Dehydrogenation of cyclohexanol on copper-containing catalysts: I. The influence of the oxidation state of copper on the activity of copper sites

Article abstract of DOI:10.1006/jcat.2000.2979

The active sites of copper-containing catalysts of both dehydrogenation of cyclohexanol to cyclohexanone and aromatization of cyclohexanol to phenol were determined using XRD, XPS, IR spectroscopy of chemisorbed CO and kinetic data. Monovalent copper and

Full text of DOI:10.1006/jcat.2000.2979

Journal of Catalysis 195, 20–30 (2000)
doi:10.1006/jcat.2000.2979, available online at http://www.idealibrary.com on
Dehydrogenation of Cyclohexanol on Copper-Containing Catalysts
I. The Influence of the Oxidation State of Copper on the Activity of Copper Sites
1
2
V. Z. Fridman and A. A. Davydov
Reflex Inc. (Catalysis Center, Research and Consulting), New York, New York
Received October 21, 1999; revised June 23, 2000; accepted June 23, 2000
mechanisms of the main reaction of ketone synthesis and
The X-ray diffraction, X-ray photoelectron spectroscopy, IR aromatization ofcyclohexanolare not fullyunderstandable.
spectroscopy of chemisorbed carbon monoxide and kinetic data
have been employed to determine the active sites of copper-
containing catalysts of both dehydrogenation of cyclohexanol to
cyclohexanone and aromatization of cyclohexanol to phenol. Two
kinds of copper active sites of the cyclohexanol dehydrogenation
reaction to cyclohexanone have been revealed (monovalent copper
and metallic copper), and it has been shown that monovalent copper
is significantly more active than metallic copper, and monovalent
copper does not catalyze the phenol formation. At the same time,
metallic copper can play the role of active sites not only in alcohol
dehydrogenation to ketone but also in aromatization of cyclohex-
anol to phenol.
�c 2000 Academic Press
A number of investigations (4–10) have focused on deter-
mining the influence of support, of the methods of prepara-
INTRODUCTION
Industrial catalysts of dehydrogenation of cyclohexanol tion, and of copper loading on the activity and selectivity of
to cyclohexanone are generally based on copper as an ac- copper-containing catalysts, with the goal of enhancing the
tive component of catalytic systems. The commercial Cu– conversion of alcohol. Attempts to increase the conversion
Mg and Cu–Zn–Al catalysts usually are used in the tem- of cyclohexanol by using new copper catalysts and higher
�
perature range from 220 to 260 C, and the conversion of reaction temperatures have not been successful because of
cyclohexanol on these catalysts is very close to equilibrium a sharp decrease of the selectivity of the process as a result
values (50–60% ) at the 99% selectivity for cyclohexanone of a sharp increase of the phenol yield in the products of the
(
1). The process of cyclohexanol dehydrogenation includes reaction. The reasons for a decrease in the selectivity of the
the complex of consequent-parallel reactions, such as de- process and a way to maintain the selectivity of the process
hydrogenation of cyclohexanol to cyclohexanone (the main for increasing the alcohol conversion were not clear. Under-
reaction), aromatization of cyclohexanol to phenol, dehy- standing the mechanisms and identifying the active sites of
dration of cyclohexanol to cyclohexene, and condensation these reactions can help solve this problem. Detailed inves-
of cyclohexanone to cyclohexyldencyclohexanone (the last tigations of copper-containing catalysts of methanol syn-
reactions are the formation of byproducts). The mecha- thesis and the water-gas-shift reaction (11–22) have been
nisms of dehydration of cyclohexanol and condensation of reported over the past 20 years, which have shown the pres-
cyclohexanone are known and the rates of these reactions ence of different copper sites on the surface of copper cata-
can be decreased by the acidity of the catalysts (2, 3), but the lysts. In contrast very few studies have been devoted to
an elucidation of the mechanism of the cyclohexanol dehy-
drogenation in an attempt to connect the activity of copper-
containing catalysts with the oxidation state of surface cop-
per (3, 23–26). Sivaraj et al. (23) have studied Cu–Zn–Al
catalysts prepared by a deposition-precipitation method in
the reaction of cyclohexanol dehydrogenation. They have
1
To whom correspondence should be addressed at 110-20 73Rd. Apt.
2
M, Forest Hills, NY 11375. E-mail: vfridman@sprintmail.com.
2
Present address: Benedum Engineering Hall, Department of Chem-
ical and Petroleum Engineering, University of Pittsburgh, Pittsburgh, PA
5261. E-mail: davydov@pitt.edu.
1
20
0021-9517/00 $35.00
Copyright �c 2000 by Academic Press
All rights of reproduction in any form reserved.

DEHYDROGENATION OF CYCLOHEXANOL
21
determined the direct correlation of catalyst activity and the cipitate was repeated four times. The precipitate was dried
�
amount of reversible carbon monoxide uptake with a max- at 110 C for 16 h to produce the precursor catalyst. The
imum at 30 at.% Cu loading. The authors concluded that Cu–Zn and Cu–Zn–Al precursor catalysts were finally cal-
�
metallic copper species are the active sites for the dehy- cined at 350 C and Cu–Mg precursors were finally calcined
�
drogenation activity of the copper catalysts. In our opinion, at 500 C for 4 h in air. The Cu–Zn, Cu–Zn–Al, and Cu–Mg
the conclusion of the authors (23) that metallic copper is catalysts were prepared with different copper loadings from
an active site for the reaction of dehydrogenation of cyclo- pure ZnO, Zn–Al system, and pure MgO up to pure cop-
hexanol is not complete. Our analysis of this results showed per oxide. Cu–Zn–Al catalysts were prepared at a constant
that these investigators (23) have not considered the pre- concentration of aluminium equal to 15 at.% . The compo-
vious studies (11–22), nor their own (25), where it was de- sitions of catalysts are presented in Tables 1, 2, and 3.
termined that Cu–Zn–Al catalysts with a 30 at.% copper
loading have not only a maximum amount of reversible
X-ray diffraction analysis. X-ray diffraction (XRD)
analysis of catalysts was carried out using DSOOSIMENS
carbon monoxide uptake but also a maximum amount of
with a copper anode and a nickel filter. The crystallite sizes
irreversible adsorbed carbon monoxide. It should be noted
of metallic copper and copper oxides were estimated from
that the results of the later study (25) are in a good agree-
Scherrer’s equation (28). These data are summarized in Ta-
ment with results of the earlier studies (21, 22). It is well
bles 1, 2, and 3 together with the BET surface areas of the
known that irreversible adsorption of carbon monoxide is
catalysts (N2 adsorption at 77 K).
directly proportional to the amount of monovalent copper
Measurements of X-ray photoelectron spectra (XPS).
on the catalyst surface (27). The observed (23) maximum
XPS spectra were obtained only for Cu–Zn–Al catalysts
of activity at 30% copper is, therefore, not only because
before and after reduction and use in the reaction of dehy-
of zerovalent copper sites but also because of monovalent
drogenation of cyclohexanol on an ESCA-3 spectrometer,
copper sites. In our previous studies of the cyclohexanol
using AlKa radiation (1486 eV). The recording was con-
dehydrogenation reaction (3, 26) the catalysts that had the
ducted in the range-binding energies from 0 to 1100 eV and
highest probability for monovalent copper on the surface
then in the spectral region corresponding to the levels Cls,
were more active than catalysts without monovalent cop-
Ols, Cu2p3/2, Zn2p3/2, Al2s, and Al2p. The C1s-binding en-
ergy value of 285 eV has been used as a reference level. The
per. On the basis of these (3, 26) and other reported studies
(
14, 15, 21, 22), we have proposed that copper-containing
surface composition of atoms in Cu–Zn–Al catalysts was
determined on the basis of the peak area intensities of the
Cu(2p3/2)/Zn (2p3/2) levels. The relative atomic sensitivity
factor used here was obtained from the experimental values
ofWagner etal. (29) (Cu(2p3/2)/Zn (2p3/2) = 0.88). The re-
duced Cu–Zn–Al catalysts were transferred to a nitrogen-
filled box attached to spectrometer without exposure to air.
catalysts have two types of active sites of copper for the
reaction of cyclohexanol dehydrogenation.
The present study deals with the surface properties of
Cu–Zn, Cu–Zn–Al, and Cu–Mg catalysts with different
copper loading with the aim of determining active sites of
the reactions of cyclohexanol dehydrogenation and aroma-
tization. The main focus is on changes in the oxidation state
of copper and on the dependence of the oxidation state of
copper on activity.
Catalytic activity. The activity of catalysts was studied in
a continuous flow-type reactor with a fluidized layer of cata-
�
lyst at 250 and 300 C of flow rates in the range of 0.5–4.0 h.
Initially the samples were reduced by mixture (3% H2 and
N2) at 250 C for 16 h. Reduction times were always three
EXPERIMENTAL
�
times the time necessary for stoichiometric reduction of
CuO to Cu . The comparison of the catalysts activities in the
Catalysts preparation. The Cu–Zn, Cu–Zn–Al, and
Cu–Mg catalysts with various atomic compositions were
prepared by coprecipitating copper, zinc, aluminum, and
magnesium nitrates with a 10% solution of sodium
carbonate. Appropriate amounts of Cu(NO3)2 � 3H2O,
Zn(NO3)2 � 6H2O, Al(NO3)3 � 9H2O, and Mg(NO3)2 � 6H2O
were dissolved in 50 mL of distilled water to obtain 5 g
0
cyclohexanol dehydrogenation reaction to cyclohexanone
�
(
Reaction 1) was made at the same temperature (250 C)
at different reactant space velocities, and at the same con-
version of alcohol (=30% ), the comparison of the catalysts
activities in the cyclohexanol aromatization reaction to phe-
�
�
nol (Reaction 2) wasmade at the same temperature (300 C)
of each catalyst. After the solutions were heated to 80 C,
and also at different reactant space velocities at the same
phenol yield (2.5% ).
4
0 mL of Na2CO3 solution was added dropwise over a pe-
riod of 20 min under vigorous stirring, followed by aging
for 1 h at the same temperature. During the aging, the fi-
IR spectroscopy. IR spectroscopy of adsorbed probe
nal pH of solution was adjusted to 7.0 by the addition of a molecules (30, 31) has been used to determine the oxida-
small portion of the Na2CO3 solution. The precipitate was tion state of the copper sites. IR spectra of adsorbed species
�
�
filtered and suspended in 100 mL of distilled water at 50 C were recorded in the temperature range from 25 to 300 C
for 10 min, followed by filtration. The washing of the pre- with an infrared spectrometer (UR-20). The catalysts were

2
2
FRIDMAN AND DAVYDOV
placed in a cell and subjected to exposure to a gaseous mix-
ture of varying compositions at different temperatures. The
recording of spectra was performed with compensation of
the gas phase. The details of this method have been de-
scribed in (30). To estimate the state of copper oxidation,
two criteria were used, namely, the value of � CO vibration
and thermal stability of the CO complexes with the copper
cation. It is known (31, 32) that in the spectra of CO ad-
sorbed on the copper cations the values of � CO depend on
the charge and the structure of the nearer environment of
TABLE 2
Properties of the Cu–Zn Catalysts before and after Use
in the Reaction of Dehydrogenation of Cyclohexanol
Catalyst
BET
Phase
composition surface
Cu crystallite
size (nm)
composition; XRD
Number
of catalyst
(Cu/Zn)
(at.% )
area
2
(m /g)
Fresh
After work d = 0.2088
1
2
3
4
5
6
7
8.0/92.0
20.0/80.0
30.0/70.0
50.0/50.0
70.0/30.0
80.0/20.0
100
67.0 ZnO
73.0 ZnO
68.0 ZnO, CuO ZnO, Cu
59.0 ZnO, CuO ZnO, Cu
61.0 ZnO, CuO ZnO, Cu
ZnO
ZnO
—
—
0
� 1
1+
12.9
12.4
18.9
24.6
31.5
copper (Cu –CO <2100 cm , Cu –CO between 2110 and
� 2+ � 1
1
2
140 cm , Cu –CO >2140 cm ) and different thermal
0
stabilities of the complexes (Cu –CO is easily destroyed at
54.0 CuO
15.0 CuO
Cu
Cu
�
1+
�
the temperature range 25–100 C and while the Cu –CO is
a stable complex, it cannot be destroyed even at 100 C (31).
IR-spectra of carbon monoxide adsorbed on Cu–Mg
catalysts, which were taken after 10 min of exposure of
RESULTS
the catalysts with copper loadings of 10, 25, and 52% to
Catalysts Characterization
�
5
0 Torr CO at 25 C, are presented on Fig. 1. For catalyst
Cu-Mg catalysts. A series of copper–magnesium cata- with 10% copper, a single symmetric band is observed cen-
�
1
lysts with final compositions ranging from pure magnesium tered at 2080 cm after reduction in an atmosphere of H2
�
oxide to pure copper oxide were prepared. X-ray analysis (Fig. 1A, a). An evacuation of this catalyst at 25 C for 3 h
showed a mixture of CuO and MgO was obtained after cal- resulted in disappearance of this band (Fig. 1A, a, dotted
cination (Table 1). Even the Cu–Mg catalyst with 10 at.% lines). The similar dependence was observed for the catalyst
copper loading (that is equal to 20 wt% ) after calcination with 25% copper loading (Fig. 1A, b). Adsorption of CO on
has in the bulk a trace amount of CuO, and as follows from the sample with 52% Cu (spectrum c) resulted in a shift of
�
1
(
33) the rest part of copper is in the solid solution of cop- the band position to 2090 cm and an appearance of a small
�
1
per in MgO. The increase of copper loading leads to en- shoulder at 2140 cm . Subsequent evacuation of this sam-
�
1
hancement of the intensity of the peaks of CuO in X-ray ple led to disappearance of the band at 2090 cm (Fig. 1A,
diffraction patterns. After reduction of Cu–Mg catalysts in c). According to (31, 32) the positions of this band, maxi-
an atmosphere of hydrogen, the peaks of copper oxides dis- mum and thermal stability of adsorbed CO with cations are
appeared and the peaks of metallic copper appeared even in characteristics of lineraly adsorbed CO on metallic copper
0
catalysts with 10% copper loading (Table 1). The crystallite (Cu
)
.
size of metallic copper after reduction in Cu–Mg catalysts
remained almost the same independent of copper loading
up to 80% . The size of these crystallites has been increased
significantly (Table 1) in Cu–Mg catalysts with copper load-
ing equal to 80% and larger.
Cu-Zn catalyst. The XRD data of the fresh and reduced
Cu–Zn catalysts presented in Table 2 indicate the depen-
dence of the bulk phase composition on copper loading,
and that X-ray diffraction patterns of Cu–Zn catalysts up
to 20% do not display any peaks of CuO in fresh samples
nor metallic copper peaks in reduced samples (Table 2).
Only the ZnO phase was observed in catalysts equal to
TABLE 1
2
0% copper and less. Based on these results and the con-
Properties of the Cu–Mg Catalysts before and after Use
in the Reaction of Dehydrogenation of Cyclohexanol
clusions of Klier et al. (14, 15) we can assume that a large
part of copper in these samples exists in the solid solution
of copper in ZnO. X-ray analysis showed a mixture of CuO
and ZnO which was obtained after calcination of catalysts
with copper equal to 30% and more. After reduction of
Catalyst
composition surface
(Cu/Mg)
(at.% )
BET
Phase
composition; XRD
Cu crystallite
size (nm)
Number
of catalyst
area
2
�
(m /g)
Fresh
After work d = 0.2088
these catalysts in an atmosphere of H2 at 250 C, copper
oxide reduces to metallic copper (Table 2). As for Cu–Mg,
the increase of copper loading in the Cu–Zn catalysts leads
to enhancement of the amount of the CuO phase. The size
crystallites of metallic copper after reduction of the Cu–Zn
catalysts remained almost the same independent of copper
loading up to 70% . The size of these crystallites increased
significantly (Table 2) in the Cu–Zn catalysts with copper
1
2
3
4
5
6
7
10.0/90.0
25.0/75.0
35.0/65.0
52.0/48.0
70.0/25.0
80.0/20.0
100
137.0 MgO, CuO MgO, Cu
121.0 MgO, CuO MgO, Cu
97.0 MgO, CuO MgO, Cu
89.0 MgO, CuO MgO, Cu
69.0 MgO, CuO MgO, Cu
12.7
10.4
11.9
11.7
13.9
21.6
31.5
53.0 CuO
15.0 CuO
Cu
Cu

DEHYDROGENATION OF CYCLOHEXANOL
23
FIG. 1. IR spectra adsorbed CO on the surface Cu–Mg, Cu–Zn, and Cu–Zn–Al catalysts with different Cu loading. (A) IR spectra adsorbed CO
�
�
�
at 25 C on the surface of Cu–Mg catalyst after reduction by H2 at 250 C for 5 h, followed by evacuation at 250 C for 3 h: (a) 10% Cu at.; (dotted
�
�
line) sample a after evacuation at 25 C for 1 h; (b) 25% Cu at.; (dotted line) sample b after evacuation at 25 C for 1 h; (c) 52% Cu at.; (dotted line)
sample c after evacuation at 25 C for 1 h. (B) IR spectra adsorbed CO at 25 C on the surface of Cu–Zn catalyst after reduction by H2 at 250 C for 5 h,
followed by evacuation at 250 C for 3 h: (a) 8% Cu at.; (dotted line) sample a after evacuation at 25 C for 1 h; (b) 20% Cu at.; (dotted line) sample b
after evacuation at 25 C for 1 h; (c) 30% Cu at.; (dotted line) sample c after evacuation at 25 C for 1 h. (C) IR spectra adsorbed CO at 25 C on the
surface of Cu–Zn–Al catalyst after reduction by H2 at 250 C for 5 h, followed by evacuation at 250 C for 3 h: (a) 5% Cu at.; (dotted line) sample a
after evacuation at 25 C for 1 h; (b) 15% Cu at.; (dotted line) sample b after evacuation at 25 C for 1 h; (c) 20% Cu at.; (dotted line) sample c after
evacuation at 25 C for 1 h.
�
�
�
�
�
�
�
�
�
�
�
�
�
loading equal to 70% and more. This fact suggests a sharp parently this spectrum can be resolved into two separate
�
1
decrease of the dispersion and as result the surface area of bands centered at 2080 and 2115 cm , corresponding to ad-
copper in these samples.
sorbed CO on metallic Cu and to lineraly adsorbed CO on
+
In contrast to results for the Cu–Mg catalysts, IR spectra Cu (31). Evacuation of the sample with 30% copper after
ofadsorbed CO on reduced Cu–Zn catalystswith 8and 20% adsorption of CO at 25 C resulted in narrowing of the band
copper displayed symmetric bands centered at 2115 cm
�
�
1
� 1
at 2115 cm with the disappearance of low-frequency com-
(
Fig. 1B, a and b). Subsequent evacuation of these samples ponent (Fig. 1B, c, dotted line). The intensity of the peak
�
�
� 1
of catalysts at first at 25 C and then at 150 C did not sig- centered at 2115 cm did not change. These results suggest
nificantly change these spectral features (Fig. 1B, a and b, that copper on the surface of this catalyst exists as mono-
dotted lines). The position of these band maxima and ther- valent and metallic copper, in agreement with the results
mal stability of adsorbed CO are characteristics of linerly and conclusions reported elsewhere (20–22). The catalysts
adsorbed CO on monovalent copper (Cu ) (31, 32). More- with copper loading more than 30% after reduction in H2
over, it can be concluded from this spectrum that first, the at 250 C completely lost transparency in the wavenumber
copper ions on the surface were homogeneous in the in- range of2000–2300cm , and IR data ofadsorbed CO could
vestigated samples, because, according to (39), the surface not be obtained for these catalysts. It should be noted that
Cu ions exhibit the formation of two carbonyl complex the loss of transparency was observed by Boccuzzi et al.
+
�
�
1
+
�
1
type, with � CO 2115 and 2140 cm , and secondly, these (34, 35) after reduction of pure ZnO, which was attributed
data testify to the absence of the Cu2 oxide phase (within to the formation of donor centers VO + (Monoionized oxy-
the sensitivity limits of the method). These results are ad- gen vacancies). Boccuzzi et al. have observed that unlike
ditional indirect support of our assumption regarding the pure ZnO, the reduced binaries in the range of copper
formation of the solid solution copper with ZnO. Unlike loading up to 30% exhibit a good IR transparency. They
spectra of adsorbed CO on catalysts with 8 and 20% copper, have concluded that reduction conditions guarantee that
+
the spectrum of adsorbed CO on catalyst with 30% copper sufficient quantities of Cu2 are still present in the ZnO
+
displayed an asymmetric band in the wavenumber range of phase as a solid solution of copper in ZnO, at least in the
�
1
� 1
2
050 to 2140 cm centered at 2115 cm (Fig. 1B, c). Ap- bulk, to prevent the formation of high amounts of VO+.

2
4
FRIDMAN AND DAVYDOV
Boccuzzi et al. have not studied the Cu–Zn catalysts with
more than 30% copper. In our investigation we have ob-
served the loss of transparency also for catalysts at 40 and
5
0% copper. Based on Boccuzzi et al.’s explanation we
have assumed that in Cu–Zn catalysts, in the loading range
greater than 30% , CuO does not form the solid solution of
CuO in ZnO. Absence of the solid solution copper in ZnO,
which can keep Cu in ZnO even after reduction, leads to
the formation of distinct phases of copper and zinc oxides.
On the other hand, CuO can be easily reduced to metallic
copper. ZnO without soluble Cu2 has a great amount of
+
monoionized oxygen vacancies VO+, which bring about a
�
1
complete loss of transparency in the range 2000–2300 cm
in the IR spectrum of these catalysts.
FIG. 2. XPS spectrum of Cu–Zn–Al catalysts (40% Cu at.): (a) after
preparation; (b) after reduction by H2 at 250 C for 5 h.
�
Cu–Zn–Al catalysts. Cu–Zn–Al catalysts were pre-
pared with different concentrations of Cu and Zn and at with 10 an 15% Cu (Figs. 1C, a, b, and c). The copper ions in
the constant concentration of Al equal to 15 at.% . X-ray these samples were homogeneous and the states of copper
analysis of the Cu–Zn–Al catalysts with 5, 10, and 15% ions on the surface of both Cu–Zn and Cu–Zn–Al catalysts
�
Cu shows (Table 3) the existence of only ZnO and an ab- were similar. Evacuation of the samples even at 150 C did
sence of peaks from any copper and aluminum-containing not change the spectral picture (Figs. 1C, a, b, and c, dot-
compounds. This is due either to the formation of a solid ted lines), which is evidence of the high level of thermal
solution of Cu and Al in the lattice of ZnO or to the high dis- stability of adsorbed CO. The position of the band and the
persion of copper and aluminum amorphous compounds. characteristic of thermal stability suggest that these bands
+
According to the XRD data (Table 3) Cu–Zn–Al catalysts can be attributed to the stretching mode of CO in Cu –
with copper loading even of 20% after calcination have not CO complexes (31, 32). As for Cu–Zn catalysts, the Cu–
exhibited bulk CuO, but after reduction of this catalyst a Zn–Al catalysts with copper loading equal to and more
trace amount of metallic copper was observed in the X- than 40% completely lost transparency in the wavenum-
�
1
�
ray diffraction patterns. Calcination of the catalysts with ber range 2000–2300 cm after reduction in H2 at 250 C.
0% copper and higher led to a mixture of CuO and ZnO According to our previous discussion for Cu–Zn this result
Table 3). The increase of the copper loading led to an in- is probably because of the weak interaction of copper with
4
(
crease in the CuO. The diameter of metallic copper crystal- ZnO in Cu–Zn–Al catalysts with copper equal to 40% and
lites increased significantly (Table 3) in Cu–Zn–Al catalysts greater.
with copper loading equal to 80% and more, and this fact
is indirect evidence for a significant decrease in the surface practically invariant with the catalyst composition. XP
area of copper.
spectra of Cu (2p3/2) of Cu–Zn–Al catalysts with 15%
X-ray photoelectron spectra of the Cu(2p3/2) level were
The IR spectra of CO adsorbed on the reduced surface of copper before and after reduction in an atmospheric hy-
Cu–Zn–Al catalysts with concentrations of 5, 15, and 20% drogen stream are shown in Fig. 2. This spectra indicates a
�
1
Cu displayed one symmetric band centered at 2110 cm
well-separated main peak at 934.5 eV and satellite peak at
�
1
2+
for the sample with 5% Cu and at 2115 cm for samples 941.5 eV, characteristic of Cu (Fig. 2a). The satellite peak
TABLE 3
Properties of the Cu–Zn–Al Catalysts before and after Use in the Reaction of Dehydrogenation of Cyclohexanol
Catalyst
Surface ratio of the atom
composition
(Cu/Zn/Al)
(at.% )
BET
Cu crystallite
size (nm)
d = 0.2088
Phase composition; XRD
Fresh After work
(Cu/Zn); XPS
Number
of catalyst
surface area
2
(m /g)
Fresh
After work
1
2
3
4
5
6
7
5.0/80.0/15.0
15.0/70.0/15.0
20.0/65.0/15.0
40.0/45.0/15.0
60.0/25.0/15.0
80.0/5.0/15.0
100
87.0
77.0
73.0
79.0
65.0
56.0
15.0
ZnO
ZnO
ZnO
ZnO, CuO traces
ZnO, CuO
CuO
ZnO
ZnO
ZnO, Cu traces
ZnO, Cu
ZnO, Cu
Cu
—
—
0.068
0.170
0.290
0.625
3.545
10.700
0.096
0.160
0.320
0.523
1.150
1.983
11.2
13.2
14.9
28.6
31.5
CuO
Cu

DEHYDROGENATION OF CYCLOHEXANOL
25
of Cu(2p3/2) band characteristic of Cu2 (36) ions disap- the MgO system leads to a gradual increase in the activity
peared completely after reduction and the binding energy for the synthesis of ketone from alcohol (Fig. 3I, a). The
level shifted to a lower energy (932.1 eV), indicating unam- rate of this reaction per unit of the mass for Cu–Mg cata-
+
biguously the complete reduction of Cu2 to lower valence lysts achieved a maximum at a copper loading of 52% . The
+
states (Fig. 2b).
subsequent increase of copper in Cu–Mg catalysts reduced
In this work we could not determine the concentration the rate of the reaction of cyclohexanol dehydrogenation
of aluminum on the surface of the Cu–Zn–Al catalysts be- (Fig. 3I, a). In contrast to results for Cu–Mg catalysts the
cause of the low intensities Al2p and Al2s lines. This result addition of even 8% copper to ZnO and 5% to Al-Zn sys-
does not seem surprising, because of the low loading of tems resulted in a sharp increase in cyclohexanone synthesis
Al (15 at.% of aluminum corresponds to 6.2 wt% ). Sim- (Fig. 3I, b and c). The maximum rates of this reaction were
ilar results have been obtained by other authors (16) for observed for Cu–Zn catalysts at 30% copper loading and
Cu–Zn–Al catalysts with Al loading of 18 and 17 at.% .
for Cu–Zn–Al catalysts at 20% copper loading. For all sys-
Surface compositions derived from XPS intensities are tems the minimum rate of the reaction of cyclohexanol de-
reported in the Table 3. The Cu/Zn ratio as a function of hydrogenation per gram of catalyst was observed for pure
copper loading increases monotonically with an increase in metallic copper. It should be noted that the dependence of
copper catalysts. The comparison of Cu/Zn ratio for calci- the rate of Reaction 1 on Cu–Mg and on Cu–Zn-containing
nated and reduced samples indicates low level stability of catalysts totally differs (Fig. 3I). Similar results were ob-
copper in catalysts with copper greater than 40% of Cu, served for the activity of these catalysts calculated per unit
since the Cu/Zn ratio after reduction of these catalysts de- of the catalyst surface (Fig. 3II), and only the position of
creased significantly.
pure copper is different. The activity of pure copper per
square meter is twice than that of catalysts with more than
Catalyst Activity
5
0% copper loading (Fig. 3II). This observation suggests
that the sufficient part of the surface of these catalysts even
with 80% copper loading is contributed by the surface of
inactive support.
Dehydrogenation of cyclohexanol to cyclohexanone.
The catalyst activity of Cu–Mg, Cu–Zn, and Cu–Zn–Al
catalysts in Reaction 1 is presented as a rate of reaction
per gram of catalyst (Fig. 3I) and a rate of the reaction per
Aromatization of cyclohexanol to phenol. The activities
specific area of catalyst (Fig. 3II). As follows from Fig. 3 of the catalysts in the cyclohexanol aromatization reaction
the initial MgO, ZnO, and Zn–Al systems do not catalyze also has been calculated per unit of the mass and per unit of
cyclohexanol dehydrogenation. The addition of copper to the specific area of the catalysts (Figs. 4I and 4II). As follows
FIG. 3. The rate of the dehydrogenation reaction of cyclohexanol to cyclohexanone per unit of the mass (mol/g � min) (I) and per unit of the
2
�
specific area (mol/m � min) (II) as a function of copper loading (T = 250 C): (a) Cu–Mg catalysts, (b) Cu–Zn catalysts, (c) Cu–Zn–Al catalysts.

2
6
FRIDMAN AND DAVYDOV
FIG. 4. The rate of the aromatization reaction of cyclohexanol to phenol per unit of the mass (mol/g � min) (I) and per unit of the specific area
2
�
(
mol/m � min) (II) as a function of copper loading (T = 300 C): (a) Cu–Mg catalysts, (b) Cu–Zn catalysts, (c) Cu–Zn–Al catalysts.
from these figures the activity of the Cu–Mg catalysts in the the surface structure of Cu–Mg catalysts led to the iden-
cyclohexanol aromatization reaction depends on amount of tification of isolated Cu2 ions in open octahedral holes
copper in catalysts (Figs. 4I, a, and 4II, a). Cu–Mg catalysts and evidence of the migration of isolated copper ions to
with 52% Cu showed a maximum in activity per unit of the form metallic clusters (33). Detailed ESR and UV spectro-
mass in Reaction 2 (Fig. 4I, a). The increase of copper in scopic studies (37) have shown the presence of isolated and
+
Cu–Mg catalysts greater than 52% reduces the rate of the strongly associated Cu2 ions in initial samples. The results
+
reaction of cyclohexanol aromatization (Fig. 4I, a).
of the present investigation have shown that in catalyst with
Phenol is not formed on the Cu–Zn and Cu–Zn–Al cata- 10% copper only a very small amount of CuO is observed
lysts with a concentration of copper less than 20% (Fig. 4I,b after calcination. Apparently the remaining copper is ei-
and c). A very small amount of phenol is formed on the ther in an X-ray amorphous phase or in a solid solution
these catalysts with 20% copper. The maximum activity CuO in MgO (39). However, after reduction, even in cata-
2+
calculated per unit of the catalyst mass for Cu–Zn catalysts lyst with 10% copper loading, both isolated ions Cu in a
in Reaction 2 is observed at 30% copper loading and for solid solution of MgO and strongly associated Cu2 were
+
Cu–Zn–Al catalysts at 40% copper loading. Pure metallic reduced to metallic copper in an atmosphere of hydrogen
�
copper also catalyzes the reaction of aromatization of cy- at 250 C (Table 1). We have considered the possibility that
clohexanol to phenol.
with Cu–Mg catalysts, the solid solution of copper in MgO
Similar results were observed for activity of these cata- could be reduced also to monovalent copper, as in the case
lystscalculated per unit ofthe catalyst surface (Fig. 4II), par- with the Cu–Zn catalysts (14, 15), at least in catalysts with
ticularly for the catalysts with less than 40% copper loading. a low loading of copper where the formation of the solid
The activities calculated per unit of the specific area for the solution of copper in a support is most probable. The re-
all catalyst system, practically, do not change in the range sults of IR spectroscopy of adsorbed CO have shown that
of Cu loading from 40 to 80% . The maximum rate of Reac- copper is reduced to a metallic state independently of the
tion 2 per square meter is observed for pure metallic copper composition of Cu–Mg catalysts. Apparently, the magne-
(
Fig. 4II, a, b and c.).
sium oxide does not have the ability to inhibit the complete
reduction of copper oxide to the metallic state in contrast
to ZnO (15). The possible lowering of the effective charge
of copper ions under the specific influence of basic magne-
DISCUSSION
Cu–Zn and Cu–Zn–Al catalysts have been investigated sium oxide has been attributed to the presence of electron-
by a gamut of physical techniques, but only a few studies donor ligands in the coordination sphere of a cation (31).
have been devoted to the investigation of Cu–Mg catalysts The difference in electron-donor ability between MgO and
(
33, 37–39). An earlier attempt to correlate the activity with ZnO accounts for the difference in the influence of these

DEHYDROGENATION OF CYCLOHEXANOL
27
supports for the reduction of copper. Thus, copper on the stagesin the preparation ofCu–Zn and Cu–Zn–Alcatalysts.
reduced surface of all Cu–Mg catalysts exists only in the Based on the conclusions of previous studies, they found a
0
oxidation state Cu , and the copper loading influences only way to increase the solubility of copper in ZnO by anion
a number of the metallic copper atoms and active sites on modification to enhance the amount of defects in the lattice
the surface. Actually, these results and activity data for the of ZnO (44–46). A solid solution of copper ions in zinc oxide
Cu–Mg catalysts strongly support zerovalent copper as an modified with anions was established as an oxide precursor
active site of the dehydrogenation of cyclohexanol. Con- of the active state (18). In the one of these studies (47),
siderable attention to the study of Cu–Zn and Cu–Zn–Al Plasova et al. showed that the reduction of a solid solution
catalysts (11–22) wasmostly due to the practical importance of Cu in ZnO carries out through the formation phase of
of using these catalysts for the synthesis of methanol. One metallic copper which very easily vanishes when the H2 at-
of the first investigations of these catalysts was performed mosphere is replaced with He. They have explained this ob-
by Herman et al. (14) and Menta et al. (15). The compo- servation by the formation of a complex proton-stabilized
sitional profile of calcined Cu–Zn catalysts, determined by system of ZnO modified by anions with epitaxially bonded
0
+
2+
X-ray diffraction (14), showed that a part of copper is in copper species [Cu � H Cu
Zn0.9O] (47).
x
2x
0.1� x
the solid solution of CuO in ZnO. According to X-ray mi-
The compositional profile of Cu–Zn and Cu–Zn–Al
croanalysis (15) up to 6% of copper after calcination exists catalysts investigated in the present study shows that in low
in the solid solution. Consequently, after reduction of the content catalysts, copper exists either in an amorphous state
oxides, fine dispersions of metallic copper appear in the 10– or in the solid solution of CuO in the lattice of ZnO (no
3
0% copper catalysts. Herman et al. (14) have shown that CuO phase was found for copper loading to 20% for Cu–
the reduction of catalysts leads to an increase in the amount Zn–Al and Cu–Zn catalysts), and that only in catalyst with
of the solid solution of monovalent copper in the lattice of 20% after reduction, traces of metallic copper have been
+
+
ZnO. Authors (14, 15) have explained that dissolution Cu
found. Based on the IR spectroscopy data of the present
study showing that the main oxidation state of copper on
2+
in ZnO is much more favorable than Cu because Cu
2+
+
is isoelectronic with Zn . The limited solubility of Cu in the surface of low content Cu–Zn and Cu–Zn–Al catalysts
ZnO is based on the requirement of electroneutrality, ei- is Cu (Fig. 2) and conclusions of the Klier et al. studies
+
ther by oxygen vacancies or by intersticial cations, upon (14, 15), we propose that only a small part (an negligible
substitution of some Zn2 by Cu (14). The catalysts with a amount) of copper in these catalysts is amorphous copper
composition close to 30/70 are characterized as a complex of oxide. The majorityofcopper in these catalystsisin the solid
two forms of copper, one appearing in and strongly interact- solution of copper in ZnO. The increase of copper loading
+
+
+
ing with the zinc oxide phase (solid solution of Cu in ZnO) in catalysts leads to the reduction in the amount of ZnO
and the other being a fine copper metal dispersion (14, 15). in the system and to the decrease of ZnO, which is able to
The Herman-Klier model of copper-zinc-contained cata- accept copper with the formation of the solid solution. Con-
lysts has been accepted by many researchers to understand sequently, copper in Cu–Zn and Cu–Zn–Al catalysts with
their own results on these catalysts. Later Klier et al. (40) Cu greater than 20% forms a distinct phases of copper ox-
estimated that the average copper concentration of 15% as ide, which coexists with the solid solution of copper in ZnO.
a solid solution in the ZnO phase is present in a 30/70 cata- Although we did not determine the oxidation state of cop-
lyst (50% of total Cu amount); they considered this number per on the surface of Cu–Zn and Cu–Zn–Al catalysts with
a limit of Cu solubility in ZnO. Similar results have been copper greater than 30 and 20% , respectively, because of
obtained by Okamoto et al. (21, 22), who found that two dis- the loss of IR transparency, we have no doubt concerning
tinct types of Cu metal species were formed on the catalyst the oxidation state of copper on the surface of these cat-
0
surface, and their proportions depend on the catalyst com- alysts. Since Cu is the most thermodynamically beneficial
�
position. In high Cu content catalysts (>25 wt% CuO), Cu state of copper for the CuO reduction by hydrogen at 250 C
metalparticlesare predominant, whereasin the lowCu con- (our own thermodynamic calculations), we propose that the
0
tent catalysts (<10 wt% CuO), a Cu metal species, which main oxidation state of copper in these samples is Cu . This
is best characterized by two-dimensional epitaxial mono- confirmation is supported by indirect evidence from XRD,
layer over ZnO, is predominant. It was concluded that Cu XPS data of the present investigation, and conclusions of
0
metal particles with charge Cu on the surface are formed the different investigators, who have studied Cu–Zn and
mainly from crystalline and an amorphous copper oxide Cu–Zn–Al catalysts by Auger electron spectroscopy (22)
0
+
phase, while two-dimensional Cu species Cu –Cu includ- and IR spectroscopy (48). Thus, in the case where the cop-
+
2+
ing Cu are derived from the Cu dissolved in the ZnO per oxide does not have any inhibiting influence of other
lattice (22). Gasbassi and petrini (16, 17) obtained results components for the reduction of CuO, there is no possi-
which showed the formation of the solid solution of CuO bility for the formation of copper in the oxidation state
in the lattice of ZnO in low content Cu–Zn catalysts. Con- another than metallic copper. Based on the limit of the sol-
currently, Yur’eva et al. (41–43) studied the mechanisms of ubility of copper in ZnO proposed by Dominiquez et al.

2
8
FRIDMAN AND DAVYDOV
Finally, we concluded that copper-containing catalysts of
dehydrogenation of cyclohexanol have on the surface two
kinds of active copper sites for Reaction 1 such as mono-
+
0
valent copper (Cu ) and zerovalent copper (Cu ), which
work independently. The ratio between these sites depends
on the composition and conditions of catalyst preparation.
The Cu–Mg catalysts do not have sites of monovalent cop-
per, and the activity of these catalysts is a function of the
metallic copper surface. The increase of the activity in the
cyclohexanol dehydrogenation reaction to cyclohexanone
of Cu–Mg catalyst happens with enhancement of copper
loading and the surface metallic copper in catalysts to at-
tain the maximum, and a subsequent decrease of the rate
is attributed to a reduced dispersion and surface area of
metallic copper.
In the case of copper-zinc-containing catalysts in the low
Cu content range, the increase in copper loading leads to
an increase in the amount of the solid solution of Cu in
ZnO, and, as result, to an increase of both the amount of
FIG. 5. Assumable phase composition of Cu–Zn catalyst, calculated
based on the Herman-Klier model of solubility of copper in ZnO in Cu–Zn monovalent copper sites and the activity of catalysts. Be-
catalysts: (a) solid solution of copper in ZnO; (b) phase of CuO.
ginning from 15% composition and greater the enhanced
catalyst activity results in the formation of both kinds of
(
40) and the assumption that the conditions of the prepara- copper active sites: monovalent copper and fine dispersion
tion of catalysts provide the formation of solid solution of metallic copper. Monovalent copper is the predominant ac-
Cu in ZnO equal to the limit solubility we have calculated tive site in copper-zinc-containing catalysts up to 20–30% .
the assumable phase composition (concentrations of the Beginning at 15–20% copper the activity of the catalysts
solid solution of copper in ZnO and distinct phase of in this range changes insignificantly, probably because the
CuO) of Cu–Zn and Cu–Zn–Al catalysts as a function of small reduction of the amount of monovalent copper sites
Cu/Cu + Zn ratio (Fig. 5). It should be noted that results is substituted by the formation of a large number of metallic
of our calculation of the phase composition Cu–Zn cata- copper sites (for instance, 20 and 40% Cu–Zn–Al catalyst
lysts are in very good agreement with those of Okamoto Fig. 3I, c). It is likely that the catalysts with 60% copper and
et al. (21) and our X-ray data (Table 3). According to this greater do not have monovalent copper on the surface at
calculation, the amount of the solid solution of copper in all, and the activity of these catalysts is a function only of the
ZnO at first increases sharply up to 15% (Fig. 5a) and then copper surface area. The shift of the maximum activity of
gradually decreases with an increase of the copper content copper-zinc-containing catalysts toward the low Cu content
greater than 20–30% . Comparisons of the calculated phase range suggests that the activity of single site monovalent
composition (Fig. 5, a) and the activity of Cu–Zn and Cu– copper is higher than the activity of single site zerovalent
Zn–Al catalysts (Fig. 3I, b and c) as a function of copper copper.
are parallel to each other. These data indicate that the solid
The dependence rate of the aromatization reaction of cy-
solution of Cu in lattice of ZnO is a precursor of at least clohexanol indicates that phenol is not formed on copper-
one kind of active site for cyclohexanol dehydrogenation zinc-containing catalysts with copper less than 20% . This
to cyclohexanone. On the other hand, the comparison of observation is evidence that monovalent copper sites are
the calculated amount of the CuO phase (Fig. 5, b) and the not the active sites of the reaction of the phenol formation.
rate of Reaction 2 (Fig. 4I, b and c) with the composition The formation of phenol is observed only after the appear-
of these catalysts indicates that the formation of phenol ance of metallic copper. The rate of Reaction 2 attains the
begins after the appearance of the distinct CuO phase in maximum at a copper content of 40% for Cu–Zn–Al cata-
Cu–Zn and Cu–Zn–Al catalysts. This result suggests that lysts and the subsequent increase of copper leads to a slowly
the amorphous and crystal phases of CuO are precursors of decreasing rate of Reaction 2. The reason for this decrease
the active sites for the reaction of aromatization of cyclo- on catalysts with 60% Cu is the decrease of the dispersion
hexanol to phenol. Since Cu–Zn-containing catalysts with and the surface area of copper. It should be noted that for
low Cu loading have on their surface only copper in the Cu–Mg catalysts the dependencies of rates of Reaction 1
+
Cu oxidation state, and they do not form phenol from cy- and Reaction 2 on composition are parallel, and this fact
clohexanol, we concluded that monovalent copper sites are is additional evidence that in these catalysts both reactions
not able to catalyze the aromatization reaction.
occur via the same active sites.

DEHYDROGENATION OF CYCLOHEXANOL
29
To compare the activities of monovalent and zerovalent anol to phenol. Two kinds of copper active sites of the re-
copper sites the activity data for pure metallic copper and action of cyclohexanol dehydrogenation to cyclohexanone
for Cu–Zn–Al catalyst with 5% Cu have been used, since have been revealed (monovalent copper and metallic cop-
there is apparently only one kind of copper site on each of per), and it was shown that the sites of monovalent cop-
them (monovalent copper on the surface of 5% Cu–Zn–Al per are significantly more active than the sites of metallic
catalyst and zerovalent copper on the surface of reduced copper. Moreover, the sites of monovalent copper are se-
CuO). The catalysts of middle compositions were not used lective, and they do not catalyze the reaction of the phenol
because of a higher probability for the existence of two formation. At the same time, the sites of metallic copper are
kinds of copper sites on the surface of these catalysts.
active sites not only for the reaction of dehydrogenation of
Several techniques have been developed and applied for alcohol to ketone but also for the reaction of aromatization
the evaluation of the specific metal surface of copper cata- of cyclohexanol to phenol; therefore, these sites are not
lysts, such as chemisorption of carbon monoxide (49, 50), selective.
hydrogen (51), and oxygen (52) and decomposition of ni-
trous oxide (53). In a recent study Dandekar and Vannice
ACKNOWLEDGMENT
have used the combined approach with irreversible CO ad-
sorption and N2O decomposition measurements (54, 55)
The authorsare gratefulto Professor T. Yur’eva (Boreskov’s Institute of
Catalysis, Novosibirsk, Russia) for consultations concerning the methods
for determination of the dispersion of supported Cu cata-
lysts. The crystallite sizes obtained from these estimates
were compared to those obtained from TEM and XRD
measurements and were found to be in very good agree-
ment (54). In our investigation we have used the values of
the total specific surface, obtained by BET, as well as the
ratio between atoms on the surface, obtained by XPS. BET
has been used by Herman et al. (14) and Okamoto et al.
of catalyst preparation.
REFERENCES
. Bedina, L. N., Polovkina, T. S., Fridman, V. Z., and Voronkova, O. N.,
Chem. Ind. (in Russian) 9, 5 (1987).
. Zadymov, V. V., Smirnov, I. V., and Topchieva, K. V., J. Phys. Chem.
1
2
(
in Russian) 54, 711 (1980).
3. Fridman, V. Z., Bedina, L. N., and Petrov, I. Y., Kinet. Catal. (in
Russian) 29, 621 (1988).
(
21) to estimate the surface of reduced CuO.
4. Kozlov, N. S., Mostovai, L. I., Sedel’zeva, M. A., and Gypalo, V. I.,
Kinet. Catal. (in Russian) 27, 244 (1986).
According to XPS data the Cu–Zn–Al catalysts have on
the surface only atoms of Cu and Zn and the surface con-
centration of Al is equal to zero (Table 3). On the other
hand, monovalent copper in catalysts with 5% copper ex-
ists on the surface and in the bulk in only the solid solution
5. Nikiforova, N. N., and Gavrenko, K. A., Petrochemistry (in Russian)
14, 116 (1974).
6. Orizarsky, I. V., Petrov, L. A., and Petrova, V. M., React. Kinet. Catal.
Lett. 17, 427 (1981).
7. Chang, H. F., and Saleque, M. A., Appl. Catal. A 103, 233 (1993).
+
of Cu in ZnO. The observation that monovalent copper
8
9
. Jeon, G. S., and Chung, J. S., Appl. Catal. 115, 29 (1994).
. Chen, W. S., and Lee, M. D., Appl. Catal. 83, 201 (1992).
substitutes zinc in lattice ZnO manifests the correctness of
+
the simplification that one atom of Cu and one atom of
10. Chang, H. F., Saleque, M. A., Hsu, W. S., and Lin, W. H., J. Mol.
2+
Zn in lattice of ZnO form equal surface areas since they
have exactly the same crystal structure and surrounding.
The results of calculations have shown that the copper sur-
face areas in the 5% Cu–Zn–Al catalyst and pure metallic
Catal. 109, 249 (1996).
1
1
1. Monnier, J. R., Hanrahen, M. J., and Apai, G., J. Catal. 92, 119 (1985).
2. Shimomura, K., Ogawa, K., Ora, M., and Kotera, M., J. Catal. 52,
191 (1978).
1
1
3. Fleisch, T. H., and Mievilie, R. L., J. Catal. 90, 165 (1984).
4. Herman, R. G., Klier, K., Simmons, G. W., Finn, B. P., and Bulko, J. B.,
J. Catal. 56, 407 (1979).
2
copper are equal to 5.7 and 11.7 m /g, respectively. The
numbers of monovalent and zerovalent sites were obtained
based on the valuesofthe copper surface areasofthese cata-
15. Menta, S., Simmons, G. W., Klier, K., and Herman, R. G., J. Catal.
+
18 � 2
57, 339 (1979).
lysts and assumptions of a Cu site density of 5.2 � 10 m
1
1
1
6. Garbassi, F., and Petrini, G., J. Catal. 90, 106 (1984).
7. Petrini, G., and Garbassi, F., J. Catal. 90, 113 (1984).
8. Bailie, J. E., Rochester, C. H., and Millar, G. J., Catal. Lett. 31, 333
(1995).
0
19
� 2
(
56) and of a Cu site density of 1.4 � 10
m
(57). Activ-
+
� 1
ities as well as turnover frequencies for Cu = 1.2 s and
0
� 1
Cu = 0.086 s sites have been calculated. The result indi-
+
19. Plyasova, L. M., Yur’eva, T. M., Kriger, T. A., Makarova, O. V.,
Zaikovskii, V. I., Solov’eva, L. P., and Shmakov, A. N., Kinet. Catal.
cates that the activity of the Cu site is 14 times greater that
0
of than Cu .
(
in Russian) 36, 464 (1995).
20. Chiotti, G., and Boccuzzi, F., Catal. Rev.-Sci. Eng. 29, 151 (1987).
CONCLUSIONS
21. Okamoto, Y., Fukino, K., Imanaka, T., and Teranlchi, S., J. Phys.
Chem. 87, 3740 (1983).
2
2. Okamoto, Y., Fukino, K., Imanaka, T., and Teranlchi, S., J. Phys.
Chem. 87, 3747 (1983).
3. Sivaraj, C., Reddy, B. M., and Rao, P. K., Appl. Catal. 45, L11 (1988).
4. Sivaraj, C., Srinivas, S. T., Rao, V. N., and Rao, P. K., J. Mol. Catal. 60,
L23 (1990).
The X-ray diffraction, X-ray photoelectron spectroscopy,
IR spectroscopy of chemisorbed carbon monoxide and ki-
netic data have been employed to determine the active sites
of the reaction of dehydrogenation of cyclohexanol to cy-
2
2
clohexanone and the reaction of aromatization of cyclohex- 25. Sivaraj, C., Reddy, B. M., and Rao, P. K., J. Mol. Catal. 47, 17 (1988).

3
0
FRIDMAN AND DAVYDOV
2
6. Fridman, V. Z., Michal’chenko, E. D., Tryasunov, B. G., Ziborov, 40. Dominiquez, J. M., Simmons, G. W., and Klier, K., J. Mol. Catal. 20,
A. V., and Plyasova, L. M., Kinet. Catal. (in Russian) 32, 922
1991).
7. Parris, G. E., and Klier, K., J. Catal. 97, 374 (1986).
369 (1983).
(
41. Yur’eva, T. M., and Minyukova, T. P., React. Kinet. Catal. Lett. 29, 55
2
2
(1985).
8. Cullity, B. D., “Elements of X-Ray Diffraction.” Addison-Wesley, 42. Kuznetsova, L. I., Yur’eva, T. M., and Minyukova, T. P., React. Kinet.
Reading, MA, 1956.
Catal. Lett. 19, 355 (1982).
2
3
9. Wagner, C. D., Davis, L. E., Zeller, M. V., Taylor, J. A., and Raymood,
R. H., Surf. Interface Anal. 3, 211 (1981).
0. Shecochihin, U. M., and Davydov, A. A., “Methodics of IR Studies in
Physicochemistry of Semiconductor Surface (VINITY),” N. 5014-83,
pp. 1–110. Moscow, 1983.
43. Yur’eva, T. M., Kinet. Catal. (in Russian) 26, 686 (1985).
44. Ketchik, S. V., Yur’eva, T. M., and Plysova, L. M., Izv. Sib. Otd. Akad.
Nauk. USSR (in Russian) 6, 109 (1983).
45. Ketchik, S. V., Yur’eva, T. M., and Plysova, L. M., Izv. Sib. Otd. Akad.
Nauk. USSR (in Russian) 1, 36 (1984).
3
3
3
3
3
3
3
3
3
1. Davydov, A. A., “Infrared Spectroscopy of Adsorbed Species on the
Surface of Transition Metal Oxides.” (Chichester), New York, 1990.
2. Gallardo Amores, J. M., Sanchez Estribano, V., Busca, G., and
Loreazelli, V. J., J. Mater. Chem. 4, 965 (1994).
3. Maksimov, N. G., Chigrina, V. A., Boreskov, G. K., Anufrienko,
V. F., and Yur’eva, T. M., Kinet. Catal. (in Russian) 13, 446 (1972).
4. Boccuzzi, F., Ghiotti, G., and Chiorino, A., J. Chem. Soc., Faraday
Trans. 79, 1779 (1983).
5. Boccuzzi, F., Morterra, C., Scala, R., and Zecchina, J., J. Chem. Soc., 50. Smith, J. S., Thrower, P. A., and Vannice, M. A., J. Catal. 68, 270 (1981).
Faraday Trans. 77, 2059 (1981).
6. Shinkarenko, V. G., and Anufrienko, V. F., Theor. Exp. Chem. (in
Russian) 12, 270 (1976).
7. Boudart, M., Derouane, E. G., Indovina, V., and Walters, A. B.,
J. Catal. 39, 115 (1975).
46. Ketchik, S. V., Yur’eva, T. M., and Plysova, L. M., Izv. Sib. Otd. Akad.
Nauk. USSR (in Russian) 4, 37 (1984).
47. Plysova, L. M., Yur’eva, T. M., Kriger, T. A., Makarova, O. V.,
Zaikovskii, V. I., Solov’eva, L. P., and Schmakov, A. N., Kinet. Catal.
(in Russian) 36, 425 (1995).
48. Bailey, S., Froment, G. F., Snoeck, J. W., and Waugh, K. C., Catal. Lett.
30, 99 (1995).
49. Cormach, D., and Moss, R. L., J. Catal. 13, 1 (1969).
51. Spenadel, L., and Boudart, J., J. Phys. Chem. 64, 205 (1960).
52. Vasilevich, A. A., Shapiro, G. P., Alekseev, A. M., Semenova, T. A.,
Markina, M. I., Vasileva, T. A., and Budkina, O. G., Kinet. Catal.
(in Russian) 16, 1363 (1975).
53. Dvorak, G., and Pasek, J., J. Catal. 18, 108 (1970).
8. Davydov, A. A., Rubene, N. A., and Budneva, A. A., Kinet. Catal. 54. Dandekar, A., and Vannice, M. A., J. Catal. 178, 621 (1998).
(
in Russian) 19, 969 (1978).
9. Davydov, A. A., Prudnikova, O. Yu., and Yur’eva, T. M., Kinet. Catal.
in Russian) 30, 477 (1989).
55. Dandekar, A., Baker, T. K., and Vannice, M. A., J. Catal. 183, 131
(1999).
56. Kington, G. L., and Holmes, J. M., Trans. Faraday Soc. 49, 417 (1953).
(