Formation and Catalytic Properties of Nickel Metal Particles Supported on Zeolite

Formation and Catalytic Properties of Nickel Metal Particles Supported on Zeolite

B. Imelik et al, (Editors), Catalysis by Zeolites © 1980 Elsevier Scientific Publishing Company, Amsterdam - Printed in The Netherlands 235 FORMATI...

408KB Sizes 0 Downloads 21 Views

B. Imelik et al, (Editors), Catalysis by Zeolites

© 1980 Elsevier Scientific Publishing Company, Amsterdam - Printed in The Netherlands


FORMATION AND CATALYTIC PROPERTIES OF NICKEL r1ETAL PARTICLES SUPPORTED ON ZEOLITE. D. DELAFOSSE Laboratoire de Chimie des Solides - ER 133 CNRS, Universite P. et 11. Curie - T. 55-54, 4, place Jussieu 75230 - Paris Cedex 05

1 INTRODUCTION Dispersed nickel supported on various oxides was used as catalysts specially in the hydrogenation, dehydrogenation, hydrogenolysis reactions (1-5). The importance of support is now well known. It has multiple effects -on the formation of metal particles, it is available to stabilize the nickel in a dispersed state and inhibit its sintering- on the catalytic reaction, it is available to provide acidic sites and to playa role in dual functional catalysis. (6,7) The peculiar properties of zeolites are so interesting that they are used like support of a lot of metal in reforming catalysis. Indeed zeolites contain cavities or tubes with defined size within metal particles can be entrapped and their strong acidic sites are available for dual functional catalysis (8,9). Taking in account the interaction between metal and support, zeolites coul? be able to provide a synergetic effect for some reactions catalysed by metals. Few years ago, several stu~ dies concerning the formation (15) and catalytic properties of nickel metal in type A,X,Y and Z zeolites were reported (10-14). In these works, reduced nickel has been found to sinter easily. As a result, large metal particles are formed at the external surface of 2eolite crystals. (14). It is unlikely that for this dispersion the zeolite support may effect drastically the catalytic properties of the metal. Attempts have been reported to prevent sintering of nicr.el towards the external zeolite surface and to precise (10, 16, 17, 18, 19) the various factors which interact on the formation and stability of nickel particles inside the framework. Indeed two main problems influence the formation of small particles of supported nickel the first is the poor reducibility of nickel cation; the second is the mobility of nickel particles with heating and their ability to form large cristallites. In order to study the catalytic action of metal catalysts, it is necessary to obtain 6 high dispersed state.(20, 21,22). The extensive body of literature of catalytic properties of metal zeolites has been reviewed in detail by Minachev and Isako (23) ; the mechanism of metallic cluster formation in zeolites has been reviewed recently by J.B. Uytterhoeven (24). Our aim is to present now a report of recent results on the formation, stability and catalytic properties of nickel metal in zeolite framework, and to show how it is possible to obtain entrapped nickel clusters within zeolite cavities and obtain the way to a comparison (25) of properties of these aggregates with the large and heterodispersed particles ones.


2 PREPARATION OF NICKEL ZEOLITES 2.1 Introduction of Nickel compounds Impregnation, Ion exchange adsorption from the gaseous phase, adsorption of metal vapour can be used to introduce active compounds into zeolite cavities or on their crystal surface. Derouane and al (26) have reported to obtain Mo, Re, Ru, Ni in Y zeolites from metal carbonyl adsorption and thermal decomposition. In the case of Mo the oxide was obtained. A second method is the interaction of bis TI allyl nickel with OH groups of the support (27). It was available to obtain small highly and homogeneous dispersed Nio particles on silica after hydrogen reduction at 400°C. However this method has not been yet used on zeolite support. It is also possible to obtain nickel particles from decomposition and reduction of nickel hexamine complex. This complex is introduced by exchange from SCN compound in liquid ammonia into NaX zeolite free of water (28). In the general case, Ni 2+ is introduced by ion exchange from nitrate or chloride (29). The size and dispersion of metal particles after reduction depends on the reducibility of Ni 2+ ions, the reducing agent and the nature of cations other than Nj2+. 2.2

Reducibility Many factors are involved in the nickel cations reducibility. First of all is the initial location of Ni 2+. In faujasite zeolite Ni 2+ ions are distributed in the SI, SII, SI', SII' sites. It is obvious that the more reducible are located in sites II. This location determined by Xray method depends on Si/Al ratio, exchange degree, pretreatment conditions, and nature of other cation present with Ni 2+ (30) and Table 1.


TABLE 1 Initial location of cations (30, 31, 32) after pretreatment under vacuum at 773°K. Samples Ni 13Na30Y Ni 20Na 16Y Ni 8Na64X Ni 31Na24X



11 N;2+

2 Ni 2+

25 Na+

12 Ni 2+

4 Ni 2+

3 Ni 2+ + 16 Na+

5 Ni 2+ 12 Ni 2+

20 N
30 Na+ 6 Ni 2+ + 24 Na+ 2 Ni 2+ + 24 Na+

9 La 3+

3 Ni 2+ + 14 Na+

3 ee 3+ 1 Ce 3+

12 Ni 2+ + 18 Na+ 18 Na+ + 2 Ni 2+

8 Ni 2+ + 2 Ce 3+

18 Na+ + S Ni 2+

12 Ca 2+ 2+ 11 Ni

Nil0Ca20Na24H12x Ni14LalSNaX Ni 24Ce 6Na X Ni8CeSNa16H40X Ni21Ce6,SNa20H4,5X

12 Ni 2+ 6 Ni 2+ + 4 Ce 3+ 4,S Ce 3+ + 6 Ni 2+

Initial location of cations after pretreatment under vacuum at 623°K. 3 Ce 3+ + 5 Ni 2+ 2 Ni 2+ + 18 Na+ 3 Ce 2+ + 1 Ni 2+ NiSCeSNa16H40X 18 Na+ + S Ni 2+ Ce 3+ + 10 Ni 2+ 5,S Ce 3+ Ni21Ce6,5Na20H4,SX (II ') 4 Ni 2+ + 13 H 20 Init{al location of cations after pretreatment with CO at 373°K.

The reducibility of i~i2+ ions is too depending on the composition a nd redox properties ~f zeolite framework. It is known now, the Ni 2+ is more reducible in framework A than X and Y zeolite. (17). In NiNa X, the sites are attacked simultaneously by molecular hydrogen with an activation energy of 28 Kcal/mole. On the contrary, in Y zeolite, the Ni 2+ in sites I are very difficult to reduce (31) with an activation energy of 40 Kcal/mole. As the concentration of proton of zeolite framework increases, the reducibility of Ni 2+ decreases (30, 32). According to Rickert (10) in zeolite media, the redox equilibrium interacts always Z - 0 ......... Ni 2+ + H _ 2 ZOH+ + Ni o (1) Z - 0..........


and is displaced towards the left by a high proton concentration. - the presence of other cations modify drastically the behaviour of Ni 2+ towards hydrogen. From a C1I20Ni14H12Na24X sample reduced in mild conditions, it was obtained N4+ species according to H \1'0 N,2+..:.:z. 2 "1' N'+ (2) (33) "'11+ 1 1 1 ' __ In presence of La 3+ which hinders the dehydration, it is assumed the formation of 3+_ La 0 - Ni 2+ entities into supercages with extraframework oxygen (34), the presence of 3+ Ce modifies the redox properties of zeolite. By studying the adsorption of TCNE and Perylen on samples NiCeNaX, it was shown that the reducing properties are increased with the


ce 00

Table II Ni 2+ Reducibility Dehydrated 773 K Dehydrated 773 L Dehydrated 623 K Dehydrated 773 K K + CO 373 K 0 Reduced 623 I( Reduced 623 K 02773 K.Reduced 623 K 2773Reduced Reduced 573 K 273 K , time 18 h HO flow , time 6 h H flow , time 18 h H , time 18 h P=50 Torrs, time 25 h H2f low 2 2flow ° ° DA u ll DA u DA u DA DA u u 1 10 90 % 60 0,3 Dehydrated 773 K



Ni 8Na64X

10 % 15 80 % 150 20 % 15 idem

Ni 31Na24X


Ni 13Na27Y


Ni 20Na 16Y





Ni 14La15NallX



Ni lDH62Na4X










85 % 7 15 % 30




85 15 92 8







• u

Ni ° =-Ni t ot



II •

%"7 % 30 % 30 % 8 24

53 % 25 40 % 7

0,95 0,5 0,2 0,95

•• Reduced during five days

81 19 80 20

%7 % 30 % 20 % 15 7

13 % 30 87 % 7


0,6 0,9 0,36

45 % 25 55 % 7 40 % 150 7 38 % 25 55 % 7






ratio Ce 3+/Ni 2+ and the oxidating ones are increased with the Lewis sites number (35). Thus with a ratio Ce 3+/Ni 2+ ~ 1 and a peculiar location of Ce3+ (in sites I '), it was possible to obtain a complete reduction of Ni 2+ ions into Nio. Small quantities of pte or PdQ in the lattice increases the Ni 2+ reducibility; the higher is the PdQ , pte concentration, the greater will be the reducibility (Table II). Ni 2+ reducibility depends on the nature of reducing agent. The influence of K, H2, NH 3 on the degree of reduction and on the dispersion of Ni ° has been studied in NiY zeolite (36) The results are summarized in Table III. TABLE III Sampl e


Condition of Dehydra tion 673 under vacuum flow He 1h 823K flow He 1hh823~

Agent K H2 NH 3

Reduction Temp. Time oK /Hours 403 723 823


Size degree of red. 0,1 0,3 0,8

DA° 99% 20% 62% 20%

< ~ ~ ~

20 60 42 10

A. new and very interesting method to increase the reducibility of Ni 2+ encaged into zeolite is to use atomic hydrogen (38, 39) at very low temperature (273°K). This method unables a better accessibility of cation towards hydrogen, as atomic hydrogen is able to enter into sodalite cavities. Besides, since the reduction temperature is low, the mobility of cations and metallic atoms is inhibited. Thus it was obtained from Ni10Ca20Na24H12X sample a complete reduction of Ni 2+ into Nio (Table II) (38). 3 NICKEL METAL DISPERSION In general case nickel particles obtained by reduction with molecular hydrogen of Ni 2+ ions located in faujasite or mordenit~ zeolite are inhomogeneous (II, 13, 14, 15). Romanovski (16) was the first to provide evidence for the existence of a bidispersed metal size particles distribution in Y zeolite. Quantitative method has been developed to determine the amount of Ni ° outside and inside the Y zeolite (39). Recently Jacobs and al (40) have evidenced using FMR, TPR and TPO, a bidispersion. For NiNaX samples, it was found also an heterodispersion with small particles into the framework and the larger ones in the outside (Table II). However Guilleux and al (19) had reported that it is possible to obtain homogeneous and small nickel particles inside the framework in two cases. In the presence of very small o amounts of pte or PdQ, it was obtained cristallites of 25 A inside the framework by breakdown of a window between two supercages (41). In the presence of Ce 3+, it was obtained homogeneous particles size the value of which depends on the Ce 3+/Ni 2+ ratio and on the acidic sites concentration (Table II) (35). In all cases, the best results are obtained when reduction with molecular hydrogen was performed in dynamical conditions. Che and al (38) have obtained an homodispersion of Nio with a size of 10 A by reducing at 273K NilOca'20H12Na24x with atomic hydrogen. This sample was pretreated by CO at 373 K in order to obtain all the Ni 2+ in supercages.


Thus it was possible under sevrreand peculiar conditions to obtain homogeneous and small nickel particles. Their size depends also on the nature of other elements introduced with nickel ° and inhibits 3+ Ce located in site I' is available to stabilize very small aggregates (7 A) their migration with atomic hydrogen reduction, the size was restricted to the dimension of ° and in the presence of small amount of Platinum to that of two superone supercage (10 A) ° cages (25 A). 4

STABILITY OF NICKEL PARTICLES FMR study (40) shows that Nio formed outside the framework corresponds to large particles They do not strongly interact with the support and exhibit some shape anisotropy. These particles are stable under vacuum and they are reoxidized by molecular oxygen above 573°K into ~ickel oxide. Nio encaged in the supercages are assumed to interact strongly with the support and a charge transfer between the Nickel particles and the zeolite framework may occur. In agreement with other searchers (10, 39, 40) we found these particles are reoxidized by molecular or activated zxygen or NO to ~i2+ cations which go back into their cristallographic sites. 2 Z-OH + Nio ~02 Ni 2+ + H (3) 20 The stability of these small particles under vacuum depends on the ~eduction conditions. So for NiCeX reduced by molecular hydrogen, the nickel particles of 7 A (Table II) are very sensitive to the presence of protons in the lattice and they are reoxidized at low temperature ; the equilibrium (1) is displaced towards the left. Some various experiments were ° Thus from the L performed to determine the state of ~ggregates < 10 A. I I I emission spectrum both for small nickel particles (10 A) in zeolite prepared by atomic hydrogen reduction and for bulk nickel, the 3d filled distribution has been obtained. But while for the bulk nickel a structureless band is observed, structures are clearly apparent for small particles. The comparison between the LIl r spectra of NiO and bulk Ni evidences that the aggregates are pure Nio. These results can be explained in terms of a decrease of interatomic interactions between 3d electrons and suggests an infinite to finite system behaviour transition (42). 5

CATALYTIC PROPERTIES Most of the studies on the nickel loaded zeolite were relative to very large metallic particles outside the framework of faujasite or mordenite zeolite in the presence of unreduced Ni 2+ cations (II, 13, 14, 43, 44). Indeed the data reported are very dispersed and controversed. But in these conditions metal containing zeolites can act as monofunctional catalysts and the zeolite can be considered as inert support for the active metals (23). The reactions catalysed by nickel zeolite should be clarified in two groups: according to Boudart (45) there are structure insensitive reactions as olefin hydrogenation or paraffin dehydrogenation and sensitive ones as hydrogenolysis, methanation ... 5.1

Hydrogenation and dehydrogenation reactions On NiNaY obtained after reduction (12, 46) the degree of conversion in n octane hydrogenation attained 81 %whereas isolefin and benzene were practically unreactive because the latter cannot penetrate the zeolite cavity where most of the nickel was located. On the contrary, NiNaA reduced at high temperature (large metallic crystallites outside) hydrogena-


ted benzene with 66 % at 180°C. There is no evidence indicating that the behaviour of metallic species outside zeolite crystals differs from that of the metal supported on other carriers. Nevertheless in a recent study on the benzene hydrogenation on various NiX (47). it was observed that the NioX activity towards benzene hydrogenation is lower than that observed on Nio/Si0 2 in the same conditions. It was available to assume benzene hydrogenation is a metallic structure i~sensitive reaction but these results lead us to consider that a portion of Nio was not accessible to benzene molecules by steric hindrance or metal carrier i nterac ti on. Ce 3+ increases the hydrogenation activity (25). This result can be compared with the data reported for PtCeY (21. 48) and explained by Ce 3+/Nio interaction. or by the additive effect of the electrostatic fiels of the Ce 3+ cations (21). lone and al (44) assume the pecularities of the behaviour of the nickel portion fixed in the zeolite cavities are due to the unstable valence state of metal atoms in small clusters. This unstability depends on the degree of interaction between the atoms of metallic clusters and electron acceptor centers of the zeolite framework. These clusters are active for n hexene hydrogenation and inactive for benzene hydrogenation. In the first case. an electron acceptor center is blocked by a hexene molecule according to Ni 2+Z + H Nio + H 2 2 The authors assume a weaker benzene ability to electron-donor.


- \z-} -


Hydrogenolysis and other reactions Nickel loaded zeolites were active in the reaction of hydrogenolysis of ethane and n hexane (15). The specific activity of reduced nickel decreases with the increasing acidity of the support in order Si02 > A1 203 > Silica-Alumina> Y faujasites but the activities remain constant in the faujasite acidity rang~ from NaY to MgY. There are two types of sites of nickel supported by faujasite : type I the most active accounts for 35 % of total surface and is responsible for n hexane hydrogenolysis. type II is responsible for n hexane isomerisation hydrocracking and benzene hydrogenation. Hydrogen sulfide poisons type II sites with one sulfur for every two nickel atoms. 5.3 The bifunctional action of nickel zeolite catalysts in the conversion of toluene were studied (43). The high activity towards the disproportionation of toluene of the reduced nickel zeolite catalysts is attributed to the fact more than 30 %of the nickel ions remain unreduced thereby being the main source of proton acidity. In the presence of hydrogen, the metallic nickel assists the elimination of some unsaturated hydrocarbons which offensive would polymerize and block the active sites. Nickel and cobalt Y zeolite and mordenite proved to be efficient in the conversion of n hexane with water vapor (49). The 3,24 % Ni Zeolon and 0,76 % NiNaY catalysts were more active than the commercial Ni/A1 203 which contains more nickel ( 15 %). Nickel cristallites located on the external surface of the Y type zeolite were investigated for the methanation reaction (50), their turnover number for this reaction was found more than an order of magnitude less than that observed for Ni/A1 203• Ruthenium clusters


inside the supercages were found very active for methanation but exhibited a marked decline in activity during the reaction, attributed to the build-up of excess carbon which results from the di'ssociation of CO. The addition of excess nickel to Ruthenium (1:7 atomic ratio) brought about a stabilization in activity. NiRu bimetallic clusters were not very dispersed (2.21 %, but the presence of Ni is believed to moderate the activity of ruthenium providing .more sites for dissociative chemisorption of H2 (51). There are no systematical studies on the behaviour of nickel zeolites towards poisons like sulfur, ammonia or coke. No relation was yet established between the size of nickel particles and the poisoning resistance of catalysts. 6

CONCLUSION The conclusions drawn from the works reported up to now do not take into account the fact that the studied samples are heterodispersed and incompletely reduced. Nevertheless it is obvious that if the large Nio particles formed outside the framework do not interact with the support, the smallest ones inside the zeolite cavities interact strongly with it and are role to modify the catalytic behaviour of the whole system. For studying the role played by Nickel aggregates entrapped within the supercages towards the hydrogenation or hydrogenolysis reactions and their poisoning resistance, it was necessary to start from completely reduced and homodispersed samples. In the first part of this report, we have seen it was now possible to obtain small Nio particles entrapped in supercages of X zeolites (Table II). So it was interesting to study the catalytic activity of these well defined samples towards the two types of reactions -insensitive or sensitiveand to evidence the interactions of Nio clusters with another cation like Ce 3+ or with Lewis sites. J.F. Tempere reports in the proceedings of this congress (25) the data we have obtained for butane hydrogenolysis. I would only say that the Ni o aggregates with a size lower ° are quasi inactive towards these two reactions. Assumptions are made to explain than 10 A these results. Probably the ability to obtain small nickel particles into zeolite, involves a high interaction between the electron acceptor sites and the metal. Conversely to the case of Platinum in Pt Y, the unstability of small Nickel particles towards the ambient atmosphere makes difficult to evidence these interactions by physical methods like ESCA or EXAFS. It seems difficult now to use for catalytic application small nickel particles but they are available to form alloys and bimetallic aggregates with peculiar properties. Besides, it would be very important to study their selectivity and thioresistance for some catalytic reactions. REFERENCES 1 R.Z.C. Van Meerten and J.~.E. Coenen, J. Cata1., 37 (1975) 37-43. 2 R.Z.C. Van Meerten and J.W.E. Coenen, J. Cata1., 46 (1977) 13. 3 G. Da1mai-Ime1ik and J. Massardier, Proc. 6th Int. Congress on Catalysis (1976) Chem. Soc London Vol 1 pp. 90. 4 G.A. Martin, J. Catal., 60 (1979) 345-355. 5 A. Frennet, L. D~gols, G. Lienard and F. Crucq, J. Catal., 55 (1973) 150. 6 M.A. Vannice, R.t. Garten, J. Catal., 56 (1979) 236-248. 7 T.E. Whyte, Catal. Rev., g (1973) p. 117. 8 J.A. Rabo, P.E. Pickert, D.N. Stamires and J.E. Boyle, Actes 2eme Congo Int. Catalyse Paris 1960, Vol. 2 (1961) p. 2055. 9 P.B. Weisz, V.J. Frilette, R.M.Maatman and E.B. Hower, J. Catal., 1 (1962) 307. 10 L. Rieckert, Ber. Bunsenges Phys. Chem., 73 (1969) 331.


11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50

V. Penchev, N. Davidova, V. Kanazirev, H. Minchev and Y. Neinska, Adv. Chem. Ser., 121 (1973) 461-468. N.V. Borunova, L. Kh. Freidlin et al., Zeolity ikh. Sintez, Svoistva i primenenie Nauka M-L (1965) p. 380. Kh. M. Minachev, V.I. Garanin and T.A. Novrusov, Izv. Akad. Nauk. SSSR, Ser Khim, (1973) 330. P. J.R. Chutoransky and W.L. Kranich, J. Catal., 21 (1971) 1-11. J.T. Richardson, J. Catal., 21 (1971) 122-129. W. Romanwski, Roczniki Chemie Ann. Soc. Chim. Polonium, 45 (1971) 427. T.A. Egerton and J.C. Vickerman, J. Chem. Soc. Faraday Trans. I, 69 (1973) 39-49. A.C. Herd and C.G. Pope, J. Chem. Soc. Faraday Trans. 1,69 (1973) 833-83(L M.F. Guil1eux, D. Delafosse and G.A. Martin, J.A. Dalmon, J. Chern. Soc. Faraday Trans. I, 175 (1979) 165-171. R.A. Dalla Betta, M. Boudart, Proc. 5th Int. Congress on Catalysis 1972, Miami Beach 1 (1973) pp. 329. P. Gallezot, Catal. Rev. Ser. Eng., 201 (1979) 121-154. P. Gelin, Y. Ben Taarit and C. Naccache, J. Catal., 59 (1979) 357-364. Kh. M. Minachev and Va. I. Isakov, In Zeolite Chemistry and Catalysis (J.A. Rabo ed) ACS Monograph 171 (1976) p. 552. J.B. Uytterhoeven, Acta Phys. et Chemica Szeged Hungary, XXIV 1-2, pp 53-69. J.F. Tempere, this review. E.G. Derouane, J.B. Nagy and J.C. Vedrine, J. Catal., 46 (1977) 434-437. U.N. Ermakov, B.N. Kuznetsov, Kinet. Catal., 13 (1972) 1355 ; DJkl. Akad. Nauk. SSSR, 207 (1972) 644. J. Fournier, Ch. De la Calle, M. Briend and r~.F. Guilleux (no published). R.A. Schoonheijdt, L.J. Vandamme, P.A. Jacobs and J.B. Uytterhoeven, J. Catal., 43 (1976) 292. M. Briend-Faure, M.F. Guil'eux, J. Jeanjean, D. Delafosse and G. Kjega-Marriadassou, M. Bureau-Tardy, Acta Phys. et Chem. Szeged Hungaria XXIV 1-2 1973 pp. 99-106. M. Briend-Faure, J. Jeanjean, M. Kermarec and D. Delafosse, J.C.S. Faraday I, 74 (197B) 1598. S. Bhatia, J.F. Mathews and N.N. Bakhshu, Acta Phys. et Chem. Szeged Hungaria XXIX 1-2, 1978, pp. 83-88. D. Olivier, L. Bonneviot, M. Richard and M. Che, Proc. (NATO) 2nd Int. Symp. on magnetic Resonance in Colloide and Interface Science, Menton, 1979, in press. M. Briend-Faure, J. Jeanjean, D. Delafosse and P. Gallezot, J. Phys. Chem. (1980)(in press) . S. Kjemel, M.F. Guil1eux, J.F. Tempere and D. Delafosse (to be published). F. Schmidt, H. Kacirek and W. Gunsser (to be published). M. Che, M. Richard and D. Olivier, J.C.S. Faraday I, 1980 (in press~. D. Olivier, M. Richard, L. Bonneviot and M. Che, Actes 32e Reunion Internationale de Chimie Physique, Lyon Septembre 1979, in press. P.A. Jacobs, J.P. Linart, H. Nijs, J.B. Uytterhoeven and H.K. Beyer, J.C.S. Faraday I 73 (1977) 1745-1754. P.A. Jacobs, H. Nijs, J. Verdonck and E.G. Derouane, J.P. Gibson, A.J. Simoens, J.C.S. Faraday I, 75 (1979) 1196-1206. J. Verdonck, P.A.Jacobs and J.V. Uytterhoeven (unpublished results). D. Fargues, F. Vergand, E. Belin, Ch. Bonnelle and D. Olivier, L. Bonneviot, M. Che, 2nd Int. Meeting on the small Particles and Inorganic Clusters, Lausanne, Sept. 1980 N. Davidova, N. Peshev and D. Shopov, J. Cata1., 58 (1979) 193-205. K.G. lone, V.N. Rommanikov, A.A. Davidov and L.B. Orlova, J. Catal., 57 (1979) 126-135. M. Boudart, A.W. Adlag, J.E. Benson, N.A. Daugharty, C. Girvin Harkins, J. Catal., 6 (1966) 92-99. P.N. Galich et al., Sb "Neftekhimiya" izd A.N. Turk SSR Ashchabad, 1963, pp. 63. M.F Guil1eux, J. Jeanjean, M. Bureau-Tardy and G. Djega-Mariadassou, VIe Symp. Ibero Americana de Catalise, Rio de Janeiro, 1978. F. Figueras, R. Gomez, M. Primet, Adv. Chem. Ser., 121 (1973) 480-489. C.J. Brooks, Adv. Chem. Ser. 102 (1971) 426-433. D.J. Elliott and J.H. Lundsford, J. of Catal., 57 (1979) 11-26.