Catalysis Today,3 (1988)387-394 Elsevier Science PublishersB.V., Amsterdam-PrintedinThe
ZEOLITE
R.
CATALYSIS
AND MOLECULAR
387 Netherlands
VIBRATIONS
LARSSON
Group of Catalysis of Lund,
Research
Inorganic
,
P 0 Box 124, S - 221 00 Lund
Chemistry, f Sweden
University
f
ABSTRACT A model for catalysis based on selective energy transfer from catalyst to the reactant is applied to some reactions catalyzed by zeolites or zeolite-related systems. It is indicated that the vibration mode that brings hydrocarbons to react is most probably the skeletal vibrations around 1200 cm-l.
INTRODUCTION A molecule
in contact
with, or forming
phase
will be in thermal
equilibrium
inter
alia, that when the molecule
collision
) it will rapidly
situation
stands
For this energy
latter
Such considerations absolute
rates
molecule
basis
supply
energy
to the vibration
phase
of the model
modes
used
excess the reaction.
,however,
of
the
source of
for the dissipation
of the condensed
for catalysis
.
(ref. 1 1.
must have a steady
that more than compensates
molecule
for treatments
its state of activation
in the condensed
by
This
of an acquired
mode may trigger
such as, e.g., the RRKM theory
energy
basis
vibration
are the principal
In order to reach
t e.g.
energy.
to that of a free gaseous
case, the statistical
into the critical
This means,
has been excited
lose its excess
in contrast
part of, a condensed
with this phase.
phase.
of
This is the
in this discussion.
ZEOLITES The above molecule
view is specially
entrapped
any other molecules
in the cavity present
enough
to give, by collision, of the potentially the molecule
in the case of a
of a zeolite.
There are few if
which have a kinetic
energy
to excite
obvious
an
reactive
energy
additive
molecule.
energy
high
to the free
The source
must thus come from the zeolite
of energy structure
333 itself by the transfer of vibrational energy from the zeolite to the molecule. It has been argued elsewhere (ref. 2 ) that the role of a catalyst is, inter alia, to supply energy by vibrational resonance directly into that vibration mode of the reacting molecule that most closely represents that distortion of the molecule which is critically needed for the reaction to proceed. Furthermore (ref. 3 1 , it
has been argued that the substrate
-catalyst complex can be regarded as an analogue to the classical concept of a forced, damped oscillator. The damping ( = friction loss of energy ) can be thought of as being analogous to the dissipation of energy of the partly excited molecule as discussed above. Applying the basic rules of classical, mechanics to this model ( ref. 3 ) molecule,V,
a relation between the frequency of the
the frequency of the catalyst,
isocatalytic temperature, 8,
W,
and the
was obtained. This relation took
the form
8 = N!E 9'
2
V -w ___-~ VI
2
. ---------~
1
____-_
(1)
"0 4.~12 - arctg -_--2 2 I li -a,*)
It has also been argued (ref. 2 ) that if
v
=w,
a
state of resonance will cause the most effective transfer of energy and thus cause the most rapid reaction.
It also follows
from this concept, and is indeed observed ( ref. 4,5 ), that for a certain catalytic reaction effected by a series of similar but non-equal catalysts, the energy of activation will change in a stepwise manner from one catalyst to another. The statements above can now be illustrated by some data from literature which are relevant to zeolite catalysts. As will be seen, some degree of subjectivity is still left in the choice of the proper vibration mode of the reacting molecule. In some cases, however,this mode can be determined by the analysis of the stepwise change of the energy of activation. Let us start with such an example.
389
BOUDART'S
OBSERVATION
OF THE COMPENSATION
( ref. 6 ),
In an early paper existed
relationship
a
cracking
of a series
alumina-silica
(2),
quantities
lnA=bE
+
related
relate
that there for the
hydrocarbons
that the logarithm related
observed parameters
on an
( the "compensation
This relation
(In A) was linearly
(E). These
the kinetic
of closely
system.
was such, cf. eqn factor
betwen
EFFECT
Boudart
effect")
of the preexponential
to the energy
to the Arrhenius
of activation
equation,eqn
(3).
c
( 2 1
In k = In A - E/RT
(3)
b = l/R@
(4)
According temperature
( ref. 6 ), the isocatalytic
to Boudart
6 was found to be 950 K.
The data that were used,
however,
seem to be have been misquoted
original
investigation
a conference
discussion
from the authors
( ref. I,8 ) .The data given by Boudart ( ref. 9 ),
and Nicholson
As the present
with the data of
( ref. 7,8 ).
author
is completely
in the numbers,
this discrepancy
in
and used by him thereafter
( ref. 6 ), are shown in Table 1 together Franklin
of the
unaware
the original
of the origin of
data are
preferred.
All the compounds rate law. Others, In Table in other
sucessive
kcal/mole
are muliples
change
of activation
formed between
energies
the
that to a very good
of one and the same factor,
. This latter value corresponds
with a wave number
did not.
as has been used
. It is found that these
form a set of values
approximation
a first order
and n-hexane
is undertaken
has been
in the table
1 obeyed
n-pentane
on the stepwise
The difference
entries
differences
Table
2 the same procedure
dicussions
( ref. 4,5 ).
shown in
viz., n-butane,
to vibrational
of 3730 cm ml or more probably
10.7 energies
3 x 1244 cm-l.
390
TABLE
1
Activation
energies
Compound
for hydrocarbon (ref. 6 )
Ea
(kcal/ mole)
cracking
on alumina-silica.
( ref. 7,8 ) In
E,
A
(kcal/mole)
Propane
43
41.7
19.56
2-methylpropane
35.3
30.5*
14.85
2-methylbutane
31.3
20.5
7.34
2,2_dimethylpropane
23.0
40.0
18.20
Cyclohexane
18.5
18.5
5.99
*
A corrected
** Calculated
TABLE
value
( ref. 8 )
from original
data
(ref. 7,8 )
2
Representation
of the differences
the five different made
**
in another
column,
substances.
between
activation
energies
If the order of the compounds
way, there will appear
other numbers
Activation Ea
energy
Difference
Factor
AE
(kcal/mole)
(kcal/mole)
Propane
41.7
11.2
1
2-methylpropane
30.5
10.0
1
2-methylbutane
20.5
19.5
2
2,2_dimethylpropane
40.0
21.5
2
cyclohexane
18.5
23.2
2
8
Total
sum of AE
85.4
Mean
BE
10.7
is
in the last
e.g. 0.
Compound
for
391 The reason that the increment occurs in sets of three is not obvious, but it is of great interest to note that the wave number 1244 cm -l is of the right size
to correspond
to a skeletal
vibration mode of the hydrocarbons under discussion. to the empirical assignments of Bellamy
( ref. 10 ),
According a
(CH3)3-C-R group has IR absorption bands at 1250 f 5 cm-l
and at
1250 -1200 cm-l whereas a (CH3)2-C group has absorption bands at 1170 f 5 cm-l
and at 1170 - 1140 cm-l.
In A 20-
15-
lo-
E/kcal
I
I 2b
I 30
’
mol’l I 40
I
Fig. 1. The compensation effect of
hydrocarbon cracking ( ref.
7,8 ) over alumina-silica. Now let us look at the compensation effect in terms of these original data. In Fig 1 In A from Table 1 has been plotted against the experimental activation energies ( ref. 7,8 ). The only point falling somewhat off a straight line is the one for 2-methylpropane. The data for activity for this compound, however,
were taken from one source ( ref. 7 ), whereas the
value for E, was from another one ( ref. 8 ). This might give the reason for a slight discrepancy. Otherwise, the data refer to measurements at only a few ( reported ) temperatures and the errors
may therefore be not completely ignorable.
392
A least square fit to these five points give a line with the slope
b = 0. 57 mol/kcal. From eqn (4) this corresponds to an
isocatalytic temperature of 877 K . Application of formula (1) tells us that under almost resonance conditions this value corresponds to n = 1226 cm-l. The agreement with the value obtained from the E, increment ( 1226 compared to 1244 ) is very good. This indicates that the molecular vibration that determines the catalytic reaction is probably the skeletal vibrations indicated above ( for the free molecules or for the corresponding protonized species ) and that the vibration of the zeolite catalyst will also be close to this value. Inspection of the infrared spectra of Y-zeolites ( ref. 11 ) indicates
a substantial splitting of the asymmetric (Al,Si)-0
stretch, giving one component around 1200 cm-l , the major one still near 1000 cm-l.
Hence a good opportunity for resonance
between catalyst and reacting molecule could exist. One must note that the errors in the estimation of the frequencies quoted above must be around +50 cm-l. THE CORRFLATION OF DEROUANE Having now discussed a case with one zeolite catalyst in common for reaction with a series of slightly different substrates, let us now turn to the contrary ; the catalytic reaction of one and the same compound at a series of slightly different catalysts. Very recently, Derouane ( ref. 12 ) has suggested a correlation between the catalytic activity of a series of zeolites and their pore dimensions. Plotting the data of Kikuchi et al.( ref. 13 ) for pentane cracking
against the pore size of the zeolites used,
he obtained a typical "volcano -shaped" relation. The position of the maximum should, in Derouane's interpretation,
correspond to the optimum pore size and optimum
possibility for interaction between substrate and catalyst. From the results above, however,
it seems to be of interest to
investigate also the relation between catalytic activity proper vibrations of the zeolite systems. give some indication of
which
and the As the above results
vibrations might be active, the
asymmetric stretch of the various zeolites has been chosen. With
393
no knowledge of the experimental details ( ref. 13 )
of the original work
a coarse estimation of the Al/Si ratios has been
used and hence, from the tabulations of Flanigen, Khatami and Szymanski ( ref. 14 ),
the corresponding vibration frequencies.
They are used in Fig 2 where the activity of pentane cracking is plotted agianst the asymmetric stretch frequencies.A good correlation is obtained.
From this curve, one can interpret the
efficiency of the ZSM-5 type catalysts as originating from the fact that their vibration frequency is closest to the frequencies of C-H deformation or skeletal vibration modes of the molecule concerned. In want of better information, it is not possible to discuss a correlation in terms of the upper branch of the split band, the one being close to 1200 cm-l.
log(TOF) -1 -
Fig. 2.
The turnover frequency of pentane cracking versus the
wave number of the asymmetric stretching vibration of the zeolites.
CONCLUSION To sum up and conclude from the two
examples of relation-
ships between activity of zeolitic catalysts and molecular vibrations, one might observe offer a pool of energy from
that any zeolite system that can
vibrations close to that one needed
394
to activate
the reaction
of vibrational 1400 cm-l
energy
( ref. 15 ).
temperatures
so low
effective.
This observation
( ref. 16 )
efficiency
will be highly
Such a source
might be found in the NH4+ vibration
of ammonium
might
explain
- containing
(<27OoC ) that ammonia
around
the
zeolites
at
( ref.17 )
is not
desorbed.
REFERENCES 1.
R.P. Wayne,
in
Comprehensive
C.H. Bamford
Chemical
Elsevier,
Kinetics,
and C.F.H.
Kinetics.
Chem. Scripta,
2.
R. Larsson,
3.
R. Larsson,
Chem. Scripta
4.
R. Larsson,
J. Mol.
, 27
Catal.,
Chem. Eng. Progr.,
J. L. Franklin
8.
J. L. Franklin
of
3.
(1987) 371.
38
R. Larsson, M. Boudart,
),
12 (1977) 78:
Z. physik.
6.
( Editors
2, The Theory
1969, Ch.
5.
7.
Vol.
Amsterdam,
Tipper
Chemie
(1986) 71
),
(Leipzig 57
268
( 1987) 721.
(1961) 33.
and D. E. Nicholson,
J. Phys. Chem.,
60
J. Phys. Chem.,
61
(1956) 59. and D. E. Nicholson,
(1957) 814. 9.
Adv. Catal.,
M. Boudart,
Methuen
& Co, Ltd, London, and
11. K. Oinuma quoted
New York,
12. E. G. Derouane,
Gakkaishi,
H. Khatami
Zeolites,
D.C.,
1971 , p. 201. Infrared
John Wiley
16. J. C. Miale,
100
8:l
( 1967)
Sieves,
and Y. Morita,
and H. A. Szymanski,
Spectra & Sons,
Am.Chem.
Murakami,
Sekiyu
of Inorganic
Molecular Sot., Wash.
and Coordination
Inc., New York, London,
N. Y. Chen and P.B. Weisz,
J. Catal.,
278. 17. C. V. Hidalgo,
;
John Wiley
(1986) 541.
K. Shimomura
Adv. Chem. Ser. 101;
Sieve
Compounds,
Molecular
(1985) 210.
14. E. M. Flanigen,
15. K. Nakamoto,
Molecules,
1974.
J. Catal.,
28
of Complex
J. Toyo Univ.
Zeolite
H. Nakano,
13. E. Kikuchi,
Spectra
1962.
H. Hayashi,
from D.W. Breck,
& Sons,
9 (1957 ) 636.
The Infra-red
10. L. J. Bellamy,
H. Itoh, T. Hattori,
J. Catal.,
85 (1984) 362.
M. Niwa
and Y.
1963.
6 ( 1966)