J. CHEM. SOC. FARADAY TRANS., 1994, 90(9), 1345-1350 1345 Zinc-exchanged Y Zeolites studied with Carbon Monoxide and Xenon as Probes Bruno Boddenberg" and Andreas Seidel Lehrstuhl fur Physikalische Chemie 11, Universitat Dortmund, Otto-HahnStr. 6, 0-44227 Dortmund, Germany The adsorption isotherms of carbon monoxide and xenon, as well as the 12'Xe NMR chemical shifts of xenon, in 55 and 74% zinc-exchanged Y zeolites with various pretreatment conditions, such as temperature of dehydra- tion, oxidation and preloading with CO, have been measured. The sets of experimental data can be explained quantitatively with a unifying approach which considers localized adsorption of both CO and xenon on well defined adsorption sites consisting of two types of zinc cations in different environments, and sodium cations.The two different zinc cation sites are characterized by their individual adsorption constants and isosteric heats of CO adsorption (63 and 46 kJ mol-') as well as by the '*'Xe NMR chemical shifts of the xenon atoms accommodated on them (185 and 135 ppm). The concentrations of these sites as a function of the degree of zinc exchange and pretreatment conditions were determined. Transition-metal-exchanged zeolites are interesting catalysts for a variety of chemical reactions.' An important question related to the performance of such catalysts concerns the location, oxidation state and distribution of the transition- metal species in the zeolite voids accessible to reactant mol- ecules. It is well known that only species in certain oxidation states and in well defined geometrical and chemical environ- ments within the aluminosilicate framework of zeolites provide catalytic activity.Although at present zinc-exchanged faujasite-type zeolites do not seem to be catalysts of great importance and have only scarcely been explored with respect to catalytic activ- ity,2-8 they appear well suited for basic studies aimed at the determination of the concentration and characteristic adsorp- tion behaviour of -active sites in zeolites. The literature reports investigations of the adsorption of CO in zinc-exchanged Y-type zeolite^,^.' which indicate the presence of a limited number of Zn2+ species at or near the supercage walls with different CO adsorption capabilities.The method employed here consists of the combined appli- cation of CO and xenon adsorption techniques as well as '29Xe NMR spectroscopy. The latter, which was introduced by Ito and Fraissard," sensitively probes the physical and chemical properties of zeolite voids accessible to xenon, whereas CO adsorption is a standard technique for locating the transition-metal cation centres that exhibit enhanced CO bonding capability. The combination of both methods should therefore provide more detailed information about cation concentration and distribution in transition-metal-exchanged zeolites. Experimental Zeolite NaY (LZ-Y52, Union Carbide, Si : A1 = 2.37 :1) was treated twice with 0.1 mol dm-, NaCl aqueous solution with subsequent intense washing, dried at 80"C and then rehydrated over saturated NH,Cl solution.The ion exchanges were performed with 0.1 mol dm -Zn(NO,), aqueous solutions at pH 5.7-5.9. After drying at 80°C (6 h) the zeolites were stored over saturated NH,Cl solution. The samples prepared are designated Zn(55)Y and Zn(74)Y, indi- cating the percentage of Zn2+ exchanged for Na+. The results of the chemical analysis for zinc (ICP/OES), sodium (FAES) and water are collected in Table 1. A differential thermal analysis/thermogravimetry (DTA/TG) investigation showed that the dehydration is completed at 300 "C and that the zeolites are stable up to 890 [Zn(55)Y] and 910°C CZn(74)Yl.The zeolites were dehydrated under high vacuum at 25 "C (ca.2 h), 120°C (6 h) and 400°C (16 h) with heating at rates of 20 and 60 K h-' in the intervals between. Subsequently, the samples were cooled to ambient temperature after either no contact or a 16 h contact with oxygen (400 hPa) at 400°C. In a similar way, samples dehydrated at a final temperature of 200 "C were prepared. For a precise characterization of the variously pretreated zeolites, the final temperature and treat- ment with oxygen (if applied) are indicated in the sample designation, e.g. Zn(74)Y/673ox. In their rehydrated form the samples were examined with X-ray diffraction (XRD) as well as with 29Si and 27Al magic- angle spinning (MAS) NMR spectroscopy. Sharp XRD lines characteristic for faujasite' ' with lattice constants a = 2.468 [Zn(55)Y] and 2.475 nm [Zn(74)Y], single 27Al NMR lines at 6 = 60.6 [Zn(55)Yl and 59.4 ppm [Zn(74)Y] relative to 0.1 mol dm-3 AlCl, solution, and 29Si NMR quintets with amplitude ratios corresponding to Si : A1 ratios of 2.2 :1 [Zn(55)Y] and 2.4 :1 [Zn(74)Y]12 were obtained.These results prove that crystallinity is retained after the zinc exchange and the subsequent dehydration/oxidation pro-cedures. A 75-80% zinc exchange seems to be the critical limit for framework stability since >80% zinc-exchanged Y zeolites have been reported to lead to framework collapse.2 The adsorption isotherms of CO and xenon were measured volumetrically using conventional glass and all-steel equipments, respectively. The adsorption was reversible, and equilibrium was attained within 30-60 min in both cases.The 29Xe NMR measurements were performed at ambient tem- perature and at the resonance frequency, 0,/27t = 21.4 MHz, using a Bruker (Karlsruhe, Germany) spectrometer type CXP 100. In each case studied, the '29Xe NMR spectrum was found to consist of a single resonance line. The chemical shifts were evaluated according to 6 = 1OB(vprobc-V,,~)/V,~~, with xenon gas at vanishing pressure taken as the reference. Table 1 Cation and water concentrations of the investigated zinc- exchanged Y zeolites number of species per unit cell Zn2+ for Na+ sample Zn2+ Na+ H20 exchange level (%) ~~~~~ ~ ~~ Zn(55)Y 15.8 20.7 253 Zn(74)Y 21.0 10.6 244 55 74 J. CHEM. SOC.FARADAY TRANS., 1994, VOL. 90 Results Fig. l(a)and (b) show the adsorption isotherms of CO in the zeolites Zn(x)Y/673ox (x = 55 and 74) at temperatures between 273 and 309 K. For comparison purposes, the ambient temperature CO adsorption isotherm of zeolite NaY dehydrated at 693 K is included.13 After initial steep rises up to pressures of several hPa, the isotherms merge into straight lines of slopes that decrease with temperature and also with the degree of zinc exchange. Fig. 2(a) and (b) show the ambient temperature adsorption isotherms of CO in the zeolites Zn(x)Y/673 (x = 55 and 74) and in the zeolite Zn(74)Y/473ox. For comparison purposes, the ambient temperature isotherms of the 673 K oxidized zeolites and of NaY are taken from Fig.1. Basically, the iso- therm shapes of the zinc-exchanged zeolites are the same as described before. Interestingly, the oxidation following the dehydration at 673 K leads to a decrease of the CO adsorp-tion. At the lower dehydration temperature (473 K) the amount of adsorbed CO is drastically reduced. Fig. 3 and 4 show the ambient temperature xenon adsorp- tion isotherms and the '"Xe NMR chemical shifts of xenon in the zeolites Zn(55)Y and Zn(74)Y dehydrated and oxidized at 673 K, as well as in Zn(74)Y/673ox preloaded with 1.7 CO per unit cell (uc) (0.21 CO per supercage). The corresponding data for zeolite NaY dehydrated at 693 K13-" are included for comparison. In contrast to Nay, the isotherms of the zinc-exchanged zeolites are each concave to the pressure axis initially and merge into straight lines at higher pressures.The preloaded zeolite exhibits lower xenon adsorption than both non-preloaded samples over the whole pressure range investi- gated. Of the latter zeolites, the sample with the higher zinc concentration exhibits larger xenon adsorption at pressures 5 4 u 53 P 0 $2 1 3 0 52 P 0Y 21 0 2 4 6 8 10 P/hPa Fig. 1 Adsorption isotherms of CO in the zeolites Zn(74)Y/673ox (a),Zn(55)Y/673ox (0)and NaY/693 (x): (a) in the range 0-300 hPa, (b) in the range 0-10 hPa. The symbol diameters in ascending order refer to the measuring temperatures 273, 298 and 309 K. The solid lines are fittings according to the model discussed in the text.5 4 0 53P 0 $2 1 0 3 0 z2 0 0 0, p1 Fig. 2 Adsorption isotherms (298 K) of CO in the zeolites Zn(74)Y/ 673 (m), Zn(74)Y/673ox (O), Zn(55)Y/673 (O), Zn(55)Y/673ox (O), Zn(74)Y/473ox (A)and NaY/693 (x). (a) and (b) as for Fig. 1. The solid lines are fittings according to the model discussed in the text. Fig. 3 Adsorption isotherms (298 K) of xenon in the zeolites Zn(55)Y/673ox (O),Zn(74)Y/673ox (O), Zn(74)Y/673ox preloaded with CO (+) and NaY/693 ( x ). The solid lines are fittings according to the model discussed in the text. 160 140 120 100 80 0 5 10 15 N/Xe per uc Fig. 4 12'Xe NMR chemical shifts (298 K) of xenon in the zeolites Zn(55)Y/673ox (0),Zn(74)Y/673ox (O), Zn(74)Y/673ox preloaded with CO (+) and NaY/693 ( x ).The solid lines are fittings according to the model discussed in the text. J. CHEM. SOC. FARADAY TRANS., 1994, VOL. 90 up to about 500 hPa, but the reverse is true at higher pres- sures. Interestingly, each of the adsorption isotherms of the zinc-exchanged samples intersects the NaY curve. The '29Xe NMR chemical shifts of xenon in each of the zinc-containing zeolites are higher than those in Nay. With increasing xenon loading they first decrease and then increase again, approaching straight lines that run almost parallel to the line found for xenon in Nay. Discussion Adsorption of Carbon Monoxide Carbon monoxide is known to be much more strongly adsorbed in zinc-exchanged zeolites Y than in their precursor form This behaviour has been attributed to the for- mation of strong complexes of CO with the accessible Zn2+ cations within the faujasite ~tructure.~~~ From IR2*16 and adsorption isotherm' studies it has been concluded that two types of Zn2+ species interact with CO molecules, which at ambient temperature cannot penetrate into the small j?-One of these Zn2+ cations was found to be present to an appreciable extent at zinc exchanges of 60%.The initial steep increase of the adsorption isotherms of CO in the zinc-exchanged zeolites (Fig. 1 and 2) and their subsequent almost linear courses indicate the presence of a limited number of strong adsorption sites that become satu- rated at pressures of several hPa, and of weak adsorption sites with characteristics similar to those of Nay.Assuming a 1 :1 correspondence between CO and the strong adsorption sites, the concentration of these sites can be estimated by extrapolating the linear isotherm portions to zero pressure. For instance, the ordinate intersections in the case of the zeolite Zn(74)Y/673ox at the three different temperatures investigated (Fig. 1) yield a common value, Np+O,of three CO molecules uc- '. Fig. 5 shows the isosteric heats of adsorption, 4s1,of CO in the zeolite Zn(74)Y/673ox as function of CO concentration. The data points shown were obtained with the aid of the well known equation" d ln(p/hPa))4,=,,.( dT from isosteres constructed by interpolating the adsorption isotherm data at the three measuring temperatures.Their accuracy was estimated to be fi5 kJ mol-'. The results obtained suggest a three-plateau feature with plateau heights at about 65, 45 and 20 kJ mo1-'. The latter value agrees quite well with literature data (23-25 kJ mol-') obtained for CO adsorbed at low concentration in the zeolite NaY.9918919 80 r 60 20 I I 1 0 1 2 3 4 N/CO per uc Fig. 5 Isosteric heats of CO adsorbed in the zeolite Zn(74)Y/673ox. Dotted lines show plateau levels at 20,46 and 63 kJ mol-'. The solid curve represents the heats calculated from the model discussed in the text. This indicates that the linear portions of the adsorption iso- therms are due to the supercage Na+ cations. Remarkably, the former two values are of the order of the low-coverage heats of adsorption of CO on the surface of ZnO (42-50 kJ mol -1)20-23 suggesting that these heats are associated with the zinc species in the zeolites under study.The discussion carried through so far suggests that the adsorption isotherms of Fig. 1 and 2 can be represented by a superposition of three isotherms, of which one is of the Henry type. Owing to the rather low concentration of the strongly adsorbing sites, it seems appropriate to choose Langmuir- type adsorption isotherms to represent the CO adsorption on them. So we write N=-+-nlklP n2kzP 1 + k,p 1 + k,p +K,P In this equation ni and ki (i = 1, 2) are the concentrations and adsorption constants of the strong adsorption sites which are assumed to accommodate one CO molecule each.Any attempt to reproduce the adsorption isotherms with one Langmuir and one Henry type expression remained unsuc- cessful. The fitting of the adsorption isotherms of Fig. 1 and 2 with eqn. (2) requires the determination of five parameters in each case. K, and the sum n, + n2 can be obtained from the slopes and ordinate intersections of the higher-pressure linear isotherm portions. Starting from appropriately chosen values of the remaining three unknown parameters, namely k,, k, and n,, an iteration procedure was applied which accom- plished optimum fits of the complete sets of the experimental isotherm data of Fig. 1 and 2 simultaneously. It was required that (i) samples with the same pretreatment exhibit identical values of n, and n2irrespective of the measuring temperature, and (ii) samples with different pretreatments exhibit the same values of k, and k, at a given measuring temperature.The main concern was to reproduce as faithfully as possible the low-pressure portions of the adsorption isotherms which required slight corrections of the values of K, and (n,+ n,) introduced initially. The results of the fittings are shown in Fig. 1 and 2 (solid curves) and in Table 2 where the optimum-fit parameter values are collected. These data are estimated to be correct within +_1digit of the last relevant figure. Obviously, the experimental results can be reproduced very well, especially as far as the low-pressure range is concerned.This lends strong support for the appropriateness of the underlying three-site model. Introducing the fitting parameters for the zeolite Zn(74)Y/ 673ox (Table 2) into eqn. (2) and constructing isosteres, iso- steric heats of adsorption at any CO concentration were calculated. These are represented by the solid curve in Fig. 5 which exhibits the three-plateau feature very well. From the data presented in Table 2 several interesting con- clusions may be drawn. The concentrations, n, and n, ,of the zeolites pretreated at 673 K both increase with zinc content. This observation indicates that the type 1 and 2 sites involve Zn2+ ions at or near the walls of the supercages of the fauja- site structure. The low concentration of these sites at the 55% zinc exchange level (ca.10% of the overall zinc content) and its considerably larger value at 74% exchange (ca. 15% of the overall zinc content) is in accordance with the previously mentioned findings of Otsuka et a!.' The treatment with oxygen of the zeolites dehydrated at 673 K decreases n2but leaves n, unchanged. This observation indicates that the type 1 ZnZ+ cations are in stable positions whereas some of the type 2 cations become displaced, probably into the sodalite cages, under high-temperature oxygen contact. The low J. CHEM. SOC. FARADAY TRANS., 1994, VOL. 90 Table 2 Parameter values used for the fits of the experimental adsorption isotherms of carbon monoxide Nay1693 298 Zn( 55)Y/673ox 298 0.26 1.3 Zn(55)Y/673 298 0.26 1.5 Zn(74)Y/673ox 273 1.2 1.8 Zn(74)YY/673ox 298 1.2 1.8 Zn( 74)Y/673ox 309 1.2 1.8 Zn( 74)Y/673 298 1.2 2.2 Zn(74)Y/473ox 298 0.18 0.18 values of n, and n2 obtained for the zeolite dehydrated and subsequently oxidized at 473 K are considered to be due to a screening of most of the Zn2+ ions by residual water mol- ecules and/or OH groups which prevent access of the CO molecules to these cations.At each measuring temperature, the adsorption constants k, and k2 differ from each other by more than one order of magnitude, indicating a much different strength of bonding of CO to these sites. We may write the temperature dependence of ki(i = 1, 2) as ki = ki, exp(q,lRT) (3) where 4iis a parameter associated with the heat of adsorp- tion on the respective site.17 Fig.6 shows the logarithmic plot of k, and k2 us. reciprocal temperature for the zeolite Zn(74)Y/673ox. Also shown is the corresponding plot of K, for this zeolite. Obviously, straight lines are obtained from which values of ki, and qiwere evaluated. The data obtained are collected in Table 3. The heats of adsorption of CO on the three types of sites considered, namely 63, 46 and 20 kJ mol-', are represented by the dashed lines in Fig. 5. Most satisfactorily, they are just at the plateau heights of the iso- steric heats, which lends further support to the three-site assumption introduced. The appropriateness of the Langmuir model to describe CO adsorption on type 1 and 2 sites requires that the values of the pre-exponential factors k,, and k,, (Table 3) are of physically reasonable magnitude.In order to prove this, we use the expression for the Langmuir constant k (for conve- nience we drop here the lower index i) provided by statistical mechanics24 In this equation 4' and U,, are the partition function and the potential energy of the adsorbed molecule, respectively, and q&* is the internal partition function of the molecule in the gaseous state. h and k, are Planck's and Boltzmann's con- stants, respectively. The two partition functions may be decomposed further according to (5) Here ql, qI1and q, are the partition functions of the vibra- tions perpendicular and parallel to the surface, and of the wagging-type motions, respectively.The latter motions corre- spond to the rotational degrees of freedom of the CO mol-ecules in the gas phase for which the partition function is 4,,,= 8n2Zk, T/h2.Here I is the moment of inertia. Finally, q:ib and 4Vpibare the stretching vibration partition functions of CO in the adsorbed and gaseous state, respectively. For the purpose of estimation we put qW= 1 and q1 =-ql1, and express the latter partition functions in the high-6.6 13 0.74 6.5 13 0.74 6.5 125 3.6 9.4 13 0.74 4.8 5.0 0.33 3.4 13 0.74 4.8 13 0.74 4.2 -5w7-5= I,.,,!,,,,,,l 3.2 3.3 3.4 3.5 3.6 3.7 103 KIT Fig. 6 van't Hoff plots of the constants k, (a),k, (0)and K, (x) of CO adsorbed in zeolite Zn(74)Y/673ox temperature approximation, i.e.(7) Here v, is the vibration frequency of the molecule with respect to the surface. The stretching frequencies of CO in the gaseous state (2143 cm-1)25 and adsorbed on Zn2+ in Y zeolites (in the range 2120-2220 cm-1)216 are close together so that the respective partition functions are the same within 1Oh. Introducing the approximations considered into eqn. (4), the pre-exponential factor is obtained to be from which relation the vibrational frequencies vSi (i = 1, 2) can be calculated from the respective values of k,. The resulting data are collected in Table 3. The frequency values obtained are of the expected magnitude for physisorbed molecules.24 The higher value of the frequency and, hence, of the force constant associated with the type 1 sites in compari- son to type 2 reflects the stronger bonding of CO to the former.This result is in accordance with the conclusions drawn from analysis of the heat of adsorption data. The nature of the sites of types 1 and 2 remains to be eluci- dated. Undoubtedly, these sites are associated with Zn2 + ions at the supercage walls of the faujasite structure. We imagine that type 1 zinc ions exhibiting the highest heat of adsorption towards CO (63 kJ mol-') have lower coordination to sur- Table 3 Parameters obtained from analysis of the temperature dependenceof the adsorption and Henry constants 1 k,/hPa -VSi/lO'~ s -qJkJ mol-' 1 9 x lo-" 1.6 63 2 5 x 10-9 0.43 46 3 --20 J.CHEM. SOC. FARADAY TRANS., 1994, VOL. 90 rounding framework oxygen atoms than the type 2 species. This is because the cations with lower coordination to oxygen atoms are expected to exhibit higher Lewis acidity towards the weakly Lewis basic CO molecules and/or, even- tually, provide easier access for the adsorbate species. These considerations suggest that type 1 zinc cations are located at or near the crystallographic S I11 positions, and type 2 zinc ions reside at or near the octahedral positions between the successive oxygen three-rings connecting the supercages with the sodalite cages. The latter positions are known to be occupied by, e.g., Cu2+ ions in copper-exchanged zeolite Y.26 Conceivably, under contact with CO, the zinc ions residing here may be pulled slightly towards the S I1 position.This effect of cation displacement is a known behaviour of, e.g., Cu' 27 and Ni2+ in faujasite-type zeolites. '29Xe NMR and Adsorption of Xenon It has repeatedly been demonstrated that cation sites which strongly adsorb CO also provide strong adsorption centres for xenon.13,28,29 The shape of the xenon adsorption iso- therms of the zeolites Zn(x)Y/673ox (x = 55 and 74) with initial concave portions and high-pressure straight lines of slopes decreasing with increasing zinc content (Fig. 3) is remi- niscent of the shape characteristics of the corresponding CO adsorption isotherms. This suggests that the rule cited at the beginning may be valid also for the zinc-exchanged zeolites studied here.In order to prove this, we use eqn. (2) likewise to treat the xenon adsorption data. As in the previous case of the adsorp- tion of CO, we assume a 1 : 1 correspondence between the xenon atoms and each of the sites available. Under the condi- tion of rapid exchange of the xenon atoms among the adsorp- tion sites available, the observable chemical shift, 6, may be formulated as where ai (i = 1-3) is the local chemical shift of a xenon atom accommodated on site i, and FN,, is a collective term taking into account the xenon density dependence of the chemical shift. 3*30,31The rapid-exchange condition is considered to be valid in the present case because a single resonance line is always observed. The fitting of the experimentally determined adsorption isotherm and chemical shift data of xenon in the zeolites Zn(55)Y/673ox and Zn(74)Y/673ox requires the determi-nation of k,, k,, K,, F and 6, -6,.The site concentrations, n, and n,, are taken over from the analysis of the CO adsorption results (Table 2). The parameter F is assumed to exhibit the same value as for Nay, namely F = 1.84 ppm uc, which is evaluated from the slope of the straight 6 us. N line of NaY (Fig. 4). Except for K,, the further parameters are assumed to be identical for both zinc-exchanged zeolites under consideration. Starting from appropriately chosen values of the parameters, an iteration procedure was applied to yield a simultaneous optimum fit of the sets of adsorption isotherm and chemical shift data.The solid curves in Fig. 3 and 4 demonstrate that both the adsorption isotherms and the chemical shifts of xenon in the zeolites without preadsorbed CO can be reproduced excel- lently. The values of the parameters, with which these fits were accomplished, are collected in Table 4. It is recognized that the adsorption constants of xenon follow a gradation similar to those of CO. However, k, and k, for xenon are about two orders of magnitude smaller than the correspond- ing values for CO, whereas the Henry constants describing the strength of adsorption on type 3 sites show contrasting behaviour, but are of a similar magnitude in both cases. Importantly, the procedure applied here, allows an accu- rate extrapolation of the chemical shifts to zero xenon con- centration (6,= o), and also the determination of site-specific chemical shifts for cases where the shifts increase at low xenon concentrations.The extrapolated values are found to be S,=, = 158 and 120 for the zeolites Zn(74)Y/673ox and Zn(55)Y/673ox, respectively, and the site characteristic shifts are 6, = 185 and 6, = 135. The shift 6, = 64 obtained for the type 3 sites is close to a,=, = 58 for xenon in NaY.'0*'3,'5 This result supports the previous conclusion that type 3 sites are sodium cations at the supercage walls. The fact that the adsorption constants k, and k, for CO are both much larger than either of the corresponding values for xenon, suggests that CO preadsorbed on sites 1 and 2 effectively blocks these sites for the xenon atoms. Knowing the number of blocked sites and assuming that the blocking does not create new sites, the adsorption isotherm and 129Xe NMR chemical shift data of xenon adsorbed in the preloaded zeolite should be predictable quantitatively with eqn.(2) and (9), and the parameter data of the unpreloaded zeolite Zn(74)Y/673ox, but with n, and n, replaced by the corre- sponding concentrations of the non-blocked sites. These con- centrations are obtained from the CO adsorption data as follows: The concentration, 1.7 CO uc-', of preadsorbed CO in the zeolite Zn(74)Y/673ox is maintained at an equilibrium pressure 0.71 hPa at 298 K. At this pressure the concentra- tions of type 1 and 2 sites occupied by CO molecules are calculated to be 1.1 and 0.6 sites uc- I, respectively, using the Langmuir-type expressions of eqn.(2) with the values of the respective adsorption constants k, and k, of Table 2. Conse- quently, the concentrations of the type 1 and 2 sites still accessible to xenon are n, = (1.2-1.1) uc-' = 0.1 uc-l and n2 = (1.8-0.6) uc-l = 1.2 uc-l, respectively. These values are listed in Table 4. The xenon adsorption isotherm and chemical shifts with the data thus obtained (last line of Table 4) and with F = 1.84 ppm uc given before, are represented by the solid curves in Fig. 3 and 4. Obviously, excellent agreement with the experimental data is obtained. Conclusions The study of zinc-exchanged zeolites with adsorption of xenon and carbon monoxide in combination with 12'Xe NMR spectroscopy has turned out in this investigation to be very useful for discriminating between zinc cations in differ- ent environments at or near the supercage walls of Y-type Table 4 Parameter values used for fits of the experimental adsorption isotherms of xenon and the I2'Xe NMR chemical shifts NaY/693 0 0 Zn(55)Y/673ox 0.26" 1.3" Zn(74)Y/673ox 1.2" 1.8" Zn(74)Y/673ox (1.7 CO uc-') 0.1 1.2 " The site concentrations were adopted from Table 2. 58 2.16 185 135 64 0.05 0.008 1.72 185 135 64 0.05 0.008 1.44 185 135 64 0.05 0.008 1.44 1350 J.CHEM. SOC. FARADAY TRANS., 1994, VOL. 90 zeolites. The whole body of experimental data of zinc-exchanged Y-type zeolites with different degrees of zinc exchange and various sample pretreatments comprising tem- perature of dehydration, oxidation and preloading with CO can be reproduced quantitatively on the basis of a simple adsorption model.Adsorption constants and 129Xe NMR 8 9 10 11 M. Ziolek and J. Kujawa, Zeolites, 1990, 10, 657. T. A. Egerton and F. S. Stone, J. Chem. Soc., Faraday Trans. I, 1973,69, 22. T. Ito and J. Fraissard, Proc. 5th Int. Con$ Zeolites, Naples, 1980, p. 510. R. von Ballmoos and J. B. Higgins, Collection of Simulated XRD Powder Patterns for Zeolites, Butterworth-Heinemann, Stone- chemical shifts characteristic for xenon atoms adsorbed on ham, 2nd edn., 1990. well defined adsorption sites were obtained. This is of great importance for the analysis of adsorption and chemical shift data of zinc-containing zeolites, or other microporous materials of unknown cation concentrations, and of the dis- tribution of zinc cations within them.The method should be applicable to other cations in zeolites, and thus an atlas of site-characteristic adsorption constants and 12'Xe NMR chemical shifts could be compiled. The 129Xe NMR chemical shifts of xenon adsorbed on the two zinc cation sites detected, presumably Zn2+ ions at the crystallographic S I11 and S II/S 11' positions, were deter-mined to be + 185 and + 135 ppm with respect to xenon gas at vanishing pressure. 12 13 14 15 16 17 18 19 20 21 G. Engelhardt and D. Michel, High Resolution Solid State NMR of Silicates and Zeolites, Wiley, Chichester, 1987. M.Hartmann and B. Boddenberg, Microporous Materials, 1994, 2, 127. J. Watermann, Diploma, University of Dortmund, 1989. T. Geilke, Diploma, University of Dortmund, 1990. C. L. Angel1 and P. C. Schaffer, J. Phys. Chem., 1966,70,1413. A. W. Adamson, Physical Chemistry of Surfaces, Wiley, New York, 5th edn., 1990. T. A. Egerton and F. S. Stone, J. Colloid Interface Sci., 1972, 38, 195. T. A. Egerton and F. S. Stone, Trans. Faraday SOC., 1970, 66, 2364. R. R. Gay, M. H. Nodine, V. E. Henrich, H. J. Zeiger and E. I. Solomon, J. Am. Chem. SOC., 1980,102,6752. A. A. Tsyganenko, L. A. Denisenko, S. M. Zverev and V. N. Financial support of this work by Fonds der Chemischen Industrie is gratefully acknowledged. 22 Filimonov, J. Catal., 1985,94, 10. E. Giamello and B. Fubini, J. Chem. SOC., Faraday Trans. I, 1983,79,1995. 23 G. L. Griffin and J. T. Yates, J. Chem. Phys., 1982,77,3751. References I. E. Maxwell, Adu. Catal., 1982,31, 1. K. Otsuka, J. Manda and A. Morikawa, J. Chem. SOC., Faraday 24 25 T. L. Hill, An Introduction to Statistical Thermodynamics, Addison-Wesley, Reading, 1960. D. H. Rank, D. P. Eastman, B. S. Rao and T. A. Wiggins, J. Opt. SOC. Am., 1961,51,929. Trans. I, 1981,77,2429. M. A. Wassel, E. A. Sultan and F. M. Tawfik, Asian J. Chem., 26 27 I. E. Maxwell and J. J. de Boer, J. Phys. Chem., 1975,79, 1874. J. Howard and J. M. Nicol, Zeolites, 1988,8, 142. 1992,4, 891. 28 R. GroDe, A. Gedeon, J. Watermann, J. Fraissard and B. Bod- M. F. Menoufy, E. A. Sultan and A. K. El-Morsi, Trans. Egypt SOC. Chem. Eng., 1990,16,33. H. Mori, N. Mizuno, T. Shirouzu, S. Kagawa and M. Iwamoto, 29 30 denberg, Zeolites, 1992, 12,909. Y-Y. Huang, J. Catal., 1974, 32,482. J. Fraissard and T. Ito, Zeolites, 1988, 8, 350. Bull. Chem. SOC. Jpn., 1991,64,2681. Y. V. S. Narayana Murthy and C. N. Pillai, Synth. Commun., 31 C. J. Jameson, A. K. Jameson and S. M. Cohen, J. Chem. Phys., 1973,59,4540. 1991, 21, 783. M. Ziolek, H. G. Karge and W. NieDen, Zeolites, 1990,10, 662. Paper 3/05015G; Received 18th August, 1993