JOURNAL OF
MOLECULAR CATALYSIS Journal of Molecular Catalysis 90 ( 1994) 323-354
ELSEVIER
Solid-state reaction as a way to transition metal cation introduction into high-silica zeolites Alexei V. Kucherov, Alexei A. Slinkin* N.D. Zelmsky Institute of Organic Chemistry, Ruwan
Academy of Smmxs,
Moscow, Russian Federation
(Received January 25, 1993; accepted November 30, 1993)
Abstract A detailed review is given of transition metal cation incorporation into zeolites by a solid-state reaction. The procedures and techniques suitable for investigation of solid-state ion exchange in zeolites are described. It is shown that a solid-state reaction is a promising way for introduction of one or several transition metal ions into cationic positions of high-silica zeolites. The majority of these polyvalent cations cannot be Introduced into these matrices by other methods. The influence of different factors (temperature and atmosphere of calcination, A13+ content in zeolitic lattice, presence of either co-cations or anionic species) on cationic species stabilization is analyzed. It is shown that pentasil-based systems with discrete types of isolated transition metal cations are unique objects for studying the relationship between catalytic properties and local topography of isolated redox sites. A solid state modification of pentasils permits the preparation of active and stable catalysts both for complete oxidation of alkanes and for NO, decomposition.
1. Introduction In the search for optimized advanced catalysts, an attractive strategy is to start from known host and stable structures and to elaborate a new architecture of active site inside the space of molecular dimension. Zeolites have been highlighted as a particularly promising class of 3-dimensional host frameworks with controlled microporosity. Synthesis of inorganic biomimetic catalysts on the basis of high-silica zeolites containing isolated transition metal cations as active redox sites may present promising opportunities for creation of new types of catalysts. Atomic-scale engineering of catalytic functions of isolated redox sites in the confined environments of zeolitic channels is also of great interest. However, conventional ion exchange from solutions [ 1,2] cannot be used for introduction *Corresponding
author; fax. ( + 7-095) 1355328, E-mail:
[email protected].
0304.5102/94/$07.00 0 1994 Elsevler Sctence B.V. All rights reserved .SSDf0304-5102(93)E0320-G
of the most interesting polycharged ions into the small cavities of high-silica zeolites. Most of the transition metal cations are strongly solvated, and prevented by their hydration shell from penetrating into the narrow channels of the zeolite structure. Therefore, it is necessary to design other ways of cation introduction. The thermal stability of high-silica zeolite frameworks allows the examination of the topochemical process of cation introduction by a solid-state reaction between zeolites and different compounds. The aim of the present review is both to summarize the dataon the solid-state introduction of transition metal ions into cationic positions of high-silica zeolites, and to elucidate the nature and catalytic properties of isolated redox site stabilized in the zeolitic matrix.
2. Solid-state exchange procedure and experimental
techniques
The typical procedure of solid-state ion exchange is quite simple: mechanical mixtures of powders of the zeolite and an oxide (or salt) of the cations to be introduced are ground and calcined in air, in inert gas or in high vacuum. Soluble compounds might also be supported on the zeolite by an incipient wetness impregnation to provide a more homogeneous distribution of the oxide (or salt) on the surface of the zeolite. The temperature required for a noticeable rate of solid-state exchange depends on the nature of both the cations and anions involved. In some cases, the nature of the gas phase also influences the process of ion migration upon calcination of zeolite/oxide mixtures. The quantity of polyvalent cations which can be introduced by a solid-state exchange is limited, and the upper limit is determined by the number of accessible vacancies (acid sites). The presence of strongly bound Na+ ions m cationic positions prevents the stabilization of polyvalent transition metal cations. Thus, the H’ or NH: forms of zeolites are the preferred starting materials. Many experimental techniques are available for studying solid-state ion exchange (IR spectroscopy, thermogravimetric analysis (TG), MAS NMR, XPS, analysis of HZ0 or HCI evolved). However, in the case of many transition metal ions (CL?+, Mn’+, CT+, Fe’+, V 4+, Cr’+, Mo5+) electron spin resonance (ESR) is the method of choice. The intensities of the lines and the parameters of the spectra allow the determination of the cation concentration and the identification of the type of coordination, respectively. In the case of isolated ions stabilized in high-silica zeolites, ESR spectroscopy can clarify the coordination of transition metal cations. The study of their interaction with different molecules is also possible. With IR spectroscopy both qualitative and quantitative analysis is feasible in many cases. The change in absorbance of OH stretching bands allows one to register the consumption of OH groups upon ion-exchange. Sorption of probe molecules ( NH3, pyridine, NzO. CO, CH,CN) can indicate the cations themselves. XPS of surface layers of zeolitic crystals permits one to follow changes in the surface concentration for supported compounds, indicating the redistribution of the transition metal cations between the zeolite and oxide (or salt) phases. The kinetics of weight loss upon the reaction of an oxide or a halide with an H form of the zeolite can be measured by TG analysis. The interaction between H zeolites and halides
A. V. Kucheror, A.A. Slinkm /Journal
ofMolecular Catalysis 90 (1994) 323-354
could be also studied by titration of evolved HCl. These techniques mation of the solid-state reaction, but also the determination of the Catalytic testing of parent and calcined [ zeolite/oxide (salt) ] unambiguously the change in both acid-base and redox properties reaction.
325
enable not only confirrate of exchange. mixtures demonstrates caused by a solid state
3. History of the problem and solid-state exchange with alkaline metal ions: brief remarks In the seventies interesting data were obtained which showed that ion exchange in zeolites might take place by interaction of the zeolite with the solid phase of alkaline metal salts. Thus, Rabo et al. [ 3-51 showed that Na, Ca, Be, and La forms of zeolite Y containing residual protons reacted with NaCl, at 575-675 K, with evolution of HCl.IR spectroscopy confirmed that this reaction eliminated acidic OH groups. Such removal of residual Bransted acid centers resulted in suppression of catalytic activity in the isomerization reactions. Clearfield et al. [ 61 demonstrated the introduction of cations, such as Zn’+, Ct?, Co’+, Ni2+, Mn’ * and Cr” *, into zeolites A, X and Y by reaction of their chlorides with the H forms of the zeolites. Evolution of HCl confirmed the reaction of OH groups with the salts, and occupation of cationic vacancies by the above-mentioned cations was indicated by the appearance of the typical ESR spectra. The latter were identical to those obtained from conventionally exchanged samples. Migration of transition metal ions in zeolites was discussed also in early papers [ 7-101, wherein the reduction and reoxidation of zeolites CuY, CUM, AgY and AgM were studied. The reduction of CuY at 400°C resulted in the formation of Cu’, part of which migrated out of the zeolite cages and formed large metallic crystallites [7]. Prolonged reoxidation, however, converted at least part of these crystallites into Cu’+ ions, that migrated back to specific locations in the small cages. Reduction of CUM by H, resulted in complete removal of cu’+ ions out of the mordenite channels [ 81. Cue crystals (average diameter 20 nm) formed on the outer surface of the zeolite. However, reoxidation at 450-500°C did not lead to the formation of large particles of CuO; instead, the formation of isolated CL?+ ions has been detected [ 81. Therefore, the possibility of migration of Cu’+ ions in H forms of zeolites in oxidative atmosphere was assumed. Also Ago agglomerates transformed almost completely into Agf cations in zeolites Y and M upon oxidative treatment of prereduced samples [9,10]. The above-mentioned earlier works gave rise to a steadily increasing interest in the phenomenon of solid-state ion exchange in zeolites. Many interesting aspects of solid-state ion exchange were reviewed recently by Karge and Beyer [ 111. In this brief review results of studies on introduction of alkaline, alkaline earth, rare earth and some transition metal cations into H, NH, and Na forms of zeolites were reported. Particular attention was paid to both the stoichiometry of exchange and the modification of acidic properties of zeolites. In our review we will discuss in detail the incorporation of transition metal cations into high-silica zeolites. Recent developments in this field give rise to the design of new promising redox catalysts.
326
A. V. Kucheror, A.A. Slir~ktr~/ Journal ofMolecular
Catalysts 90 (1994) 323-354
4. Solid-state exchange with transition metal ions 4.1. Copper ions 4.1.1. Incorporation of CLI(II) cations irtto tnorderlite and pentasils Systematic ESR studies on solid-state ion exchange with copper compounds were carried out on mordenite and ZSM-5 [ 12,131. The analysis of the results was facilitated by the availability of ESR spectra of copper( II)-containing samples prepared by conventional ionexchange in aqueous solution [ 14-171. cl?+ ion migration and redistribution upon the solid state interaction at 52G8OO”C between Cu’, CuO, CuCl,, CuF,, Cu,( PO_+)? or Cu,S and different samples of mordenite and ZSM-5 (Table 1) was studied [ 12,131. Thermal treatment resulted in the appearance of ESR spectra of isolated Cu2+ ions whatever the CL?+ compound. It is clear that the oxidative calcination resulted in the rapid formation of CuO both from Cu” and copper hydroxycarbonate. The dispersion of CuO influences the rate of solid-state interaction only. The rate of appearance of CL?+ ESR signals was dependent on the particle size of both the zeolite and the Cu-compounds; the larger the particles the lower the rate. However, upon calcination of different mixtures in severe conditions, maximum intensities of Cu( II) ESR signals were achieved, being determined by the type of zeolite only. Fig. la shows the ESR spectrum for the mixture of HM and CuO calcined in air at 550°C for 1 h and evacuated at 20°C. Ingress of air gave rise to a considerable and reversible broadening of lines (Fig. I b). The calcination of the mixture at 800°C for 1 h resulted in the significant increase of Cu(I1) ESR signal intensity with formation of the spectrum shown in Fig. lc. The oxidative treatment of a mixture of HM and Cu” powder gave rise to the same spectra, only the intensity of Cu( II) ESR signal for the (HM + Cu”), calcined at 550°C for 1 h was smaller than that in Fig. la by a factor of 5. The calcination of a mixture of NaM and CuO at 550-800°C did not cause the appearance of a Cu( II) ESR signal. Fig. 2a shows the ESR spectrum of a mixture of H-ZSM-5 and CuO calcined at 520°C for 1 h in air, and evacuated at 20°C. The calcination of the mixture at 800°C for 1 h resulted in the formation of the ESR spectrum shown in Fig. 2b. The treatment of the mixture at 800°C for 4 h caused no change in the spectrum (Fig. 2b). Entry Table 1 List of samples studied in [ 121 Sample
SiO,/AIZO, rat10
Na’ substituted for H’ (7~70)
NaM HM HNa-ZSM-5” H-ZSM-5 pentasil- 140” pentasil-280h amorphous sihca-alumina
10 10 69 69 140 280
0 95 40 95 95 95
14
_
“Prepared by the burmng-off bAnalogues of ZSM-5.
of orgamc cations from the sample with 407~ Na’ deficiency
A. I’ Kucherov, A.A. Slmkin /Journal
of Molecrrlnr Catalwrs 90 (1994) 323-354
327
b.
200 I
Oe I
H
Fig 1. ESR spectra, at 20°C. for the calcmed mixture of H-M and CuO: (a) calcined in air at 550°C for evacuated at 20°C; (b) after the ingress of air; (c) calcined in air at 800°C for I h and evacuated at 20°C
I h and [ 121
of oxygen led to a considerable and reversible broadening of the spectrum, as shown in Fig. 2c. The calcination of a mixture of H-ZSM-5 and CuO in high vacuum gave rise to the same picture (Fig. 2). The ESR spectra of a calcined mixture of HNa-ZSM-5 and CuO was less intense and differed in hyperfine splitting (hfs) parameters (Table 2). Therefore, it was concluded that Na+ cations block Cu( II) sites. The solid-state reaction between CuO and H-ZSM-5 samples of different SiOJAl,O, ratio gave Cu(II) ESR signals which were identical in hfs, but differed in the signal intensities. A linear correlation between maximum intensity of Cu( II) ESR signal and Al’ + content in ZSM-5 framework was established [ 12,131, demonstrating clearly that the number of cationic vacancies for CL?+ ions (protonic sites) is determined by the Al content in ZSM-5. The calcination of a mixture of H-ZSM-5 and dried CuCl, in high vacuum at 400°C for 0.5 h caused the appearance of the Cu( II) ESR signal shown in Fig. 3a. The treatment of
328
A V Kucheror.
A.A. Slinkm / Jmmal
oj~hfoledur
Catulwis
90
(I 994)
323-354
b.
Fig. 2. ESR spectra, at 20°C. for the calcined mixture of H-ZSM-5 and CuO: (a) calctned in air at 550°C for I h and evacuated at 2O”C, (b) calcined in atr at 800°C for I h and evacuated at 20°C; (c) after the ingress of air
[121. Table 2 ESR-parameters Sample
HM HNa-ZSM-5
of Cu’+ tons
[ 121
Copper compound
Calcinatlon temperature
cue cue
550 550 and 800 550
H-ZSM-5
cue
800
H-ZSM-5
CuF, or CUl( pwz
550 and 700
sI
Al, (Oe)
A, (Or)
2 325 2.37
2 055 2 075
I44 135
19 _
2 33 2.27 231 2.29 2.31 2 29 2.37 2.38
2.07 2 045 2 06 2.05
142 175 153 156 153 156 120 120
175 29
(“C)
23
the mixture at 550°C for I h resulted in a considerable increase of the ESR signal intensity. The calcination at 800°C for 1 h resulted in the formation of the Cu(I1) ESR spectrum shown in Fig. 3b. The thermal treatment at 550-800°C in high vacuum of the mixtures of H-ZSM-5 and CuF, or Cu,( PO,)z also resulted in the appearance of the ESR signals of isolated Cu’+ ions. X-ray analysis showed no destruction of the zeolitic framework upon the calcination. Fig. 4 gives the hfs on g,, for the mixtures of H-ZSM-5 and different copper compounds, calcined at 700°C for 1 h. The ESR parameters for the CL?+ ions in different samples are listed in Table 2.
A. V Kucherov. A.A. Shnkm / Jorrnml of Molecrrlar Catalysis 90 (I 994) 323-354
Fig. 3. ESR spectra, at 20°C. for the mixture of H-ZSM-5 and C&l2 calcmed m vacuum: for 30 mm. (b) calcined at 800°C for 1 h [ 121,
329
(a) calcined at 400°C
The results summarized in Table 2 [ 12,131 show clearly that the thermal treatment of heterogeneous mixtures of copper compounds and H-forms of high-silica zeolites results in the appreciable dispersion of copper and the migration of Cu(I1) ions into the zeolitic channels. In the case of CuO, the Cu’ l tons, which migrate from the outer surface of zeolite crystals, are coordinated in the same positions as Cu2+ cations, introduced by conventional ion-exchange [ 1.51. The ESR signal of the mixture (CuO + HM) calcined at 550°C for 1 h (Fig. la) is identical to the ESR spectrum of an exchanged sample KuH-M ( 1 Cu2+ ion per 3 elemental cells). Most of the copper in 8CuH-M is located as isolated Cu2+ cations in a square pyramidal coordination. The considerable change in Cu( II) ESR spectra in air (Fig. lb), caused by the dipole-dipole interaction of Cu2+ cations with oxygen, demonstrates that all cations are accessible for 0, molecules. Severer calcination causes considisolated Cu2 + erable increase of the Cu( II) ESR signal intensity, i.e., the quantity of Cu2+ ions migrating in the HM channels has increased markedly. At the same time, the form of the ESR signal (Fig. lc) is indicative of clustered Cu 2+ ions with dipolar interaction. The ESR spectrum in Fig. Ic is identical in both hfs and intensity to that of an exchanged sample SICuH-M (2 Cu2+ cations per elemental cell). The thermal treatment of the mixture (CuO + HM) therefore results in the gradual increase of CL?+ ion concentration in cationic positions of mordenite. In mild conditions, copper( II) location as isolated Cu’ l cations occurs; clustered cations are formed with increase in CL?+ content. The calcination of a mixture of CuO and an exchanged sample 5 lCuH-M did not result
Fig, 4. The hfs on R,, of ESR spectra, at - 196°C. for the mixtures of H-ZSM-S and different copper compounds. cdcmed at 700-800°C. (a) CuO; (h) CuF?. (c) CU~(PO,)~ L121
in any increase of the Cu( II) ESR line intensity [ 121. It was concluded that the quantity of copper which may be introduced into HM by a solid state reaction is limited to 50% of the exchange capacity. The upper limit for Cu(II) in mordenite is only determined by the number of accessible sites for their location and the solid-state interaction does not increase the number of Cu’ + cations in this zeolite. The Cu( II) migration into H-ZSM-5 structure was examined in detail [ 12,131. The hfs of spectra of the mixture of H-ZSM-5 and CuO, calcined at 550 and 800°C (Fig. 2a and 2b) were fully identical to that of the exchanged sample CuH-ZSM-5 [ 141. The samples, calcined at 550°C contain two types of isolated Cu2+ cations located in a square planar environment (g,, = 2.27; A,, = 175 Oe) and five-coordinated one (g,, =2.31;A,, = 142 Oe). The spectra of the samples, calcined at 800°C show the preservation of the two discrete types of Cu2+-’ ion sites but distorted with respect to the two previous ones. The strong influence of 0, on the hfs of Cu( II) ESR spectra (Fig. 2c) is also typical of the exchanged sample CuH-ZSM-5 [ 141. It is concluded that the Cu(II) ions, which migrate into H-ZSM5 from the outer surface of the crystals, are coordinated at the same positions as the isolated cl? + cations introduced by conventional ion-exchange. Upon calcination the number of isolated Cu*+ Ions in H-ZSM-5 reaches a maximum of 3040% of the framework Al?+ [ 12,131. “Weak associations” of Cu’+ ions are not formed at higher loading. The calcination of a mixture of CuO and exchanged sample CuH-ZSM-5 (20% exchanged forCu’+ ) resulted in the increase of Cu(I1) ESR signal intensity by a factor of 1.5. Thus, the solid
state exchange increases the quantity of isolated CL?+ cations in H-ZSM-5 in comparison with the ion-exchanged samples (three-fold exchange of NH,-ZSM-5 with Cu( NO,), water solutions). 4.1.2. Peculiarities of interaction betnleen H-ZSM-5 artd different compounds of copper The acid sites in high-silica zeolites may be considered as powerful traps for Cu( II) ions which migrate into zeolitic channels upon calcination. Formally, the process may be regarded as the reaction of copper compound with a strong acid (protonic sites in highsilica zeolites). In H-ZSM-5 Cu( II) compensates one elemental charge of the lattice and has an additional link to an extra-lattice ligand (Cu( OH) + ) [ 14,151. The study of the interaction of H-ZSM5 with CuCL, CuF, and Cu,( P01)2 [ 12,131 provided additional information on the influence of the anion (or extra-lattice ligand). The solid-state reaction of H-ZSM-5 with these compounds resulted in the appearance of the ESR signals from isolated CL?+ ions (Figs. 3 and 4; Table 2), i.e. the migration of Cu( II) ions to the cationic positions takes place in all cases. The appreciable migration of copper( II) at 400°C in the mixture of CuClz and HZSM-5 was explained by the low melting point of CuCI, (596°C). However, differences in the types of Cu(I1) coordination due to the presence of different anions were noted. At least 4 types of isolated CL?+ cation coordination were clearly distinguished in (H-ZSM5 + CuF,) or (H-ZSM-5 + Cu,( PO,),) samples (Fig. 4). Two types of coordination coincided with those in the (H-ZSM-5 + CuO) sample and two additional types corresponded to crystal fields of octahedral symmetry (Table 2). It was supposed that a considerable part of the isolated Cu’+ cations in zeolite channels interact with extra-lattice ligands, such as F- or PO:-. The study of interaction between H-ZSM-5 and Cu,S [ 131 demonstrated that copper( I) ions also migrate to cationic positions in zeolite channels, upon the high temperature treatment of the mixture of H-ZSM-5 and Cu,S in high vacuum. 4.1.3. Sitiilg of Cu’ + cations in H-ZSM-5 and mteraction with d#erent molecules Interaction of different molecules with Cu”+ cations located in H-ZSM-5 channels was studied in [ 16-2 1] and compared with the data of INDO calculations [ 18,191. The sample CuH-ZSM-5 with only one type of Cu2+ (1 Cu2+ ion per IO3 tetrahedra) showed, after dehydration at 673 K, only one species: g,, = 2.310; A,, = 172 Oe and g I = 2.052. For a ZSM-5 sample with 7 times greater loading of Cu2+, three Cu’+ species were detected after calcination at 673 K. Dehydration in high vacuum at 873 K was accompanied by the disappearance of one species and two species remained (g 1,= 2.3 1, A :, = Oe; g; = 2.30, A i = 175 Oe) Samples of CuH-ZSM-5 studied in [ 12-15,20,21], which have a significant copper loading ( 1 ion per 180 tetrahedra) and contain two types of cations (Table 2), may be compared with the sample studied in [ 171. There is only one XRD work [ 221 on the location of Nil+ in H-ZSM-5. Two sites were detected: the first site near the wall of the straight elliptical channel, and the second one inside the element T,401-0, composed completely of five-membered rings. The cation in the main channel is linked with three lattice oxygens and must be coordinated with extra-lattice ligands. The second site with six lattice oxygen ligands is recessed from the main channels of ZSM-5, and its geometry is presented in Fig. 5. In [ 18,191 it was assumed that Cu2+
kg 5 Cluster model [ CuL’ ( Ogm )6jv of isolated Cu” ion inside H-ZSM-5 structural five-membered rings (taken from X-ray data [ 12I ) ( I8 I.
fragment composed
of
cations are located in the same sites of ZSM-5 as Ni2+ cations, and the properties of such a location (Fig. 5) were analyzed by means of quantum-chemical calculations. It was shown that calculated components g, , and gz2 agree well with the g I value of 2.05 1 [ 18,191. A satisfactory agreement was noted also between the theoretical g3? value and g,,, obtained experimentally. At the same time, the quantum-chemical calculations showed that a small geometric displacement of the Cu 2+ ion ( LLZ) in its chelate site may result in a drastic change m the values of the g factor. A noticeable change in g,, components (especially for g33) accompanied even a negligible Cu ” ion displacement (dZ=O.Ol A). The sharp dependence of the g factor upon the change in donor-acceptor properties of the Cu2+ hgand environment, at LQ= 0, was also noted [ 18,191. In this case also a small change in the charge of Og- led to a significant change in g tensor components. especially g,,. Thus, it was concluded that effects other than coordinative complex formation in CuHZSM-5 channel intersections must be taken into account [ 18,191. The above-mentioned INDO calculations confirmed the surprising effect of measurable changes in the hfs of the CL?+ ESR signal upon the physisorption of Xe or n-hexane inside CuH-ZSM-5 channels [ 20,2 I 1. The sample CuH-ZSM-5 (0.6 wt.% Cu) studled was prepared by ion-exchange and then calcined at 8OO”C, but identical results were obtained with the sample 0.75 wt.% CuO/H-ZSM-5 prepared by a solid state reaction [ 20.2 1,231. Fig. 6 presents the parallel components of spectra, taken at - 196°C. The ESR spectrum of parent CuH-ZSM-5 (Fig. 6a) shows two signals from isolated Cu( II) ions. Sorption of Xe (200 mmHg) at 20°C leads to measurable changes in the hfs of the spectrum, as shown in Fig. 6b. An evacuation of the sample for 10 s results in complete restoration of the parent spectrum (Fig. 6a). Fig. 6c shows the changes in the parallel part of the Cu( II) ESR signal due to hexane sorption. Evacuation of the sample at 20°C for 1 h does not change the new ESR signal formed (Fig. 6~). However, a slow rise in temperature up to 200°C in vacuum is accompanied by gradual, slow restoration of parent ESR spectrum.
A.V Kucherov. A.A. Sltnkrn /Jotmud
oj’kkdecular
Catdyrs
90 (1994)
323-354
333
a.
B.
c.
-T-:
2500
3000
oe
Fig. 6. Part of the ESR spectrum (parallel components), at - 196°C. of Cu( II) cations m H-ZSM-5: (a) m vacua; (b) after the mgress of Xe; (c) after the sorption of n-hexane [20].
Therefore, the parent sample contains two discrete types of isolated five-coordinated Cu(I1) cations, located in square pyramids with rhombohedral distortions [ 12-141. The changes in the ESR parameters due to physisorption were noticeable for one of two types of Cu( II) cations. Thus, Cu( II) cations located in two sites which differ slightly in square pyramid distortion, differ markedly with respect to xenon (or hexane) sorption. It is difficult to imagine a linkage between Cu’+ ion and a saturated non-polar molecule (n-hexane) or inert gas atom (xenon). Such an intluence by non-localized physical adsorption was explained by a slight geometrical displacement of the Cu (II) ion due to dispersion (London) forces [ 20,2 11. The Cu( II) ion symmetry change due to the formation of a coordination bond is, however, most typical for the majority of compounds (unsaturated and polar molecules) [ l&19,2426]. The interaction of Cu” located in the position, shown in Fig. 5, with such a strong ligand as NH,, was analyzed by quantum-chemical calculation [ 18,191. In this case a discrepancy between calculated and experimental parameters was noted. Therefore, either NH, must be located far from Cu’+ or the interaction takes place with extraction of Cu2+ from the position shown in Fig. 5. That is, upon adsorption of polar molecules, such as water, ammonia and pyridine, CL?+ migrates into positions near the channel intersections of ZSM5 [ 16,251. The free space at the intersection of two channels permits the formation of species with four molecules of ammonia or pyridine. For low copper content CuNaH-ZSM-5 it was observed that CL?+ forms complexes with three molecules of pyridine and four of ammonia [ 16,251. In the work performed at higher Cu2 + loading the formation of complexes with two molecules of pyridine and three of ammonia was detected [ 171. Therefore, the con-
centration of copper appears to have a marked effect on the complexation of CuZt in the zeolite lattice. It was assumed [ 18,191 that ammonia sorption does not draw all the CL?+ ions to identical positions in channel intersections. It was supposed that slight changes in the CL?+ ESR parameters for one type of species is due to a weak interaction betw,een the CL?+ cations located in recessed positions and NH, molecules. The sorption of (CH,),O in CuH-ZSM-5, at 20°C. was accompanied by very noticeable changes in the CL?+ ESR spectra, resembling the changes induced by CH,OH sorption. It was concluded [25] that adsorption of methanol resulted in a copper complex involving two molecules of CHJOH complexed through the oxygen. Adsorption of ethanol gave [ 251 very similar ESR parameters but interaction with three molecules of ethanol was assumed from the spin-echo spectrum. It was concluded that complex formation requtres that Cu’+ be located at the intersection of two channels [ 16,251. However, it was difficult to suppose the formation of such a coordinative complex with the more bulky CH,OCH, molecules [ 17,181. It was assumed that both a small change in donor-acceptor properties of Cu2* ligand environment and a slight geometric displacement of Cu’+ ion in its chelate site caused by the CH,OCH, molecule may result in the change of the observed Cu’+ ESR signal [ 17,181. The ability of CL?+ in ZSM-5 to form complexes with methanol (or ethanol) depends also upon the co-cation [ 251. When the co-cation is Na’, K+ or Ca’+, Cu( II) coordinates to three molecules of CH,OH, but it coordinates to four alcohol molecules in H-ZSM-5 ~251. Upon ethene sorption on CuH-ZSM-5 two CL?+ complexes are formed [ 161. Equilibrium was reached slowly, and it was concluded that a very slow migration of Cu’+ into accessible cation sites took place. In a similar manner, benzene seemed incapable of causing the migration of Cu ‘+ into the main channels [ 161. No migration of CL?+ occurs as a result of C,H, or p-xylene sorption on CuH-ZSM-5, and the changes in ESR parameters were due to the displacement of Cu’+ cations in two different chelate sites [ 18,191. The interaction of CjH, molecules with both Cu2+ cations and redox sites (typical of pure H-ZSM-5) was registered [ 181. The ESR signal at g = 2.002 was identical to that of oligomeric cation-radicals, observed upon C,H, sorption on H-ZSM-5. The same effect was observed upon p-xylene sorption on CuH-ZSM-5: both Cu’+ ion complexation and pxylene cation-radical formation took place simultaneously. It was concluded that the introduction of Cu2+ cations in H-ZSM-5 does not lead to the disappearance of strong redox sites inside zeolitic channels [ 18,191. 4.2. Introduction
of Mo( V) cations into high-silica zeolites
The solid-state interaction between mordenite or ZSM-5 (Table I), and MOO, does not result in an ESR signal for Mo( V) ions after calcination in air or in vacuum [ 27,281. Reduction of both uncalcined and precalcined mixtures by H2 at 300-4OO”C produced identical weak Mo( V) ESR signals. MOO, forms polymer fragments in vapor phase but not cationic species. Therefore, Mo(V1) ions cannot enter the zeolitic channels, despite the high volatility of MOO,. After heating at 150°C the ampoule containing the mixture, H-ZSM-5 + MoCI,, m vacuum, a strong ESR signal was generated (Fig. 7a). Calcination of the ampoule at 350°C
A.V. Kucherov. A.A. Slirrkrn /Jomxal
of Molecular Cntuiysis 90 (1994) 323-354
335
resulted in the sublimation of excess MoCI, ( TbO,,= 268°C) but the intensity of the ESR signal did not decrease. Air was introduced at 20°C and this resulted in considerable broadening of the ESR signal (Fig. 7b). However, evacuation of the sample at 20°C for 1 min led to a complete restoration of the spectrum (Fig. 7a). After calcination of the mixture at 450°C for 1 h and at 550°C for 30 min, the ESR signal retained 70% of its initial intensity. Calcination of HM with MoCl, in vacuum produced an intense Mo( V) ESR signal but the interaction of NaM with MoCI, did not result in noticeable formation of isolated MO(V) ions. Calcination of the mixture, amorphous Si02-Al,O, + MoCl,, at 150°C led, as in the case of H-ZSM-5, to a considerable Mo( V) signal (I= 8 X 1 OL9spin/g; g = 1.95; LIH= 48 Oe) . However, calcination of the sample under severer conditions resulted in a continuous decrease in signal intensity. After calcination at 450°C for 30 min the signal intensity fell by a factor of two, and the signal disappeared completely when the sample was calcined at 550°C for 30 min. Thus, a solid-state interaction of MoCl, in vacuum with both zeolites and amorphous silica-alumina resulted in appearance of the intense ESR spectrum from isolated MO(V) ions (Fig. 7). High temperature treatment, however, showed that Mo( V) in cationic positions of zeolites was more strongly bonded than on the surface of SiO,-AI,O, [ 27,281. Mo( V) is not stable in H-ZSM-5 upon oxidative treatment [ 27,281. The calcination of a Mo( V) /H-ZSM-5 sample in air at 300°C resulted in irreversible disappearance of Mo( V) ESR signal shown in Fig. 7. The subsequent reduction by H2 produced only a very weak Mo( V) ESR signal. 3.3. Reactions behveerl morderlite or perltasils and chromium compounds The thermal treatment of mixtures of high-silica H-zeolites and CrOj or Cr?03 resulted in the appearance of an ESR spectra from isolated Cr(V) ions [ 27-301. Calcination of the
Fig. 7. ESR spectra. at 20°C. of the mixture [ H-ZSM-5 + MoCI,} calcined in vacuum at 150°C (a) in vacuum; (b) after the mtroduction of air [27]
336
A I’ Kucherov.
A.A Siinkm / Joumal
of Molecular
Caralps
90 (I 994) 323-354
different mixtures under severe conditions produced Cr(V) ESR signals, the maximum intensities of which were determined by the number of acid sites in the zeolites. The interaction of NaM with chromia did not result in the appearance of Cr( V) ESR spectrum. Fig. 8 shows the ESR spectra, at 20°C for the mixture of H-ZSM-5 and CrO,, calcined in air or in vacuum. Cooling the sample to - 196°C resulted in an increase in ESR signal intensity (g = 1.98; AH= 50 Oe). Introduction of air at 20°C gave rise to a considerable but reversible line broadening (Fig. Xc) [ 271. Such a change in the ESR spectrum in 07, caused by the dipole-dipole broadening of all components of hfs, demonstrated the accessibility of Cr(V) cations to Oz. Chemisorption of ammonia or pyridine at 20°C produced new ESR signals (superposition of two or three anisotropic signals from Cr( V) with no hfs structure) [ 27,281. The uncalcined mixture of H-ZSM-5 and 1.2 wt.% Cr,O, only showed a broad symmetric ESR line (AH= 460 Oe) at temperatures > 80°C. This signal is typical of the bulk antiferromagnetic a-Cr203. Calcination of the mixture in air at 800°C led to a considerable decrease in intensity of the broad line and a new narrow (AH = 50 Oe) ESR signal appeared at room temperature. The hfs and intensity of the ESR signal formed were identical to those shown in Fig. 8b. In contrast, calcination of the same mixture in vacuum did not result in a change in intensity of the broad line and no new ESR signal was seen. The results obtained show that Cr( V) is introduced via a solid state reaction into highsilica zeolites H-ZSM-5 or HM [ 27-301.
Fig. 8 ESR spectra of the mixture ( H-ZSM-S + 00,) at 20°C (a) Calcined m au at 550°C for I h and evacuated at 20°C. (b) calcmed in air at 800°C for 1 h and evacuated at 20°C. (c) after the mtroduction of air 127 I
A V. Kucherov,
A A. Sltnkttt / Journal
oj’M&cular
Catalyic
90 (I 994) 323-354
331
In ZSM-5, the average distance between Al atoms in the framework is large. Therefore, it is difficult to imagine the site capable of polyvalent cation coordination without additional ligands. It is suggested [ 27-301 that isolated complex cation species CrO+ , and not isolated ions Cr’ +, are coordinated in cation positions. The presence of hfs in the ESR spectra from Cr( V) cations located in H-ZSM-5 (Fig. 8) is an unusual peculiarity of these samples. This splitting demonstrates the interaction between the unpaired electron of Cr( V) and the nuclear spin (I= 5/2) of lattice Al’+ ions. ESR spectra with similar hfs were published for chromium-containing X and Y zeolites [ 31,321. The spectra from Cr( V) in ZSM-5 are very well resolved in the Q band (Fig. 9) [ 33,331. Sextets on g, (splitting 6 Oe) and g,, (splitting 7.5 Oe) were distinguished (Fig. 9a), showing the interaction of the Crv ion with one lattice Al”+ ion. It was supposed that the Al’+ ion is situated in the second coordmation sphere of the Cr” ion according to: 0
0
cr+& \ ./O\ /O\ ./ 7l\ /“\ s’\ \
The Cr( V) and Cr( III) distribution in H-ZSM-5 was investigated in relation to chromium content and redox treatment [29,30]. Most of the chromium cations in fresh samples, calcined in air at 140°C were present as chromate. But some Cr(V) cations were found inside the zeolitic channels in the form of hydrated octahedral complexes. Increasing the temperature gives an increase in the concentration of isolated Cr( V). However, the amount of isolated Cr( V) cations located inside the zeolitic channels does not exceed 20% of the number of Al’+ lattice ions. Excess chromium is located on the outer surface of the zeolite crystal in the form of Cr,O, crystals [ 29,301. A repeated redox treatment at 500°C of the
(0)
12.2
KCk
Fig 9. ESR spectra III Q band, at 20°C. of Cr(V) cations Introduced mto (a) H-ZSM-5; (b) H-[Ga]ZSM-5
[ 33 1
sample 2.5% Cr/H-ZSM-5 does not result in any change in the distribution Cr( III) ions both in the channels and on the outer surface of the zeolite. 4.4. Peculiarities
of V(W) cation introduction
of Cr( V) and
into high-silica zeolites
V( IV) is placed in cationic positions by the solid-state reaction of V20j with H-ZSM-5 [ 27,28,35-371. Calcination of mechanical mixtures of VzOs with H-ZSM-5 at 550-800°C. resulted in the appearance of V(IV) ESR signals (Fig. lOa), which are typical of isolated vanadyl species (g,, = 1.93; g, = 2.00; A,, = 198 Oe; A, = 85 Oe) [27,35]. Calcination of the mixture in air, in vacuum or in H2 (60 mmHg), led to spectra identical to that shown in Fig. 10a. The introduction of air at 20°C led to drastic but reversible line broadenmg. Calcination of HM with V,O, produced a broad ESR signal with poorly resolved hfs due to interaction between closely arranged vanadyl species. The vanadyl ion with 3d’ configuration is an electron analogue of Cr(V) ion, and superhyperfine splitting (shfs) due to A13+ .interaction is expected. A shfs of 7 Oe was detected by recording the ESR spectra at 200°C (Fig. 10) [ 361. This is experimental evidence of an electronic interaction between the vanadyl cation and lattice Al’+ ion of ZSM-5 [36]. Consequently, both ions must be in close proximity. A possible picture is shown below where the d,, orbital, carrying the unpaired spin, is in the place perpendicular to the paper and V=O is directed into the channels.
ESEM confirmed that paramagnetic vanadium species are formed upon a solid-state reaction of V,Os with H- and Ca-ZSM-5 zeolites. It was also shown that ( I ) the vanadium species were more accessible to H20 than to 02, and (2) the yield of vanadyl cations depended upon the cationic form of ZSM-5, heating conditions, and water adsorption. The yields were greater in H-ZSM-5 than in Ca-ZSM-5 [37]. ESEM results indicated that framework Al was located within 0.5 nm of the paramagnetic vanadium species [ 371. Differences in color and ESR parameters resulting from different sample treatments were noted. V(IV) ions in H-ZSM-5 can be reduced by H2 treatment [ 371. However, this was observed on a sample with 4 wt.% V,O,, i.e. with an excess of V,OT on the outer surface [ 27,281. Thus, the color change from yellow to grey may be due to V,OT. The interaction of isolated V( IV) cations in H-ZSM-5 with different molecules (ammonia, pyridine, 2,4,6_trimethylpyridine (TMPy), p-xylene, mtrobenzene) was also studied by ESR spectroscopy [ 351. It was shown that the initial ESR spectrum is split mto two spectra upon adsorption of all of the molecules studied except for TMPy. Smce adsorption
A.V.
Kuclwov.
A.A. Shkrn
/ Journal
of Molecular
Catalyws
90 (1994)
323-354
339
a.
b.
C.
Fig. 10. ESR spectrum of Isolated vanadyl cations (51V( IV), 3d’. I= 7/2) introduced into H-ZSM-5 by a solidstate reaction with Vz05 (a) at 20°C; (b), (c) fragments of spectrum (a) taken at 200°C. g,, = 1.93, gI = 2.00, A,,=2000e;A1=850e
[36].
of sterically hindered molecules like Th4Py did not affect the parameters of the ESR spectrum, it was concluded that both V( IV) species are localized inside the structure of the zeolite and not on the outer surface [ 351. The parameters of the ESR spectra changed little upon sorption of weak ligand such as p-xylene. However, the shfs lines from the interaction of the unpaired electron with the nuclear spin of “Al disappeared due to a slight displacement of V( IV) ions in chelate sites
340
A. V. Kucherm:
A A. Slrnkin /Journal
of Molecular Curulyrs 90 (1994) 323%3%
under influence of the adsorbate [ 351. This is analogous to observations on Cu( II) cations [20,21]. The p-xylene molecules not only interacted with cationic sites. but also reacted with redox sites, as indicated by the appearance of an intense ESR signal of cation-radicals (g = 2.004). Thus, incorporation of V( IV) into H-ZSM-5 does not block one-electron transfer sites in this zeolite [ 351. Adsorption of NH3, pyridine or C,H,NO? is accompanied by a significant change in the parameters of the V( IV) ESR spectrum, due to incorporation of these strong ligands in the coordination sphere [ 3.51. It was assumed that adsorption of ammonia or pyridine results in an increase in the covalency of the bonds due to delocalization of the unpaired electron from the &,. orbital to the orbitals of the ligands. On the contrary. in adsorption of nitrobenzene it was noted that delocalization of the electron decreases and the V-O bonds become more polar [ 351. The physical reason of this difference is not yet clear. The valence state of V( IV) in H-ZSM-5 channels is extremely stable upon redox treatment: the cation is not oxidized to V(V) even after heating in air at 800°C. This feature was discussed in [35] from the positions set out in [38], where it was shown that the parameter B = AG,, /AC, characterizes the strength of the V=O vanadyl bond. The values of B for V(IV) stabilized in H-ZSM-5 are unusually high, as compared with V205/Si02 or Vz0,/A1?03, and the V=O bond must be significantly stronger than that when vanadyl ions are supported on the surface of SiOz or AI,O, [ 351. However, the physical reasons of such strong stabilization of vanadyl cations by H-ZSM-5 structure are not yet clear. The solid-solid reaction between zeolite NaY and V,O, at 690-750 K was studied in recent work [ 391 with X-ray diffractometry, “Si NMR, adsorbed ‘19Xe NMR, ESR, and TEM. At low R values (atomic ratio R = V/ (Si + Al) = 0.05) the V,O, structure disappears, whereas the zeolite is not damaged nor is the cavity size reduced. However some cavity entrances were blocked by a compound containing vanadium atoms and, in particular, a few V( IV) atoms. It was supposed that a reaction between V205 and Na+ of NaY zeolite crystals may result in vanadium bronze phase (NaV;+ V”O,, ) formation. When the R value was higher, the zeolite lattice collapsed. Thus, an intensive solid-state interaction between zeolite and V,Os at quite low temperatures was observed [ 391.
4.5. Introduction
of nzarzgarzese cations into ZSM-5 zeolites
The state of Mn ions in ion-exchanged, ion-impregnated zeolites and in physical mixtures of Mn salts (or oxides) with H- or Na-ZSM-5 zeolites was investigated with ESR, XPS, IR, ammonia TPD and catalytic testing [40-42]. The solid-state reaction between Mn ions (in the form of MnO, Mn,O.,, MnCl,, or MnSO,) and the acidic OH groups of H-ZSM-5 takes place at temperatures above 700 K, resulting in incorporation of the Mn*’ ions into cationic sites. This is supported by: ( 1) the decrease of the number of strong acid sites; (2) the disappearance of Mn-containing phases and migration of the cations from the outer surface inside the zeohtic crystals; (3) the formation of isolated manganese Ions with defined coordination whose redox properties are identical with those of the Mn ions at the exchange sites. As a consequence, the catalytic activity of the modified samples in transformation of methanol and toluene was greatly decreased, similarly to the ion-exchanged zeolites [ 401.
A. V Kuclwov.
A.A. Slinkrn /Journal
of Molecular Catalysts 90 (I 994) 323-354
331
The upper limit of acid site replacement for Mn cations in H-ZSM-5 reached ca. 50% only. No insertion of Mn cations into Na-ZSM-5 was registered [ 40,411. The rate of Mn cation insertion into H-ZSM-5 depends on the nature of the manganese compound. The rate of the solid-state reaction decreased in the sequence: MnCl, > Mn30, > MnSO,, with MnCl, already reacting at 570°C [ 41,421. Part of the Mn ions stabilized in H-ZSM-5 channels as Mn2+, is very stable upon redox treatment in H, and O2 [ 40-421. However, the presence of Mn ions in some other valence states cannot be excluded. A peculiarity of the Mn’+ ESR spectra from MnH-ZSM-5 samples is that the typical well-resolved signal at g = 2.00 with six hyperfine lines (AH= 98 G) is observed in the X band for hydrated samples only. Dehydration of the samples resulted in a substantial broadening, making the signal unobservable [ 4&42]. 4.6. Interaction between H-ZSM-5 and iron compounds Location, coordination, and reactivity of Fe(II1) cations in H-ZSM-5 were studied in comparison with those of Fe( III) lattice ions in ferrisilicate H- [Fe] ZSM-5 [ 40,43,44] . The solid-state interaction of H-ZSM-5 with Fe0 or Fe30, at temperatures up to 800°C is not accompanied by the formation of isolated Fe3+ ions [43,44]. The use of FeCl, (T_,,= 309°C) enabled the registration of Fe 3f ion migration and redistribution at temperatures of 250-300°C. Fig. 11 shows the changes in the Fe3+ ESR spectra from FeCI,/ H-ZSM-5 upon different treatments. The formation of the ESR signal with g =4.27 (Fig. 11b, c) is typical of isolated Fe’+ ions being located in a strong crystal field of low symmetry. An anomalous temperature a.
b.
-
Fig. II. ESR spectra of the mixture (H-ZSM-5 + FeQ): (a) parent uncalcined mixture; (b). (c) calcined in vacuum at 300°C for I h; (d) calcined in air at 300°C for min; (e) calcined in air at 550°C for 1 h and evacuated at2OT;(f)aftertheingressofair.$ectrum(c)takenat -196”C[441.(1)~=427;(2)g=5.65;(3)g=6.25.
312
A. V Kucherm,
.4 A. Slrnkm
/ Jounral
of Molecular
Cntulws
90 (1994)
3?3-3%
dependence of the signal intensity was a peculiarity of the ESR signal, as shown in Fig. 1 lb, c. It was believed that cationic species with one elemental charge (FeO+. FeCl’ ) neutralizes one elemental lattice charge (AlO; ) [ 43,441. The evolution of the ESR spectrum Fig. I I b + Fig. 1 Id -+ Fig. 1 le upon oxidative calcination of the sample is due to the gradual substitution of anionic ligands FeCI’ +FeO+. Appearance of ESR lines with g = 6.25 and 5.65 in place of the anomalous line with g = 4.27 was treated as a further lowering of crystal field symmetry [43,44]. Reversible broadening of Fe3+ ESR lines upon the introduction of oxygen, caused by the dipole-dipole interaction with 02, demonstrated the accessibility of Fe3+ ions to gas phase molecules. Interaction of coordinatively unsaturated Fe’+ ions with such strong ligands as ammonia or pyridine resulted in an increase in field symmetry, provoking the shift of ESR lines to smaller g values (replacement of ESR lines with g = 6.25 and 5.65 by a low-temperature line with g = 4.27) [43,44]. Two UV bands (250 and 320 nm) are due to Fe3+ ions, which are located in the crystal field of very low symmetry. The increase of field symmetry upon NH, sorption resulted in the disappearance of these lines [ 441. Fe(II1) cations in H-ZSM-5 could not be reduced by NH, at temperatures up to 300°C. The severer treatment of the sample, at 550°C for 30 min, led to the complete disappearance of isolated Fe3+ ions with formation of Fe” [43,44]. Reoxidation of the reduced sample led to complete disappearance of the Fe” particles, but the ESR spectrum of isolated Fe”+ ions was not fully restored (their concentration was lower by a factor of 3-5 than that in the parent sample). This is due to the formation of Fe oxide particles on the outer surface of the zeolite crystals and the temperature of reoxidation was too low to allow migration of Fe( III) (as in the case of mixtures of Fe oxides with H-ZSM-5). However, the number of cationic sites was not reduced, and all vacant acid sites were occupied as a result of the repeated treatment with FeCI, [ 43,441. Changes in the Fe3+ ESR spectra upon interaction of Fe-H-ZSM-5 with weak ligands such as Hz0 or p-xylene confirmed the great coordinative unsaturation of the isolated Fe’+ cations [43,44]. Moreover, measurable changes in ESR spectra were registered upon Xe or hexane physisorption on FeH-ZSM-5 [ 201. Such changes in Fe3+ ion location upon the weak dispersive influence of molecules filling the zeolitic channels is indicative of ions located in a very strong crystal field of low symmetry. Formation of an intense ESR signal from p-xylene cation-radicals (g = 2.004) upon xylene sorption on the Fe-H-ZSM-5 at 20°C showed that the number of accessible redox sites (typical of pure parent H-ZSM-5) was not reduced noticeably despite the location of Fe”* cation in the zeolite channels [43,44]. The reduction of isolated Fe”+ cations by sorbed p-xylene took place at Trcd as low as 200°C [ 441. The ESR spectrum of ferrisilicate H- [ Fe] ZSM-5 resembles the spectrum of Fe’ ’ cations in Fe-H-ZSM-5 despite some differences in the g factor values. However, the properties of the Fe3+ lattice ions in ferrtsilicate (tetrahedral and distorted tetrahedral environments) differed drastically from those of Fe3 l cations in H-ZSM-5. Molecules, such as NH, for example, sorbed in ferrisilicate channels did not influence the symmetry of the crystal field (opposite to the Fe3 l cations), and Fe3 + lattice ions were resistant to the reductive treatment [43,44].
A V. Kucherov. A.A. Slinkin / Jourd
4.7. Ni2’ introduction
oj Molecular Cuta1.v.w 90 (1994) 323-354
333
into H-ZSM-5
In the case of ( NiCI? + H-ZSM-5 }, calcination of the mixture in O2 at 770 K for 6 h resulted in a complete exchange of strong acid OH groups in the zeolite for Ni*+ cations (taking into account the replacement of two protons for one Ni’+ ion). Subsequently, both reduction of the sample by hydrogen at 720 K, or ion-exchange with NH,N03 solution resulted in a full restoration of acidic OH groups in the zeolite. The rate of solid-state exchange depends strongly on the nature of the anionic fragment in the nickel compound: the process was substantially less effective upon reaction between H-ZSM-5 and NiS04, and no exchange was registered in the cases of NiO or Ni( CH,COO) 2 [ 4 1,421. 4.8. Interaction
of H-ZSM-5 with Zn arzd Ga oxides
The modification of catalytic properties of H-ZSM-5 upon the introduction of ZnO or Ga,OJ was demonstrated in many works, and in this review only a few examples are presented [45-48] where the change in physical and chemical properties of mixtures ( ZnO + H-ZSM-5 } or ( GazO, + H-ZSM-5 ) was studied in more detail by various methods (IR, XPS, X-ray diffraction, TPR, NH3 thermodesorption). Migration of Zn’+ ions from the outer surface of zeolitic crystals into H-ZSM-5 channels was assumed [45] with a concomitant decrease in acid site concentration upon thermal treatment of the (ZnO + H-ZSM-5}. Zn’+ ions introduced into H-ZSM-5 are more easily reduced by Hz, as compared with bulky ZnO phase [ 451. Formation of active sites via solid-state reaction of GazOx and H-ZSM-5 is possible [ 46481. For the intimate physical mixture of Ga,OJ and H-ZSM-5, the acidity of the zeolite, and H2 are all necessary to generate the highly active form of the catalyst. Oxidative calcination of Ga( NOII) ,/H-ZSM-5 resulted in the location of dispersed Ga,O, phase on the outer surface of zeolitic crystals and no incorporation of Ga”’ into zeolite exchange sites took place. Upon reductive treatment, migration of Ga,O into the zeolite channels occurred with formation of Ga( I) on the exchange sites [46,48].
5. Peculiarities H-[Ga]ZSM-5
of transition metal cation introduction and H-[Fe]ZSM-5
into H-[AI]ZSMd,
The properties of acid sites in high-silica zeolites can be modified through isomorphous substitution of Si’+ for different Me” .Ions in lattice positions. These acid sites in the zeolite act as traps for migrating cationic species, and replacement of Al for Ga and Fe in the H-ZSM-5 framework may change their properties. Furthermore, the substitution of Al’+ (1=5/2) for Ga3+ (Z==3/2) must change the shfs in the ESR spectra of CrSf cations, permitting insight into the structure of both the acid site and the cationic species. The solid state interaction, at 300-8OO”C, between H- [ Ga] ZSM-5 and compounds of Cu, Fe, Cr or V was studied [ 33,341 and the results were compared with the data obtained for H- [Al] ZSM-5. Both the concentration and the coordination of isolated CL?+ cations in H- [ Ga] ZSM-5 and H-ZSM-5 are essentially identical. Only a minor difference in the distortion of the CL?+
334
A. V Ktrcherov.
AA
Slurkm / Jorrrrd
of Molecular
Cutal,vsrs 90 (1994) 323-354
environment is noted for the two zeolites [ 33,341. The ability of Cu2+ cations to interact with different adsorbents is also identical for CuH-ZSM-5 and CuH- [ Ga]ZSM-5 [ 33,331. The process of Fe’+ introduction into cationic positions of H- [ Ga] ZSM-5 was identical to that examined for H-ZSM-5. The properties of isolated Fe3+ cations in both zeolites also essentially coincide [ 33,343. The process of the introduction of Cr5’ mto H- [ Gal ZSM-5 and the properties of isolated CrO’ species were identical to those found for H-ZSM-5. However, the presence of shfs in the Cr5’ ESR spectra provided additional information on the structure of the site where the cationic species were located [ 33,341. The changes in shfs due to substitution of Al”+ for Ga”+ in the ZSM-5 framework are clearly seen from the comparison of the spectra taken in the Q band (Fig. 9) [ 341. In accordance with theory, the replacement of Al (nuclear spin 1=5/2) for Ga (I= 3/2) reduced the number of shfs components from 6 to 4. At the same time, the splitting increased from 6 to 24 Oe, i.e., the interaction of the unpaired electron of CrO: with lattice Ga3+ was substantially stronger than that with lattice Al’ +. This may be caused by a slight change in the geometry of the cation site due to the difference in ionic radii between AI”+ (0.50 A) and Ga’+ (0.62 A) [ 33,341. Cationic sites in H-ZSM-5 and H- [ Ga]ZSM-5 differ noticeably in their thermostability: the strength of Ga’+ bonding in the ZSM-5 framework is lower than that of the Al’+ ion [ 33,341. The destruction of cationic posittons did not occur upon calcination at 750-800°C of H-ZSM-5 containing Cr( V), Cu(I1) or Fe(II1) cations, whereas the calcination of H[ Ga] ZSM-5 containing the three ions was accompanied by a sharp drop in the concentration of isolated cations despite the fact that the samples maintained high crystallinity as found from X-ray diffraction [ 33,341. In contrast to the case of H-ZSM-5, the interaction of H-[Ga]ZSM-5 with vanadium compounds was accompanied by the appearance of a trace V( IV) ESR signal only [ 33,341. Thus, acid sites in gallosilicate were too weak to stabilize vanadyl cations. A less thermostable matrix of H- [Fe] ZSIv-5 with weak acid sites was not able to stabilize even such cations as Cu’+ and Cr”. It was shown that no stabilization of isolated Cu’+ cations in ferrisilicate occur upon interaction of copper compounds with H-[Fe]ZSM-5 [43,44].
6. Co-introduction
of different
ions into cationic
positions
of H-ZSMS
The co-introduction of (Cu’+ +CP’ } and (Cu’+ +V4+} into zeolite H-ZSM-5 by a solid state reaction is possible under different treatment conditions [49,50]. Two different ions can be co-introduced into cationic positions inside zeolitic channels or one type of cation can replace the other. The final result depends on the rate of migration, redistribution and substitution of polyvalent ions in cation sites of high-silica zeohtes [49,50]. Different ions can be located at the same sites and migration and redistribution processes proceed concurrently [ 49,501. The ESR spectrum of the calcined mixture of H-ZSM-5 and CuCrO, was the superposition of ESR signals from isolated Cu( II) and Cr( V) ions. These ions were located in the same sites as after introduction of the individual cations from CuO or CrO, [ 12,271. The ratio of the Cu(II) and Cr( V) ESR signal intensities was estimated as L-3 [ 49,501.
A. V K~tclrerov.
A.A. Slmkbl
/ Journul
of Molecrrlctr
Catnips
90 (1994)
323-354
345
d. /
Fig. 12 ESR spectra, at 20°C. for ( [ H-ZSM-5 + 5% VzOz I + 5% CuO] evacuated at 20°C. (a) [ H-ZSM-5 + V,05] calcined in au at 8OOT for 1 h. (b) binary mixture calcined in an at 550°C for I h; (c) binary mixture calcined at 800°C for I h; (d) after reduction by H, at 400°C for 2 h [50]
Calcination of H-ZSM-5 with CrO, at 800°C resulted in the appearance of an intense ESR signal from Cr( V) ions i.e., all accessible cation sites were occupied by isolated Cr( V) ions. The subsequent interaction of the sample with CuO was accompanied by the following two processes: the CP’ ESR signal intensity dropped and, at the same time, the Cu*+ ESR signal intensity rose. Thus, Cr(V) ions in the cation positions of the H-ZSM-5 were exchanged for Cu( II) ions upon oxidative calcination of the binary mixture. The CL?+/ Cr5+ ratio was 20-30. Reduction by Hz of the sample containing co-introduced Cu( II) and
346
A C’.Kucherm,. .4 A
Slinkrrl/ Jorrmal o$Molecrrltrr Catulws 90 (I 994) 323-354
Cr( V) cations resulted in a rapid disappearance of ESR signals from both CL?+ and Cr5+ isolated cattons [ 49,501. Examination of the interaction of H-ZSM-5 with V20s and CuO was conducted under different conditions [49,50]. These particular oxides were chosen because Cu( II) is easily reduced by Hz at 400°C and V( IV) in H-ZSM-5 is not [ 27,351. The maximum number of isolated V(IV) cations in H-ZSM-5 was obtained. when a mixture of V,O, and H-ZSM-5 was calcined at 800°C (Fig. 12a). Subsequent interaction of the sample with CuO in air resulted in a gradual decrease of V( IV) signal intensity and an increase in Cu(I1) ion signal intensity (Fig. 12b, c). From the ratio of signal intensities it was concluded that at least 90% of V( IV) cations were replaced by Cu(I1) cations, i.e. the Cu( 11) ion was bound more strongly in the cationic position than the V( IV) ion [49,50]. It was supposed that the vanadium was displaced from the zeolite channels to the outer surface of crystals and oxidized to V(V). Reduction of the calcmed binary mixture by Hz at 400°C resulted in a sharp decrease in Cu( II) cation concentration but no subsequent increase in V( IV) ion concentration (Fig. 12d). This temperature was too low to promote the migration of vanadyl ions from the outer surface of the crystals. However, the rise in temperature to 800°C in Hz led to a quick remigration of V( IV) ions into vacant cationic sites and the intensity of the VJt ESR signal returned to the initial value (Fig. 12a). A new oxidative treatment of the reduced binary sample resulted in the repeated replacement of almost all V( IV) cations by the Cu( II) ions formed. The effect of cation substitution was more pronounced m the system Fe3+/H-ZSM5 + CuO [ 43,441. At least 99% of the isolated Fe” + cations were replaced by Cu’ + cations as a result of high-temperature interaction between Fe3+ -containing H-ZSM-5 and copper oxide. Also, the concentration and coordination of Cu( II) ions. introduced into the sample by a solid-state exchange, coincided completely with those of Cu( II) cations in CuH-ZSM5 [ 14,151. The solid-state interactton does not permit the exchange of Na+ cations in H-ZSM-5 for polyvalent cations, i.e. the strength of linkage for Na + ion exceeds noticeably the bonding strength for all transition metal ion studied. Formally, the strength of retention of different ions in cationic positions of H-ZSM-5 decreases in the order: Naf >> Cu’+ > Cr5+ > V”+ = F$+.
7. The influence of gas phase (0,, vacuum, solid-state exchange in high-silica zeolites
Hz, H,O) and anionic
fragments
on the
In most cases solid-state exchange in high-silica zeolites was conducted in air in the presence of water vapor. It must be noted, however, that solid-state exchange was also achieved with such insoluble compounds as CuO, Cr,O, or V20,. In many cases the reaction between H-ZSM-5 and transition metal compounds ( V,Os, CuO, CuCl,. FeCI,, MoCI,) was realized in high vacuum ( lo-” Torr), i.e. in complete absence of water. Thus, the presence of residual water is not a necessary condition for the solid-state introduction of transition metal ions into high-silica zeolites. This conclusion agrees with that for the solidstate introduction of alkaline metal ions into zeolites [ Ill.
A. V. Kuchem~.
A.A. Slrnkrrl /Journal
oj Molecular
Cotalym
90 (1994) 323-354
347
A comparative study of the solid-state exchange in air and in vacuum permitted the conclusion that in many cases (reactions between H-ZSM-5 and CuO, copper salts, V209, CrO,) the presence of O2 does not change the process of ion migration. In contrast, a study of the interaction between HM or H-ZSM-5 with Cr,O, enabled the decisive role of 0, to be demonstrated [ 27,281. The reaction did not take place in vacuum at temperatures up to 82O’C but the interaction in air resulted in migration of Cr( V) ions into cationic positions inside the zeolites. It was assumed that in this case the surface oxidation of Cr,OX ( T,,,I= 2340°C) created mobile species of higher oxides which are CrO+ [ 27,281. It is clear that treatment of H-ZSM-5 with different salts of transition metals enlarges the possibility of solid-state exchange because the presence of different anionic fragments change the mobility of cationic species. This was demonstrated in many works for Cu. Fe, Ni or Mn salts [ 12,27,42-44]. At the same time, the introduction of different anionic ligands results in the change of the local topography of the cationic site (or the symmetry of the local crystal field for isolated polycharged ion). In some cases the anionic ligand IS quite mobile (Cl- ) and it can be replaced by oxygen upon oxidative calcination [ 43,441:
./O\ _/O\./ P\ F\ P\
ä
\
y”\ /O\ / / \ ,“; P\
However, in many cases the ligand is too strongly bonded to the cation and the calcination of the samples in air at 800°C is not capable of eliminating anionic fragments (PO:-, SO;-, F-). The influence of a reductive treatment (H,, hydrocarbons) on the properties of different samples Me” + /H-ZSM-5 was studied widely. For the reactive ions which easily form Me” (Cu, Ni, Fe), reduction results in the migration of metal atoms out from the zeolitic channels with restoration of the protonic sites. However, the solid-state introduction of cations into H-ZSM-5 under reductive conditions may be realized in some cases. It was noted that reductive treatment is a necessary condition for introduction of Ga species into H-ZSM-5 (probably in the form of Ga( I) cations) [4648]. The reaction between H-ZSM-5 and V20, took place both in O2 and in H2 and vanadyl cations, withstanding redox treatment, were stabilized in cationic positions [ 27,281. For the binary sample (CuO + V?Os -t HZSM-5) reduction by Hz resulted in remigration of vanadyl cations into vacant cationic positions [ 49,501. Therefore, the change in the solid-state interaction conditions (atmosphere, type of the salt) increases the possibilities of transition metal ion introduction into cationic positions of high-silica zeolites. At the same time, the data obtained demonstrate the possibility of ion redistribution and substitution upon change in treatment conditions, i.e., upon catalytic testing of multi-component zeolite catalysts containing transition metal compounds.
8. Cations
of transition
metals
as active catalytic
sites in H-ZSM-5
Transition metal ion introduction into high-silica zeolites is a way of preparing quite simple diluted systems with regular location of isolated cations in a few discrete sites. The
ability of the H-ZSM-5 matrix to stabilize these ions as isolated cationic species in unusually low-coordination environments ( Cu2+ , Fe’ ’ ) and non-typical valence states (Cr” + , VJ + j must be noted. Such coordinatively unsaturated sites are very reactive towards a variety of ligands. Therefore, such systems are interesting catalytically. Moreover, thermal treatment of the samples as well as the introduction of different anionic ligands (F-, SOf ~, PO:- ) permits the local topography of isolated cationic sites to be changed. Thus, the comparative catalytic testing of such systems enables one to elucidate the influence of local crystal field symmetry on the Intrinsic catalytic properties of isolated transition metal ions. Despite the fact that the number of active sites in high-silica zeolites is small ( 1-I .5 wt.% of transition metal oxide may only be located in the zeolitic matrix in a catalytically active form), there are some examples of very high intrinsic redox activity in H-ZSM-5. Some recent works show that modification of pentasils by transition metal cations permits one to prepare active and stable catalysts for both total oxidation of alkanes and NO_, decomposition.
9. CuH-ZSM-5 hydrocarbons
zeolite as catalyst for both NO decomposition
and NO reduction with
Catalytic decomposition of nitric oxide was studied on a number of zeolites containing transition metal cations [ 5 I-591, but of these, only the CL?+ forms showed appreciable activity. It was noted that CuH-ZSM-5 was especially active for NO decomposition [ 521. The conversion of NO showed S-shape dependence on the exchange level; the decomposition rate was very small at low exchange levels. but increased rapidly above ca. 50% of exchange [54]. It was assumed that at least two kinds of sites exist in the ZSM-5: one IS most easily exchanged with a copper ion but is inactive [54]. The pure H-form of ZSM-5 was inactive in the stoichiometric decomposition of NO. SO, deactivates CuH-ZSM-5 for NO decomposition due to formation of CuSOA-type species [ 59,601. Nitric oxide reduction with hydrocarbons is a complex process which takes place on H forms of zeolites [57]. However, the introduction of different cations into the zeolite changed noticeably the activity of catalysts. CuH-ZSM-5 was the most active for the selective reduction of NO by ethene in the presence of oxygen at temperatures as low as 437-573 K [ 561. No detailed study of the reaction mechanism or active site structure was realized but the assistance of Cut ion in the process was assumed.
10. Complete oxidation
of alkanes
Catalytic oxidation of CH, traces with O2 is enhanced using cation-containing H-ZSM5 zeolites instead of H-ZSM-5 [61-661. The system permits comparison of the intrinsic activities of different ions (or of the same ions in different local environments j.
A. V. Kucheror. A.A. Slmkrn / Jound
of fvlolecular Carulysrs 90 (1994) 323-354
349
10.1. C/H-ZSM-5
Supported chromiacatalysts (Cr/SiO,; Cr/SiOz-Al,O,) are well-known active catalysts for complete oxidation. A comparative study of Cr/H-ZSM-5 with known physical properties is of interest. Fig. 13 shows the activity of Cr/H-ZSM-5 m CH4 burn-up at 500°C [ 61-631. Catalytic activity of the Cr/H-ZSM-5 system increased linearly with the rise of chromium content up to 1.5 wt.%, showing considerable site homogeneity. The activity of the catalyst reached a maximum, when the number of isolated Cr( V) cations, capable of changing valence state upon redox treatment, reached a maximum. Thus, the excess of chromia ( a-Cr203 phase on the outer surface of zeolitic crystals) did not contribute to the catalytic activity of the samples. At the same time, it was shown that the specific activity of Cr(V) cation in H-ZSM-5 is much higher (by a factor of 30) than that of Cr(V) ion stabilized on the surface of amorphous supports [61-631. Stabilization of isolated Cr(V) ions in coordinatively unsaturated environment inside H-ZSM-5 may be a reason for such unusually high intrinsic catalytic activity.
10.2. Ah/H-ZSM-5
Complete oxidation of CH, over Mn/H-ZSM-5 proceeds with very high activity and stability [ 63,641. Fig. 14 shows the dependence of catalytic activity upon the Mn content in H-ZSM-5 samples with different SiO,/AI,O, ratios [ 641. It was concluded that: ( 1) the maximum number of catalytic sites (i.e. isolated Mn cations) depends linearly on the concentration of vacant acid sites in H-ZSM-5; (2) intrinsic activity per Mn ion in Mn/HZSM-5 exceeds that in Mn/SiOz by a factor of 10 [ 641. sv per
I6
1
Cr
sv per
- h-’
h-’
1
-
cr5+
W
a
*Cr/Si02 I 1 Fig. 13. Space velocity (SV) (99% CH, conversion, pergramofCr(V) [621.
I 2
1
Cr,
wtX
3 500°C) for different Cr/H-ZSM-5
samples
(I
) Total; (2)
2
6
4
Fig 11. Space velouty (SV) (998 CH, conuerwn. SOO”C) for hln/H-ZSM-5 m zeolite SIO,/AI,O, = ( I ) 32; (2 J 40. ( 3) 80 [ 641
1 Fig 15. Space velouty per gram of Cu 165
(SV) (99% CH, conversion.
2
samples with different Al content
3
5Oo’C) for dlfferent G/H-ZSM-5
samples.
t I I Total, (2)
I
10.3. G/H-ZSM-5 Catalytic oxidation of CH, with O2 using CL?+ -containing H-ZSM-5 was studied in detail [ 65,661. Fig. 15 shows the activity fordifferentCu/H-ZSM-5 samples [66] .Catalytic activity of Cu/H-ZSM-5 Increased within the first 5-10 min on stream and then reached a steady state. No drop m the effectiveness of the catalysts was found at 500°C after IO h of continuous use. The calcination of the sample in air at 550°C for 15 h did not result in any decrease in catalytic activity [ 651. It is clear from Fig. 15 that stabilization ofCu( II) ions in H-ZSM-5 results in the formation
A.V. Kucherov. A.A. Slmkin /Journal
of Molecular Caralws
90 (1994) 323-354
351
of active sites. At the same time, an increase in the activity per gram of Cu (Fig. 15, curve 2) demonstrates an inhomogeneity of catalytic sites. Comparison of catalytic data with the results obtained from the study of Cu2* location and coordination showed that the high catalytic activity is determined only by the formation of the most coordinatively unsaturated square planar Cu( II) ions [ 65,661. A linear correlation between the amount of square planar Cu( II) and the activity of Cu/H-ZSM-5 demonstrated the decisive role of these [ 651. When the number of isolated Cu( II) cations in the zeolite reached a maximum (at 1.52.0 wt.%), the activity of the catalyst also reached a limit (Fig. 15, curve 1). The excess CuO on the outer surface of the zeolite did not contribute noticeably to C&H-ZSM-5 catalytic activity in CH, oxidation [ 65,661. The calcination of Cu/H-ZSM-5 at 800°C for l-2 h did not reduce noticeably the number of isolated Cu( II) ions accessible to reagents, but a considerable change in the Cu( II) ligand environment occurred. The most unsaturated square planar Cu(I1) coordination disappeared completely as a result of high-temperature treatment, and a lOOO-fold drop in catalytic activity of the samples accompanied the change in the local topography of isolated Cu(I1) in catalytic sites [65,66]. The samples Cu/HM and CuO/SiO, (which did not contain square planar Cu( II) ) also showed a much lower specific activity than Cu/H-ZSM5 [65,66]. The relation between catalytic activity and Cu2 -site local topography was also studied [ 67,681. The same samples Cu/H-ZSM-5 and Cu/HM were tested in catalytic ethane oxidation, but catalyst activities were measured at low alkane conversion, in contrast with previous results [ 65,661 where space velocity data under 99% CH4 conversion at 500°C were used to compare catalyst activities. The samples did not differ noticeably in the total number of accessible active CL?+ sites diluted in an inert support, and the specific activity per CL?+ ion was compared in each case. It was shown that a lOOO-fold change in specific activity of isolated CL?+ site may occur as a result of local topography transformation [ 67,681. Activity of the sample CI.I/H-ZSM-~~~~~ (which contained the most coordinatively unsaturated square planar Cu(I1) ions) exceeded activities of the samples containing five-coordinated CuZf ions by a factor of 30-100. Subsequent transformation of copper(I1) environment to the most symmetrical, octahedral one resulted in a further activity fall by a factor of 30. Suppression of complete oxidation was accompanied by a selectivity change: activity in high temperature, partial oxidation became measurable for the samples with more symmetrical Cu’* environments [ 67,681. l
10.4. V/H-ZSM-S The system V/H-ZSM-5 with isolated V( IV) cations was completely inactive in CH, burn-up [ 631, due to the stability of the V( IV) valence state [27,35]. Introduction of V into H-ZSM-5 zeolite does not permit the preparation of active catalyst for complete alkane oxidation.
11. Alkane oxidation
with N,O
Ethane oxidation with NzO over H-ZSM-5 and Cu/H-ZSM-5 was studied in [ 69,701. C2H, oxidation by N,O, as distinct from O?, proceeds on both H-ZSM-5 and Cu/H-ZSM-
352
A V Kucherov. A A Slinkrn /Journal
of Molecular Catalys~x 90 (1994) 323-354
5. At T>430”C, a stable 100% conversion took place on H-ZSM-5 but the decrease of temperature to 4 10-4OO”C led to a drastic drop in C,H, and NzO conversion with fast coking of zeolite. Cu/H-ZSM-5 system was more active but two fields of reaction were distinguished. Complete oxidation took place at T= 400°C and a pale-green lighting of catalyst bed accompanied this process. At 380°C both a fast coking and drastic drop in activity occurred. The oxidation of ethane with NzO was treated as a complex heterogeneoushomogeneous process (catalytic decomposition of N20 and subsequent gas-phase C,H, oxidation with participation of excited molecules and coupling products) [ 69-7 11.
12. Concluding
remarks
( 1). A solid-state reaction may be treated as a new and environmentally friendly method of high-silica zeolite modification by one or several transition metal cations (CL?+, Fe3’, V 4+, MoS+, CrS+, NiZf, Mn’+, Zn’+, Ga”+). The majority of these cations cannot be introduced into pentasil-type zeolites by conventional ion-exchange. It is possible to co-introduce two different ions into cationic positions inside the zeolitic channels or substitute one transition metal cation for the other. Different ions are located at the same sites and the migration and redistribution processes proceed concurrently. (2). The pentasil-type matrix is capable of stabilizing isolated cations in both non-typical valence states and unusually low coordination environments (i.e., in local crystal fields of low symmetry). Such coordinatively unsaturated transition metal ions have very high reactivity. (3). The possibility of solid-state ion introduction into high-silica zeolites is determined by the presence of acid sites, i.e., strong acid sites serve as powerful traps for migrating cationic species. The treatment of H-ZSM-5 both with different salts of transition metals or in the presence of gas phase molecules enhances the possibility of solid-state exchange due to the change in cationic species mobility. At the same time, the presence of different anionic ligands results in the change of the local topography of cationic site. Therefore, atomic-scale engineering of catalytic functions of isolated active sites in confined environments of zeolitic channels becomes possible. (4). Pentasil-based catalysts with discrete isolated redox sites permit a correct study of structure/activity relationship for isolated active sites of known local topography. (5). Despite the fact that the concentration of active sites in high-silica zeolites cannot be high, there are some examples of very high specific redox activity of cationic sites stabilized by H-ZSM-5 matrix. Some recent works show that modification of pentasils by transition metal cations permits preparation of promising catalysts for both complete oxidation of alkanes and NO_, decomposition.
References [ 11 D.W. Breck. Zeolite Molecular &eves Structure, Chemistry and Use, Wiley, New York, 1974 [21 R.M. Barrer.Zeolites and Clay Mineralsas Sorbentsand Molecular &eves, Academic Press, London, New York. 1978
A. V. Kucherov. A.A. Shkm
/ Journal of Molecular Catalysm 90 (1994) 323-354
353
[3] J.A. Rabo, M.L. Poutsma and G.W. Skeels, in J. Hightower (Ed.), Proc. 5th Int. Congr. Catal., August 2026, 1972, Miami Beach, FL, North-Holland, New York, 1973, p. 1353-1361. [4] J A. Rabo, Salt Occlusion in Zeolite Crystals, in J.A. Rabo (Ed.), Zeolite Chenustry and Catalysis, ACS Monograph 171, American Chemical Society, Washington, DC, 1976, p. 332-349. [5] J.A. Rabo and P.H. Kasai, Progr. Solid State Chem.9 ( 1975) 1. [6] A. Clearfield. C.H. Saldamaga and R.C Buckley, Proc 3rd Int. Conf Molec. Sieves, Sept. 3-7, 1973, Zurich. Switzerland, Recent Progress Reports, J.B. Uytterhoeven (Ed.), Univ. Leuven Press, 1973, Paper 130, p. 241-245. [7] R G Herman, J.H. Lunsford, H. Beyer, P.A. Jacobs and J B Uytterhoeven, J. Phys. Chem., 79 ( 1975) 2388 [S] A.A. Shnkin, G.V Antoshin. M I Loktev et al., Kinet. Katal., 19 (1978) 754 [9] H.K. Beyer and P.A. Jacobs, J.R. Katzer (Ed.), Molecular Sieves II. ACS Symp. Ser., No. 40, 1977, p. 493503. [lo] H. Beyer. P.A Jacobs and J.B. Uytterhoeven, J. Chem. Sot., Faraday Trans. 1.72 ( 1976) 674. [ 111 H.G Karge and H.K. Beyer, Stud. Surf. Scl. Catal.. 69 ( 1991) 43. [ 121 A.V. Kucherov and A.A. Slinkin, Zeolites, 6 ( 1986) 175 [ 131 A.V Kucherov and A.A. Slinkin, Kmet. Katal ,27( 1986) 671. [ 141 A.V. Kucherov. A A. Slinkin, D.A. Kondratyev. T.N. Bondarenko, A M. Rubinstein and Kh.M. Minachev, Zeolites, 5 ( 1985) 320. [ 151 A V. Kucherov, A.A. Slinkin, D.A Kondratyev. T.N. Bondarenko, A.M. Rubinstein and Kh.M. Minachev, Kinet. Katal.. 26 (1985) 409 [ 161 M.W. Anderson and L Kevan, J. Phys. Chem., 91 (1987) 4174 [ 17) Y Sendoda and Y. Ono, Zeolites, 6 ( 1986) 209 [ 181 A.A. Slinkin, A.V Kucherov, N D. Chuvylkin, V A. Korsunov. A.L Khachko and S.B. Nikishenko. J. Chem. Sot., Faraday Trans. 1,85 ( 1989) 3233. [ 191 A.A. Slinkin, A.V. Kucherov, N.D. Chuvylkm, V.A Korsunov, A L. Kliachko and S.B. Nikishenko. Kmet. Katal., 31 (1990) 698. [20] A.V Kucherov and A.A. Slinkin, J Phys. Chem, 93 (1989) 864 [21] A V Kucherov and A.A Slinkin. Kinet. Katal., 30 ( 1989) 497. [22] Lm Zhenyi. Zhang Wangjin and Yu Qin. Proc 7th Int. Zeolite Conf., Tokyo, 1986, Kodansha-Elsevler, Tokyo-Amsterdam, 1986, p 1091. [23] A.V. Kucherov, unpubhshed data [24] M. Narayana and L. Kevan, J Chem. Sot., Faraday Trans. 1.82 ( 1986) 213 [25] C.E. Sass and L. Kevan, J. Phys. Chem.. 92 ( 1988) 5192 [26] J.S. Yu and L. Kevan, J Phys. Chem., 94 ( 1990) 7620 [27] A.V. Kucherov and A.A Slinkin, Zeohtes, 7 (1987) 38 [28] A.V. Kucherov and A.A Slinkin. Kinet. Katal., 27 (1986) 678. [29] A.A. Slinkm. A.V. Kucherov, S.S. Goqashenko, E G Aleshin and K I. SlovetskaJa, Zeolites. 10 ( 1990) 111. [ 301 A.A. Slinkin, A.V. Kucherov, S.S. Gojashenko, E.G. Aleshm and K 1. Slovetskaja, Kinet. Katal (30 ( 1989) 184. [31] J.-F Hemidy, F. Delavennat and D. Comet, J. Chim Phys. Phys.-Chim. Biol , 70 (1973) 1716 [32] J.-F Hemidy and D. Comet, J. Chim. Phys., Phys.-Chim Blol .71 ( 1974) 739 [33] A.V. Kucherov. A.A. Shnkin, H.K BeyerandG. Borbely, J. Chem Sot., Faraday Trans 1.85 ( 1989) 2737. [34] A V. Kucherov, A A. Slinkm, H.K Beyer and G. Borbely, Kinet. Katal., 30 ( 1989) 429. [35] A A. Slinkm, A.V Kucherov and S.B Nikishenko, Kinet. Katal.. 31 ( 1990) 692. [36] A.V. Kucherov and A.A Slinkin. Zeolites, 7 (1987) 583. [37] C.E. Sass, X Chen and L. Kevan. J. Chem Sot., Faraday Trans., 86 ( 1990) 189. [38] V.K. Sharma, A. Wokaun and A. Baiker, J Phys. Chem ,90 ( 1986) 2715. [ 391 C. Marchal, J. Thoret, M. Gruiaet al, in M.L. Occelly (Ed.), Fluid Catalytic Cracking, Vol. II. Concepts in Catalysis Design, ACS Symp. Ser., 1990, p. 452. [40 1B. Wichterlova, S. Beran. S. Bednxova. K Nedomova, L. Dudlkova and P. .lux, Stud Surf. Sci. Catal , 37 (1988) 199 [41] S. Beran, B. Wichterlovaand H.G. Karge, J Chem. Sot , Faraday Trans. I, 86 ( 1990) 3033. 1421 B. Wlchterlova. S Beran, L. Kubelkova et al.. Stud. Surf. Sci. Catal ,46 ( 1989) 347
354
A.V. Kucheror’. A.A. Sldm
/Journal
of Molecular Caralws
90 (1994) 323-35-J
[431 A.V. Kucherov and A.A Slinkin, Kinet. Katal.. 28 ( 1987) 1199. [W] A V. Kucherov and A.A Slinkin, Zeolites. 8 ( 1988) I10 [45 ] Y. Yang, X. Guo, M. Deng et al., Stud. Surf. Sci. Catal ,46 ( 1989) 849. [46] G.L Price and V. Kanazirev. J Catal.. 126 (1990) 267 1471 P. Menaudeau and C. Naccache. Stud Surf. SCI. Catal., 69 ( 1991) 767. [48 1J.F Joly, H. Ajot. E Merlen et al., Appl. Catal. A, 79 ( 1991) 249. 149 I A V. Kucherov and A.A. Slinkin. Kmet. Katal., 27 ( 1986) 909 [50] A.V Kucherov and A A Slinkin, Zeohtes, 7 ( 1987) J3 15 I ] M. lwamoto. S. Yokoo, K Sakai and S. Kngawa. J. Chem. Sot., Faraday Trans. I. 77 ( 198 I ) 1629. [52 I M. Iwamoto, H. Furukawaand S Kagawa, New Development in Zeolite Science and Technology, Kodanshn, Tokyo. 1986, p. 943. [53] M lwamoto, H. Furukawa, Y. Mme et al.. J. Chem. Sot.. Chem. Commun, ( 1986) 1272 [54] M. Iwamoto, H. Yahiro, Y Mine and S. Kagawa. Chem. Lett., ( 1989) 213 [55] Y Li and W.K. Hall, J. Phys. Chem.. 94 ( 1990) 6145. [56] S. Sato, Y Yu-u, H Yahiro, N Mizuno and M. lwamoto, Appl Catal.. 70 ( 1991) LI 1571 H. Hamada, Y. Kmtarcht. M. Sasaki et al., Appl. Catal.. 64 ( 1990) LI [58] M. Iwamoto, H. Yahiro. S. Shundo et al, Shokubai (Catalyst). 32 ( 1990) 430. [59] M Iwamoto. H Yahtro. S Shundo et al , Appl. Catal.. 69 ( 1991) L15 1601 H. Hamada. N. Matsubayashi, H. Shimada et al., Catal. Lett ,5 ( 1990) I89 [ 61 J A A. Slinkin, A.V Kucherov. S.S. Goryashenko and K I. Slovetskaya Proc. All-Umon Conf Use of zeolites m catalysis, Nauka, hqoscow, 1989, p. I9 I (in Russran). [62 I A.A. Shnkm. A V Kucherov, S.S. Goriashenko. E G. Aleshm and K.I. Slovetskaya. Proc 8th Sovtet-French Seminar Catal., Novostbirsk, 1990. p. 153. [63] A.A. Shnkm. A V Kucherov. S S. Goriashenko. E.G Aleshin and K.1 Slovetskaya, Proc 9th SovtetJapanese Seminar Catal.. Novostbirsk, 1990, p. 224. [64] S S Goriashenko, K.I. Slovetskaya, M.A Alimov and A A. Slinkin. Kinet Katal., 33 ( 1992) 350. [65 ] A.V Kucherov, A A. Slmkin, S.S. Goryashenko and K.I. Slovetskaya, J. Catal, II8 ( 1989) 459. [66] S.S Gonashenko. K.I. Slovetskaya, A V Kucherov and A A. Shnkin, Kmet. KataJ,30 ( 1989) 249 [67] A.V Kucherov, T.N. Kucherova and A A. Slmkin. Catal. Len., IO (1991) 289. 1681 A.V Kucherov, T N Kucherova and A.A. Shnkin. Kinet Katal ,33 ( 1992) 618. 1691 A.V. Kucherov. TN Kucherova and A A Slinkm, Recent Res. Rep lnt Symp. Zeolite Chem Catal.. Prague, Sept. 8-13, 1991. p. 69. [ 701 A V Kucherov. T N Kucherova and A A. Shnkin, Kinet. Katal., 33 ( 1992) 877. [7l I A V Kucherov. S-S. Goryashenko. T.N Kucherova, K I Slovetskaya and A.A Shnkin, Kmet Katal., 34 (1993) 1089