вход по аккаунту


Fourth International Catalysis Congress[.196808961.pdf]

код для вставкиСкачать
Fourth International Catalysis Congress‘*’
The Moscow Catalysis Congress from June 23 to 29, 1968
was the fourth in a series of four-yearly international congresses on heterogeneous catalysis. The kinetics and mechanisms of special reactions, new research methods, and the
properties of little-known catalysts were dealt with in six
plenary lectures and 86 discussion lectures, but the principal
topic was the relation between the chemical composition,
crystalline and electronic structure, and the activity and selectivity of catalysts.
1. The Active Center
Though many systems are now known in which the entire
catalyst surface is uniformly active, the particularly high
activity of special centers on the surface of others is partly due
to geometric or topochemical factors. Uniformly active surfaces have been found on metal oxides i n simple combustions
and on metals in simple hydrogenations. This is so e.g. according to P. C. Aben, J . C. Platteeuw, and B. Stouthamer
(Netherlands) (31) in the hydrogenation of benzene on Pt, Pd,
and Ni catalysts supported o n SiOz, AI2O3, A1203-Si02, and
MgO-SiO2, and according to M . Kraft and H . Spindler
(Leuna) (69) in the dehydrogenation of cyclohexane with Pt
on A1203. However, these authors also found that the surface
of the same catalyst is not uniformly active in the dehydrocyclization reaction of n-heptane. R. van Hurdeveld and F.
Hartog (Netherlands) (70) also point out that a catalyst can
have several types of centers, which can act differently in different reactions; nickel o n aerosil is reported to have a uniformly active surface in the deuteration of benzene to give
cyclohexane, but not in the H / D exchange between
benzene and Dz. Apart from crystallite corners, edges, surfaces, and lattice defects, active centers in metal catalysts include crystallite steps of atomic height (Bs sites), as was
described by G. C . Bond (England) (67).
The importance of the topochemical factor was discussed by
G. Purruvuno (USA) (11) for the 12C/14C exchange in CO&O
mixtures and for the decomposition of Hz02 and NzO on
cobalt ferrites having the composition c03-~
Fex04 (x =
1.9 . . . 2.1). Whereas the stoichiometric ferrite ( x = 2.0)
gives a minimum rate for the first reaction, it is found to be
the optimum catalyst for the other two reactions; this is
attributed to the change in the distribution of ion vacancies
with the Co/Fe ratio. B. V. Erofeev, N. V. Nikiforova, I. I .
Urbunovich, and L. D . Dmitrieva (USSR) (47) reported that
Cu and Mo o n A1203, MgO, SiOz, BeO, ZnO, and some
spinels are particularly effective in the dehydrogenation of
cyclohexane or the dehydrocyclization reaction of n-hexane
if the metals are atomically dispersed and situated in cation
sites of supports with octahedrally arranged anions. According to K. Turama, s. Yoshida, and Y . Doi (Japan) (13), Cr5+
ions, which are characterized by a n unusual crystal field, are
active in the copolymerization of ethylene on chromic oxide.
A geometric factor is particularly important for systems in
which multi-point adsorption of the reacting materials occurs.
For example, according to 0. V. Krylov and E. A. Fokina
(USSR) (64)’ the activity of metal oxides in the reaction of
acrolein with methanol increases with the lattice constant.
[*I In this report, the lecture number is given (in parentheses) in
each review, cf. Proc. IV. Internat. Congress Catalysis, Moscow
1968, in press.
0. N . Bragin and A . L. Libermann (USSR) (27) showed that
the hydrogenolysis of alkylcyclopentanes and cyclohexane on
Ru, Rh, Os, and Ir, which proceeds by two-point adsorption,
yields a large number of products, whereas on Pt, owing to
six-point adsorption, only the alkylcyclopentanes, and not
the cyclohexane, react. According to H. Noller, P. AndrPu,
E. Schmitz, S. Serain, 0. Neufang, and J . Girdn (Hannover
and Venezuela) (81), two-point adsorption is also responsible
for the stereospecific action of many catalysts in the dehydrochlorination and dehydrobromination of 2,3-dichloro- or
2.3-dibromobutane to form 2-halogenobutene. For example,
while KB02 and K2CO3 give a high cis-trans ratio in the
end products, CaC12 and Ca3(PO& give a low ratio; this
is explained by differences in the adsorption of substrate
halogen on the catalyst cation and of substrate hydrogen on
the catalyst anion.
In most insulating catalysts and in reactions with uncharged
adsorbed intermediates, acid groups o n the catalyst surface
act as active centers. This was shown by T. Yamaguchi and
K. Tanaba (Japan) (80) in the hydrolysis of dichloromethane
o n NiS04, ZnS. SiOz, and A1203-Si02, by Y. Noto, K . Fukuda, T. Onishi, and K . Tamaru (Japan) (37) and J. M .
Criado, J. Dominguez, F. Gonzhlez, G. Munuera, and J. M .
Trill0 (Spain) (38) in the dehydration of formic acid on
A1203, Si02, CrZ03, and Ti02 and by B. D . Flockhart. S. S.
Uppal, I. R . Leith, and R . C . Pink (Ireland) (79) in the H/D
exchange between n-propane and Dz on Al2O3. According to
H . Bremer and K . H. Steinberg (Merseburg) (76), T. V.
Antipina, 0. V. Bulgakov, and A . V. Uvarov (USSR) (77). and
T. Nishizawa, H. Hattori, T. Uematsu, and T. Shiba (Japan)
( 5 5 ) . acid groups also bring about the dehydrogenation of
isopropanol on MgO-SiO2 and cracking processes, as well
as the polymerization of ethylene and propylene and double
bond migrations in 1-butene o n partly fluorinated or hydrofluorinated A1203 catalysts. In all these cases the activity
increases with the concentration of Lewis- or Brmsted
2. The Catalyst-Substrate Bond
A knowledge of the nature and strength of the chemisorption
bond is particularly useful for the establishment of reaction
mechanisms and for the estimation of reaction rates. As was
emphasized by J. L. Garnett, R . J. Hodges, and W . A . SollichBuumgartner (Australia) (l), there is an analogy between
homogeneous and heterogeneous catalysis, as a result of
which the interaction between individual substrate molecules
and isolated catalyst structural units also approximately
describes the chemisorption on the surface of solid catalysts.
In the H/D exchange of monohalogenobenzenes, alkylbenzenes, and polycyclic aromatic compounds with DzO on
Pt, for example, the analogy is attributed to the n-complex
mechanism occurring in homogeneous and heterogeneous
catalysis. V. S . Feidblium (USSR) (I 5 ) referred to the analogy
in the dimerization and isomerization of a-olefins on Ziegler
catalysts. P. Cossee, P. Ros. and J. H . Schachtschneider
(Netherlands) (14) presented results of quantum-chemical
calculations that give a detailed picture of the energy relationships of the catalyst-substrate complex in the firstmentioned case. A. Cimino, V. Indovinu, F. Pepe, and M .
Schiuvello (Italy) (12) studied the action of isolated catalyst
structural units on the decomposition of NzO on solids
containing atomically dispersed CrzO3 as catalyst doped with
Angew. Chem. internut. Edit. / Vol. 7 (1968) 1 No. I 1
various amounts of Liz0 in a catalytically inactive MgO
H. H. Dunken and C . Opitz (Jena) (2) calculated the chemisorption bond of CO, HCN, OH, and hydrocarbons on
metals by the MO-LCAO method. The authors did not
examine only the interaction between substrate molecules
and isolated metal atoms, but also considered small metal
domains. The results correspond qualitatively t o a number
of important experimental results.
Some authors also examined the interactions between substrates and the solid’s electrons. However, a n electronic
factor can be expected only with semiconducting catalysts
and possibly with metallic catalysts, but not with insulators,
and also only in reactions with complete electron exchange
between substrate and catalyst (ionosorption). It is therefore
not surprising e.g. that F. Bozon- Verduraz and S. J. Teichner
(France) (6) found no change in the rate of hydrogenation of
ethylene on ZnO whose electron concentration was varied by
doping, since the rate-determining step in this case is not
associated with a n electron exchange. On the other hand,
T. Freund, S . R . Morrison, and W. P . Gomes (USA) (4)
found that in the oxidation of formic acid to Hz02 and COz
o n ZnO exposed to light, the concentration of electrons and
holes on the catalyst surface determines the reaction rate.
According to J. M . Criado. J. Dominguez, F. Gonzaler,
G. Munuera, and J. M. Trill0 (Spain) (38). the rate of dehydrogenation of formic acid o n Crz03 and o n Ti02 also
increases with the electron concentration, which is varied by
doping. C . Borgianni. F. Cramarossa, F. Paniccia, and E.
Molinari (Italy) (7) found that p-semiconductors (CrzO3, NiO)
are more active than n-semiconductors (ZnO, FezO3) in the
combustion of Hz. E. G. Schlosser and W. Herzog (Frankfurt/Main) (9) presented a semiquantitative interpretation of
the electronic factor on the basis of the combustion of CO
and HZ on doped NiO catalysts. Taking into account the
space-charge layers o n the catalyst surface, they were able
to show that both the increase in reaction rate with increasing
hole concentration and the decrease in reaction rate with
decreasing hole concentration are much smaller than expected. J. Scheve and I . W . Schulz (Berlin) (84) believe that
the electronic factor is due to energetic excitation of adsorbed
molecules as a result of collision with electrons of the solid.
The best catalyst is accordingly characterized by maximum
energy transfer with agreement between the thermal energy
of the electrons and special energy term differences of the
molecule, as was shown for the decomposition of NzO on
CuO/CrzO3 mixed catalysts.
N . P. Keyer (USSR) (8) mentioned that the “field effect”, in
which changes in the electron concentration o n the surfaces
of solids are brought about by a n electric field, should be
particularly useful for the investigation of electronic factors,
since the catalyst mixture does not change in this case. Results have already been obtained for the decomposition of
isopropanol o n TiOz. On the basis of investigations on the
hydrogenation of benzene on Cu/Ni alloys, D . A . Cadenhead,
N . J. Wagner, and R. L. Thorp (USA) (26) warn against
overestimation of the electronic factor. Whereas the decrease
in the reaction rate observed in this case with increasing Cu
content had previously been attributed entirely to the filling
of the electron vacancies in the 3d band, the authors now
believe that the only important factor is the Ni concentration
o n the catalyst surface, which can differ from that in the
interior as a result of a separation effect.
3. Catalytic Activity and Thermodynamic
AS was shown by W. M . H. Sachtler, G. J. Dorgelo. J. Fahrenfort, and R . J. H . Voorhoeve (Netherlands) (34) for the oxidation of benzaldehyde to benzoic acid o n MnOz, V205, and
Vz05/SnO~mixed catalysts, by V. E. Ostrovsky and N . N.
Angew. Chem. internat. Edii. 1 Vol. 7 (1968) / No. 11
Dobrovolsky (USSR) (46)for the combustion of Hz on Cu,
Ag, and Au, and by A . Cimino, V. Indovina, F. Pepe, and
M . Schiavello (Italy) (12) for the decomposition of N 2 0 o n
Cr203iMgO mixed catalysts doped with LizO, the activity of
the catalysts increases as the chemisorbed oxygen becomes
more loosely bound. This is not surprising, since the reaction
consists essentially of two steps, i.e. the combination of the
organic substrate or of hydrogen with chemisorbed oxygen
and the re-chemisorption of oxygen o n the catalyst, the first
step being rate-determining. Similar relationships were found
by G. K . Boreskov, V . V. Popovsky, and V. A. Suzonov
(USSR) (33) in the 160:180 exchange in oxygen isotope mixtures on oxides of the first series of transition metals and in
the combustion of Hz and methane o n cobaltites and
chromites. D . G. Klissurski (Bulgaria) (36) confirmed these
findings in the oxidation of methanol to formaldehyde o n
transition metal oxides and o n SbzO3, W03, and Sn02, and
K . Kochloej, M . Kraus, and V. Baiant (CSSR) (85) mentioned
in this connection the dehydration of secondary alcohols on
SiO2, Ti02, ZrOz, and Alz03.
In all these cases there is a linear relationship between the
logarithm of the reaction rate or the activation energy and
the bonding energy of the surface oxygen i.e. the free energy
of surface oxygen release. This linear relationship, which is
also known as the linear free energy relationship (LFER),
may according to V. A. Roiter, G. I. Golodetz, and Y. I.
Pyatnirzky (USSR) be regarded as a solid analog of the
Hammett or Taft relationship found for homogeneous
organic reactions. The authors also showed in the combustion
of Hz and propane on transition metal oxides that the activity
cannot be increased indefinitely by the use of catalysts having
very low oxygen-binding energies; on the contrary, there is an
optimum catalyst for which the oxygen-binding energy is
roughly equal to half of the enthalpy of the overall reaction.
If the oxygen-binding energy is further decreased the catalyst
activity decreases again, since the oxidation of the catalyst
now becomes rate-determining instead of its reduction. A
plot of the activity against the oxygen-binding energy gives a
diagram having the volcano shape described by Balandin.
R. A . Gnrdner (USA) (83) also assessed the catalytic activity
from the strength of the catalyst-substrate bond by evaluation
of special bond strengths of the intermediates, which affect
the IR vibration frequency, as shown for the combustion of
Hz and the combination of H2 and CO.
4. Statistical Methods for the
Discovery of Catalysts
The work of I . I. Joffe. V. S. Fyodorov, B. Y. Gurevitch, M . A .
Ustrayh, I . S . Fux,and K . M . Muhenberg (USSR) (63) may be
of special practical interest. The authors use the parameter
calculation method known from statistics t o find active and
selective catalysts, e.g. for the oxidation of hydrocarbons, and
illustrated the process in the case of the combustion of CO on
metal oxides. It was found that out of 20 tabulated physicochemical quantities of the catalysts, only eight are significant
for the catalytic activity. These quantities in order of increasing influence, are density, color index, melting point, vacancy
volume, electronic structure factor, susceptibility, concentration of d-electron vacancies, and ionization potential of the
metal. On the basis of these quantities, a number of metal
oxides that have not yet been tested were chosen as promising
catalysts for the combustion of CO.
D . A . Dowden, C. R. SchneN, and G. T. Walker (England) (62)
showed how active and selective catalysts for complicated
reactions can be found by the use of an arrangement scheme
by means of which a large volume of experimental data can
be taken into account, e.g. catalysts for the steam reforming
of paraffins.
[VB 169 IE]
German version: Angew. Chem. 80,917 (1968)
Без категории
Размер файла
285 Кб
international, congress, catalysing, pdf, 196808961, fourth
Пожаловаться на содержимое документа