0.342(23) 0.279(2) 0.1535 2.4690(4) Na17.25Ag-Y 3.9(4)Ag+ 13.5( 1.7)Na+ 0.3(4)Agf 8,0(2.7)Na+ 1.8(6)Ag+ 6.0 7.25 1.25 27.50 23.5(3.2)HzO 0.250(5) 0.311(5) 0.392(8) 0.249(6) 0.249(6) 0.384(6) 0.249(6) 0.253( 8) 0.302( 8) 0.352(9) 0.277(3) 0.1747 2.4690(2) - Na/ 1 4Ag-Y 4.4(6)Ag+ 1 1.7(9)Na+ 1.8(4)Ag+ 4.0(3)Na+ 6.0(7)Ag+ 12.2 14.0 1.8 27.90 18.3(3)H20 0.247(4) 0.312(4) 0.3 8 8 (6) 0.249(6) 0.296(5) 0.3 8 1 (6) 0.249( 6) 0.307(9) 0.255(5) 0.303(5) 0.340( 6) 0.272(2) 0.1376 2.4690(4) a Estimated standard deviations on the last figures in parenthesis.Upon increasing the Ag content to 7.25 ions/u.c. the observed electron density on site I is equivalent to 3.9 Ag+ ions. Using the same argument as for Na/2Ag-Y it is physically impossible to attribute the electron density in site I to 17.6 Na+ ions, since the exclusion rule (I+$ I’ < 16) cannot be obeyed in this case. According to the premise that w 17.5 charges are located around the O3 oxygens in all the hydrated samples, and assigning the entire electron density in site I to 3.9 Ag+ ions, the distribution in site I’ must be 13.5 Na+ and 0.3 Ag+ ions. The observed electron density on site 11 being equivalent to 16.1 Na+ ions, exceeds that for the Na/2Ag-Y and Na-Y and is thought to be due to the presence of silver.I34 SITE GROUP INTERACTION I N ZEOLITES By analogy with Na-Y lo and Na/2Ag-Y where 10 charges (Na+ ions) (see table 1) were located on site 11, 8 Na+ ions and 1.8 Ag+ ions are assigned to site 11.On the basis of the preceding arguments the entire electron density on site I is attributed to 4.4Ag+ ions in the Na/l4Ag-Y structure. Taking account of the standard deviation it is seen in table 1 that the observed distances in all the samples are very similar to the Na-Y. The 1’-03 distance (0.247+_0.004nm) in the Nal 14Ag-Y sample has a tendency to be even smaller than in the other structures, suggesting that minor amounts of Ag+ are present on that site. Since, moreover, it was assumed that the charge distribution/oxygen ratio will only slightly vary, 1.8 Ag+ and 11.7 Na+ ions are assigned to site 1’.The sum of the charges around 0, then amounts to 17.9 units ofcharge/u.c. (see table 1). The high electron density observed on site I1 suggests a major Ag+ occupancy. Keeping the number of charges on site I1 equivalent to 10 the observed electron density is satisfied by 4 Na+ and 6 Ag+ ions, In Part 2 the selectivity of Ag+ for Na-Y will be studied.’ The high preference for Ag+ in the early stages of the Na+-Ag+ exchange in the presence (ternary system) or absence (binary system) of a large excess of Cs or NHZ, will be considered to originate from a high preference for a few Ag+ ions in the small cages, the ideal coordination in the hexagonal prism being the most probable position. From the foregoing X-ray assignments in the sequence of samples studied it emerges that site I is preferentially occupied by Ag+ ions at low loadings of silver in the zeolite.Beyond a limit of ~4 Ag+ ions, additional introduction of Ag+ into the zeolite fails to increase site I occupancy, but results in a distribution over different sites 1’, I1 and the unlocalized part according to the preference pattern : site I1 > site I’ 21 unlocalized. The preference of Ag+ for site I may originate from its large radius. The ideal distance Ag+-W- = 0.266 nm, which is close to the observed distance site I-O3 of 0.275 nm. However, K+ which is theoretically still better suited (Kf-02- = 0.273 nm; 1’-03 = 0.280 K-Y)4 to fit the hexagonal prism has an insignificant occupancy on site I.4 The higher hydration energy of Agt- and its ability in general to form complex ions must therefore favour the formation of a type of coordination compound with the 6 oxygens of the hexagonal prism.Sodium forms a stable coordination complex with H20 in the sodalite units lo and does not occupy site I. Silver with its still higher than Naf hydration energy is reluctant to remain in the sodalite cages. The diameter of the Ag+ ion being 0.06 nm larger than Na+ ion, it is sterically hindered in its coordination with site 1‘, while site 11’ is occupied by the H20, and therefore Ag+- prefers site 11. The occupancy in the small cage sites seems to be a subtle inter-play between hydration and coulombic forces governed by steric factors. As a consequence, the distribution of exchangeable cations over small and large cavities fails to follow a fixed pattern. The specific behaviour of any kind of ions or mixture of ions along with the pore system characteristics of zeolite-Y result in a neutralization which cannot solely be discussed in terms of small and large cavities.The authors thank the Belgian Government (Programmatie van het Wetenschaps- beleid) for financial support. H. S. Sherry, Adu. Chem. Ser., 1971, 101, 350. B. K. G. Theng, E. F. Vansant and J. B. Uytterhoeven, Trans. Firday Soc., 1968, 64, 3370. E. F. Vansant and J. B. Uytterhoeven, Adu. Chern. Ser., 1971, 101,426. W. J. Mortier and H. J. Bosmans, J. Phys. Chem., 1971, 75, 3327. W. J. Mortier, H. J. Bosmans and J. B. Uytterhoeven, J. Phys. Chem., 1972, 76, 650. A. Maes and A. Cremers, J.C.S. Faraday I, 1975, 71, 265.M. COSTENOBLE AND A . MAES 135 ’ A. Maes and A. Cremers, J.C.S. Faraday I, 1978, 74, 136. A. Maes and A. Cremers, Adu. Chem. Ser., 1973, 121,230. J. V. Smith, Ado. Chem. Ser., 1971, 101, 171. lo M. L. Costenoble, W. Mortier and J. B. Uytterhoeven, J.C.S. Faraday I, 1976, 72, 1877. l1 W. C. Hamilton, POWOW (Brookhaven National Laboratory, Brookhaven, New York, 1962). l2 W. R. Buning, K. 0. Martin and H. A. Levy, ORFFE (Oak Ridge, National Laboratory, l3 D. H. Olson, J. Phys. Chem., 1970, 74, 2758. l4 W. M. Baur, Amer. Mineral., 1964, 49, 697. l5 J. J. Pluth and J. V. Smith, Mar. Res. Bull., 1973, 8,459. l6 T. H. Hseu, Ph.D. Thesis (University of Washington, 1972). Oak Ridge, Tenn., 1964). (PAPER 6/1919)