Liquid-phase oxidation of alcohols by oxygen and nitrous oxide catalysed by Ru-Co oxide

Article abstract of DOI:10.1016/j.jcat.2005.01.006

Chemoselective catalysts in bulk or supported on γ-Al ₂O₃ binary oxides Ru^IV-Co^III (Ru/Co = 1:1-1:2), prepared by co-precipitation, were used for liquid-phase oxidation of saturated and unsaturated primary and secondary alcohols to aldehydes and ketones with O₂ or N₂O. The catalysts can be separated by filtration and reused. No leaching of Ru or Co in solution was observed. The oxidation is enhanced by the presence of hydration water in the Ru-Co catalyst, which indicates the participation of active Ru^IV hydroxo species in the reaction. From XRD and TGA, the Ru-Co oxide can be approximated as a hydrous binary oxide comprising the amorphous RuO₂ and heterogenite-3R cobaltic acid CoO(OH). The alcohol oxidation appears to occur by a nonradical mechanism, which may be viewed as an oxidative dehydrogenation of alcohols to form an aldehyde or ketone. H₂-TPR shows that Co^III practically does not affect the oxidising ability of RuO₂. This suggests that the cobalt is likely to enhance catalyst reoxidation by O ₂ rather than to play a significant role in the alcohol dehydrogenation. The alcohol oxidation by N₂O exhibits a close similarity to the oxidation by O₂ but is much less efficient. Much more active catalysts are required to make the oxidation with N₂O synthetically useful.

Full text of DOI:10.1016/j.jcat.2005.01.006

Journal of Catalysis 231 (2005) 41–47
www.elsevier.com/locate/jcat
Liquid-phase oxidation of alcohols by oxygen and nitrous oxide catalysed
by Ru–Co oxide
Tatiana L. Stuchinskaya, Mehdi Musawir, Elena F. Kozhevnikova, Ivan V. Kozhevnikov ∗
Department of Chemistry, University of Liverpool, Liverpool, L69 7ZD, UK
Received 9 November 2004; revised 23 December 2004; accepted 10 January 2005
Abstract
IV
III
Chemoselective catalysts in bulk or supported on γ -Al O binary oxides Ru –Co (Ru/Co = 1:1–1:2), prepared by co-precipitation,
2
3
were used for liquid-phase oxidation of saturated and unsaturated primary and secondary alcohols to aldehydes and ketones with O or N O.
2
2
The catalysts can be separated by filtration and reused. No leaching of Ru or Co in solution was observed. The oxidation is enhanced by
IV
the presence of hydration water in the Ru–Co catalyst, which indicates the participation of active Ru hydroxo species in the reaction.
From XRD and TGA, the Ru–Co oxide can be approximated as a hydrous binary oxide comprising the amorphous RuO and heterogenite-
2
3R cobaltic acid CoO(OH). The alcohol oxidation appears to occur by a nonradical mechanism, which may be viewed as an oxidative
III
dehydrogenation of alcohols to form an aldehyde or ketone. H -TPR shows that Co practically does not affect the oxidising ability of
2
RuO . This suggests that the cobalt is likely to enhance catalyst reoxidation by O rather than to play a significant role in the alcohol
2
2
dehydrogenation. The alcohol oxidation by N O exhibits a close similarity to the oxidation by O but is much less efficient. Much more
2
2
active catalysts are required to make the oxidation with N O synthetically useful.
2
2005 Elsevier Inc. All rights reserved.
Keywords: Alcohol; Oxidation; Oxygen; Nitrous oxide; Ruthenium oxide; Cobalt
1. Introduction
method for the aerobic oxidation. Typically, the aerobic ox-
idation of alcohols involves the use of catalysts based on
platinum-group metals [1,2]. Supported platinum and palla-
dium catalysts have long been used for alcohol oxidation [2].
More recently, ruthenium catalysts have attracted significant
interest. These involve soluble complexes or solid catalysts
[1,2,5, and references therein]. Most efficient Ru-based het-
erogeneous systems for the aerobic oxidation of alcohols in
liquid phase include recently developed Ru–Co–Al hydrotal-
cite [6], Ru-hydroxyapatite [7], Ru–Co(OH)2–CeO2 [8], and
Ru/Al2O3 [9]. Frequently, Ru catalysts are sufficiently selec-
tive to avoid overoxidation of aldehydes to acids and tolerant
toward many other functional groups, including C=C dou-
ble bonds, that may be present in alcohol molecules [2].
Recently we reported that hydrous RuIV–MIII and PdII–
MIII binary oxides (M = CoIII, FeIII, and MnIII) are active
solid catalysts for the oxidation of primary alcohols to alde-
hydes with O2 [10–12] or N2O [13] in liquid phase, where
the RuIV–CoIII oxide is the most efficient catalyst. These cat-
The catalytic conversion of primary saturated and unsat-
urated alcohols to aldehydes is essential for the preparation
of fragrances, food additives, and many organic intermedi-
ates [1,2]. Traditional methods for the synthesis of aldehydes
involve the use of stoichiometric amounts of inorganic oxi-
dants (e.g., CrVI) and generate large quantities of waste. The
development of effective catalytic aerobic oxidation of alco-
hols with the use of environmentally benign and inexpensive
oxidants such as oxygen or air is an important challenge
[1,2]. Nitrous oxide is a potentially interesting oxidant for
the clean oxidation of organic substrates, as the only by-
product in these reactions is dinitrogen [3,4]. Heterogeneous
catalysis is generally considered to be the most attractive
*
Corresponding author. Fax: +44(0)151 794 3589.
E-mail address: kozhev@liverpool.ac.uk (I.V. Kozhevnikov).
0021-9517/$ – see front matter
doi:10.1016/j.jcat.2005.01.006
2005 Elsevier Inc. All rights reserved.

42
T.L. Stuchinskaya et al. / Journal of Catalysis 231 (2005) 41–47
alysts are robust, purely inorganic materials resistant to ox-
idative degradation, in contrast to organometallic catalysts.
The aim of this work is to investigate in detail the oxidation
of saturated and unsaturated primary and secondary alcohols
to aldehydes and ketones with O2 and N2O in liquid phase
with bulk and supported RuIV–CoIII mixed oxide catalysts.
The catalyst testing is complemented by mechanistic stud-
ies and catalyst characterisation by XRD, TGA, XPS, and
H2-TPR.
calculated from the measured XPS peak area corrected for
intrinsic sensitivity and spectrometer transmission factors.
The metal content in catalyst samples was measured by ICP
analysis. Temperature-programmed reduction (TPR) of cat-
alysts was carried out on a Micromeritics 2900 TPD/TPR
apparatus equipped with a thermal conductivity detector.
Catalyst samples (20–30 mg) were heated to 500 ◦C at a rate
of 10 ◦C/min in a H2–Ar (5:95) gas flow (60 cm3/min).
2.3. Catalyst preparation
2. Experimental
Bulk hydrous Ru dioxide was prepared by precipitation
from a 0.1 mol/l aqueous solution of RuCl3 with 1 mol/l
NaOH at pH 10 and 65 ◦C. The RuIV–CoIII binary oxides
were prepared similarly by co-precipitation of 0.1 mol/l
RuCl3 solutions containing appropriate amounts of CoCl2.
Supported catalysts were made by (co)precipitation by aque-
ous NaOH in the presence of a support such as acidic, neu-
tral, or basic γ -alumina (Al2O3-a, Al2O3-n or Al2O3-b) or
SiO2. The suspensions were aged with stirring for 2 h (24 h
for supported catalysts), filtered off, washed with water un-
til Cl− was removed (AgNO3 test; [Cl−] < 10−8 mol/l in
washings), and finally dried at 60 ◦C/0.5 Torr for 2 h, un-
less stated otherwise. During the preparation, RuIII and CoII
underwent aerobic oxidation to RuIV and CoIII [14], respec-
tively. Catalyst characterisation is given in Table 1.
2.1. Materials
Solvents, chemicals, and catalyst supports were pur-
chased from Aldrich, BDH, or Lancaster and used without
further purification. Carveol and carvone were kindly do-
nated by Quest International, and RuCl3 and RuO2 by John-
son Matthey Catalysts. O2 and N2O of 99% purity were from
BOC Gases.
2.2. Techniques
Thermal gravimetric analysis (TGA) of catalysts was per-
formed on a Setaram TG-DSC 111 analyser. BET surface
areas were obtained from nitrogen physisorption measured
on a Micromeritics ASAP 2000 instrument. Powder X-ray
diffraction (XRD) patterns were obtained with a Siemens D-
5005 diffractometer (Co-Kα radiation) and attributed with
the use of a JCPDS-ICDD database. The particle size of
RuO2 and CoO(OH) was estimated from the Scherrer equa-
tion; no correction for instrumental peak broadening or mi-
crostrain was made. X-ray photoelectron spectra (XPS) were
measured with an M-Probe SSI spectrometer with the use
of monochromated Al-Kα X-rays (1486.6 eV). The catalyst
samples were degassed for 24 h under vacuum before analy-
sis. The atomic percentage of Ru and Co at the surface was
2.4. Oxidation procedure
2.4.1. Oxidation with O2/air
The oxidation of alcohols was carried out in a 50-ml
round-bottomed three-neck glass flask equipped with a re-
flux condenser, a magnetic stirrer, and a gas inlet allow-
ing a flow of oxygen or air (25 ml/min) to be bubbled
into the reaction mixture. Because of the inherent danger
of mixing oxygen with hot organics, appropriate precautions
should be taken with this work, particularly if it is scaled up.
Table 1
Catalyst characterisation
a
b
c
d
Catalysts and supports
pH
Water content
(wt%)
Metal content (wt%)
Ru
Co
2
γ -Al O -n (neutral), S
= 163 m /g
7.0
4.5
9.0
7.0
7.0
7.0
9.0
4.5
7.0
BET
2
3
2
γ -Al O -a (acidic), S
= 155 m /g
BET
2
3
2
γ -Al O -b (basic), S
= 155 m /g
BET
2
3
2
SiO , S
= 320 m /g
BET
2
RuO /γ -Al O -n
4.3
5.8
5.7
5.1
5.3
2.00
2.05
3.23
2.66
2.99
38.0
30.5
2
2 3
RuO · CoO(OH)/γ -Al O -n
1.12
2
2 3
RuO · CoO(OH)/γ -Al O -b
1.73
1.41
1.61
19.5
24.3
2
2 3
RuO · CoO(OH)/γ -Al O -a
2
2 3
RuO · CoO(OH)/SiO
2
2
RuO · CoO(OH) · 3H O
21.6
23.0
2
2
RuO · 1.5CoO(OH) · 4H O
2
2
a
b
c
d
Ru/Co atomic ratios are given from the preparation stoichiometry; the catalyst composition was determined by XRD and TGA.
The pH of 10% aqueous suspension.
From TGA weight loss in the temperature range of 30–300 C.
◦
From ICP analysis.

T.L. Stuchinskaya et al. / Journal of Catalysis 231 (2005) 41–47
43
Typically, a mixture of an alcohol (2.5 mmol), Ru catalyst
(alcohol/Ru = 10:1–45:1 mol/mol), and decane (GC inter-
nal standard) in toluene (10 ml) was charged in the reactor
and saturated with oxygen at room temperature for 5 min
while it was rapidly and thoroughly stirred. Then the reactor
was placed in the oil bath, which was preheated to a certain
temperature to start the reaction. We took samples of the re-
action mixture at appropriate time intervals to monitor the
reaction by GC (Varian 3800 gas chromatograph equipped
with a 25 m BP5 capillary column). Reaction products were
identified by GC-MS with the use of authentic samples.
◦
Fig. 1. XRD pattern for RuO pretreated at 130 C/10 h; matches the pat-
2
2.4.2. Oxidation with N2O
tern of rutile RuO .
2
The oxidation of alcohols with N2O was carried out in
a 50-ml glass-lined stainless-steel autoclave equipped with
a pressure gauge and magnetic stirring. Typically, the reac-
tion mixture contained 1.0 mmol substrate and ca. 0.1 mmol
dodecane (internal GC standard) in 5.0 ml solvent and an ap-
propriate amount of a catalyst. The mixture was placed in the
autoclave, which was then pressurised with N2O and vented
three times at room temperature with stirring to remove air
from the system. Finally, the autoclave was pressurised with
N2O and placed in the oil bath, which was preheated to
the reaction temperature, to carry out alcohol oxidation with
thorough stirring. After that, the reactor was cooled, depres-
surised, and opened, and the reaction mixture was analysed
by GC.
◦
Fig. 2. XRD patterns: 1) Ru–Co (1:1.5) oxide pretreated at 60 C/0.5 Torr,
2 h; matches the pattern of heterogenite-3R CoO(OH); 2) Ru–Co (1:1.5)
◦
oxide pretreated at 130 C/10 h; shows the patterns of rutile RuO and
2
heterogenite-3R CoO(OH).
3. Results and discussion
an amorphous RuO2 hydrate and the crystalline cobaltic acid
CoO(OH) (heterogenite-3R) with a crystallite size of 100 Å.
The peak at 42.6◦ may indicate the presence of some Co3O4.
The catalyst might also include a mixed Ru–Co oxide phase.
Thus the catalyst composition can be approximated as a bi-
nary oxide RuO2 · (1–2)CoO(OH) · (3–5)H2O. The catalyst
dehydrated at 130 ◦C/10 h to an inactive form with the pat-
terns of both rutile RuO2 and heterogenite-3R CoO(OH)
(Fig. 2, pattern 2), with crystallite sizes of 90 and 80 Å, re-
spectively. The hydration water may play an important role,
generating active RuIV hydroxo species (vide infra).
The XPS analysis of the oxidation state of Ru in the cata-
lysts, although inconclusive because of the unavoidable car-
bon contamination [17], was compatible with the presence
of RuIV. The Ru/Co atomic ratio at the surface of Ru–Co
oxide (1.55:1) was found to be considerably higher than the
bulk ratio (1:1.5). This indicates segregation of Ru and Co
in the catalyst, with Ru concentrating at the catalyst surface.
H2-TPR profiles for bulk and supported catalyst samples
are shown in Figs. 3 and 4, respectively. Bulk RuO2 (ru-
tile) reduces quantitatively to Ru0 at about 97 ◦C (Fig. 3,
profile 4). A small peak at 176 ◦C may be due to the
reduction of unknown ruthenium species present in the
catalyst. RuCl3 reduces to Ru0 at a higher temperature
(215 ◦C), as expected (profile 1). CoO(OH) exhibits reduc-
3.1. Catalyst characterisation
The Ru–Co mixed oxide catalysts were characterised by
ICP (chemical composition), TGA (water content), XRD
(phase analysis), and H2-TPR (redox properties). Only hy-
drous oxides were catalytically active in the oxidation of
alcohols [11–13]. These oxides were obtained by a mild
thermal pretreatment of the precursor hydroxides, typically
at 60 ◦C/0.5 Torr/2 h. From TGA, the active bulk oxides
contained three to five water molecules per Ru atom, as
determined from weight loss in the temperature range of 30–
300 ◦C (Table 1). Thoroughly dehydrated oxides were inac-
tive in the oxidation of alcohols. The hydrous RuO2 has been
reported to be different from the anhydrous form [15]. As
shown by powder XRD, the hydrous RuO2 was amorphous;
it had a BET surface area of ca. 200 m2/g. After dehydration
at 130 ◦C/10 h, the Ru dioxide transformed to the inactive
crystalline RuO2 with the rutile structure (Fig. 1), which had
a low surface area (ca. 10 m2/g), in agreement with the liter-
ature [15]. Hydrous RuO2 has been reported to chemisorb a
significant amount of oxygen, whereas the anhydrous form
adsorbs little oxygen [15]. These differences might greatly
affect the activity of RuO2 in the oxidation of alcohols [16].
From XRD data (Fig. 2, pattern 1), the hydrous Ru–Co
mixed oxide of the optimal 1:1–1:2 atomic ratio comprises

44
T.L. Stuchinskaya et al. / Journal of Catalysis 231 (2005) 41–47
Fig. 3. H -TPR profiles: (1) RuCl · 2H O, (2) CoO(OH), (3) RuO ·
2
3
2
2
CoO(OH) · 3H O, (4) RuO .
2
2
tion to Co0 around 285 ◦C (profile 2), in agreement with
previous reports ([18,19] and references therein). In bulk
RuO2 · CoO(OH) catalyst, RuIV reduces quantitatively to
Ru0 practically in the same temperature range as in the case
of RuO2 (profile 3). Interestingly, no separate peak for Co
reduction is observed up to 500 ◦C. This is not unexpected,
as doping Co catalysts with ruthenium has been shown to
significantly increase the temperature of Co reduction [18].
In alumina-supported RuO2 and Ru–Co (1:1) oxide, RuIV
reduces to Ru0 in the same temperature range as it does in
bulk oxides, regardless of the acidity of alumina (Fig. 4, pro-
files 1–4). On silica, the reduction occurs at a slightly higher
temperature, 136 ◦C (profile 5). No Co reduction peak is ob-
served for the supported Ru–Co catalysts up to 500 ◦C. This
Fig. 4. H -TPR profiles for supported Ru–Co catalysts: (1) RuO /
2
2
γ -Al O -n, (2) RuO · CoO(OH)/γ -Al O -a, (3) RuO · CoO(OH)/
2
3
2
2
3
2
γ -Al O -n, (4) RuO · CoO(OH)/γ -Al O -b, (5) RuO · CoO(OH)/SiO .
2
3
2
2
3
2
2
may be explained by the hampering of Co reduction with Ru
and the formation of a more stable Co aluminate [18]. The
TPR results show that the CoIII additives have practically no
effect on the oxidizing power of RuIV.
3.2. Oxidation of alcohols by O2
Table 2 shows representative examples of oxidation of
saturated and unsaturated primary and secondary alcohols
by O2 with bulk and supported Ru–Co oxide as the catalyst.
Table 2
◦
a
Oxidation of alcohols by O catalysed by Ru–Co (1:1) oxide (in toluene, 110 C)
2
Entry
Catalyst
Alcohol
Time
(h)
Product
Conversion
(%)
Selectivity
(%)
TOF
(h
−1
)
b
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
Ru–Co
Cinnamyl
Cinnamyl
Cinnamyl
Cinnamyl
Cinnamyl
Cinnamyl
Cinnamyl
Cinnamyl
Benzyl
1-Decanol
1-Decanol
9-Decenol
9-Decenol
2-Decanol
0.5
0.5
1.0
0.25
0.25
0.5
0.25
0.5
0.25
3.0
4.0
3.0
4.0
4.0
1.0
1.0
Cinnamaldehyde
Cinnamaldehyde
Cinnamaldehyde
Cinnamaldehyde
Cinnamaldehyde
Cinnamaldehyde
Cinnamaldehyde
Cinnamaldehyde
Benzaldehyde
Decanal
Decanal
9-Decenal
9-Decenal
2-Decanone
t-Bu(Ph)CO
Carvone
96
100
100
97
100
76
96
56
100
86
90
75
92
87
100
83
94
> 99
> 99
98
> 99
> 99
99
> 99
> 99
84
> 99
76
38
40
20
78
80
68
77
22
80
5.6
4.5
4.9
4.6
4.4
20
17
Ru–Co/Al O -n
Ru–Co/Al O -n
Ru–Co/Al O -n
2 3
Ru–Co/Al O -a
Ru–Co/Al O -a
Ru–Co/Al O -b
Ru–Co/SiO
Ru–Co/Al O -n
Ru–Co/Al O -n
Ru–Co/Al O -n
Ru–Co/Al O -n
Ru–Co/Al O -n
Ru–Co/Al O -n
Ru–Co/Al O -n
Ru–Co/Al O -n
2
3
c
2 3
2
3
d
2 3
2
3
2
2
3
2
3
e
e
2
3
2
3
97
2
3
> 99
> 99
> 99
2
3
t-Bu(Ph)CHOH
2
3
f
Carveol
2
3
a
Alcohol (1.0 mmol), alcohol/Ru = 20:1 mol/mol, toluene (5 ml) and O (25 ml/min). Turnover frequencies (TOF): mol alcohol reacted per mol Ru and
2
per hour.
b
c
d
e
f
Catalyst, RuO · 1.5CoO(OH) · 4H O; 2.5 mmol alcohol; alcohol/Ru = 20:1 mol/mol [11].
2
2
Air (25 ml/min) instead of O as oxidant.
2
Alcohol/Ru = 45:1 mol/mol.
Radical scavenger 2,6-di-t-butyl-p-cresol (0.01 M) was added.
Carveol consisted of 57% trans and 43% cis isomer.

T.L. Stuchinskaya et al. / Journal of Catalysis 231 (2005) 41–47
45
The bulk oxide is useful for catalyst characterisation (vide
supra). The supported Ru–Co oxide is a much more efficient
catalyst because of the isolation of Ru sites, making more
Ru sites available for the reaction. The activated aromatic
and allylic alcohols, such as benzyl and cinnamyl alcohol,
are oxidised by O2 with an almost 100% yield of aldehydes
in 15 min (entries 5, 9). These last reactions may be affected
by oxygen transport into the liquid reaction mixture. The
oxidation also proceeds easily when air is used instead of
pure O2, though more slowly, as expected (entry 6). Non-
activated primary alcohols (e.g., 1-decanol) are more dif-
ficult to oxidise. They, too, give aldehydes in good yields,
but overoxidation to acids takes place. Addition of a radical
scavenger, such as 2,6-di-t-butyl-p-cresol, greatly improves
the yield of aldehyde (entries 10, 11). This indicates that
the overoxidation occurs via a radical mechanism. The ox-
idation of 9-decenol occurs chemoselectively to 9-decenal
without affecting the position of the double bond (entries 12,
13). Therefore, the Ru–Co oxide catalysts possess both high
activity and high chemoselectivity for alcohol-to-aldehyde
oxidation. It should be noted that the alumina-supported Ru–
Co oxide catalyst shows the highest output in the oxidation
of cinnamyl alcohol among the most efficient solid Ru cata-
lysts. It has a turnover frequency (TOF) of 80 h−1 at 110 ◦C
(entry 5), which is higher than those reported for Ru–Co–Al
hydrotalcite (14 h−1 at 60 ◦C) [6], Ru-hydroxyapatite (6 h−1
at 80 ◦C) [7], Ru–Co(OH)2–CeO2 (10 h−1 at 60 ◦C) [8], and
Ru/Al2O3 (27 h−1 at 83 ◦C) [9].
Fig. 5. Oxidation of cis and trans isomers of carveol (0.80 mol/l)
◦
with O catalysed by RuO · CoO(OH)/γ -Al O -n (in toluene, 110 C,
2
2
2 3
alcohol/Ru = 20:1 mol/mol): conversion of trans (open circles) and cis
(solid circles) carveol; selectivity to carvone (dotted line).
The Ru–Co oxide catalysts are also efficient for the oxi-
dation of secondary alcohols to ketones (Table 2, entries 14–
16). Interestingly, the aliphatic secondary alcohols oxidise
more slowly than the corresponding primary alcohols. Thus
the competitive oxidation of an equimolar mixture of 1-
and 2-dodecanol gives 81% aldehyde and 15% ketone. This
shows that steric effects play a significant role and indicates
that alkoxy RuIV intermediates may be involved in the reac-
tion,
Fig. 6. Oxidation of cinnamyl alcohol (0.20 mol/l) to cinnamaldehyde
◦
with O catalysed by supported Ru–Co oxide (in toluene, 110 C, alco-
2
hol/Ru = 45:1 mol/mol): (1) RuO · CoO(OH)/γ -Al O -a, (2) RuO ·
2
2
3
2
CoO(OH)/γ -Al O -n, (3) RuO · CoO(OH)/γ -Al O -b, (4) RuO ·
2
3
2
2
3
2
CoO(OH)/SiO .
2
double-bond migration,
.
The cis-isomer is more reactive than the trans-isomer
(Fig. 5).
Catalyst support plays a very important role (Fig. 6).
γ -Alumina is much better than silica. This is in agreement
with the H2-TPR results (Fig. 4), as Ru–Co/Al2O3 reduces
at a lower temperature than Ru–Co/SiO2. The acidity of alu-
mina is essential too; the acidic alumina is better than the
neutral and basic alumina. The acidity of support may affect
the state of RuIV and CoIII in the catalyst. It could also facil-
itate the formation of alkoxy RuIV intermediate (vide infra).
The Ru–Co catalyst can be reused several times after a
simple workup (Fig. 7). In repeated runs, the selectivity re-
Carveol (cis/trans = 43:57) is readily oxidised to carvone
with almost 100% selectivity (Table 2, entry 16), without

46
T.L. Stuchinskaya et al. / Journal of Catalysis 231 (2005) 41–47
Fig. 7. Catalyst reuse for cinnamyl alcohol oxidation by O : RuO ·
IV
2
2
Fig. 8. Mechanism of alcohol oxidation catalysed by Ru
.
◦
CoO(OH)/γ -Al O -n, 110 C, alcohol/Ru = 20:1 mol/mol, 0.5 h. The
2
3
catalyst was filtered off, washed with CH Cl , dried in vacuum and reused.
2
2
occur by a free radical mechanism, as it is effectively inhib-
ited by the addition of a radical scavenger (Table 2).
The Ru-catalysed oxidative dehydrogenation of alcohols
is likely to occur via the formation of RuIV alkoxide fol-
lowed by β elimination of Ru hydride species to yield alde-
hyde or ketone and subsequent catalyst regeneration by O2
(Fig. 8) [9]. This mechanism is supported by the follow-
ing evidence: (i) the results of oxidation of t-Bu(Ph)CHOH
favour a two-electron mechanism; (ii) catalyst hydration en-
hances the reaction that supports the participation of Ru–
OH groups; (iii) the substrate activity (primary alcohol >
secondary alkohol) is indicative of the participation of RuIV
alkoxide intermediates. In addition, Yamaguchi and Mizuno
[9] obtained important evidence supporting the formation of
Ru hydride species: they observed transfer hydrogenation
of a ketone that was used as the oxidant instead of O2 for
alcohol oxidation catalysed by alumina-supported Ru hy-
droxide. It should be noted that in this mechanism RuIV
does not change its oxidation state. If the mechanism is
correct, the RuIV acts as a Lewis acid rather than a redox
agent.
mains almost 100%; the conversion gradually declines, how-
ever. This may be due to catalyst dehydration during the
workup, which is difficult to avoid in batch operation. Wet-
ting the recovered catalyst with a drop of water improved its
reuse to some extent. It should be noted that no leaching of
Ru or Co from the catalyst was observed during the reaction
(ICP analysis).
3.3. Mechanistic features
The oxidation of alcohols to aldehydes and ketones in
the Ru–Co oxide/O2 system can be viewed as an oxida-
tive dehydrogenation, whereas the formation of acids is an
oxygenation; that is, it involves oxygen atom incorporation.
The test alcohol t-Bu(Ph)CHOH has been used to probe the
mechanism of alcohol oxidation—one-electron versus two-
electron transfer ([20] and references therein). With two-
electron oxidants, such as PdII, this alcohol gives the ke-
tone, t-Bu(Ph)CO, with the same carbon backbone. With
one-electron oxidants (e.g., Ce4+ or S2O82−) the C–C bond
between α and β carbon atoms cleaves to yield benzalde-
hyde and t-butyl radical as the primary products
Another important question is: what is the role of CoIII? It
is conceivable that CoIII could enhance the oxidation power
of Ru by increasing its oxidation state or facilitate cata-
lyst regeneration by activating O2. From H2-TPR (Figs. 3
and 4), the presence of CoIII does not increase the oxidation
power of Ru. Hence CoIII might facilitate catalyst regener-
ation (steps 3 and 4). These steps could be rate-controlling
in the catalytic process. Such a role of Co is in agreement
with the well-known ability of cobalt ions to activate O2 and
peroxides [14].
With supported Ru–Co oxide catalyst, t-Bu(Ph)CHOH
was found to selectively oxidise to t-Bu(Ph)CO (Table 2,
entry 15), which indicates that the oxidative dehydrogena-
tion of alcohols to aldehydes or ketones is a two-electron
process, and one-electron processes do not play important
role in these reactions. Similar behaviour has been observed
in alcohol oxidation with O2 catalysed by [n-Pr4N]RuO4 in
a homogeneous system [21] and by bulk RuO2 and Ru–Co
oxide in a heterogeneous system [11]. On the other hand, the
overoxidation of aldehyde to acid (oxygenation) appears to
3.4. Oxidation of alcohols by N2O
Oxidation by N2O has attracted considerable interest
[3,4]. N2O is a notorious air pollutant. It is produced in large
amounts as a by-product in oxidations by HNO3, for exam-
ple, in the production of adipic acid [4]. The use of N2O for
catalytic oxidation will therefore bring “green” benefits by
reducing its environmental impact. N2O is a powerful oxi-

T.L. Stuchinskaya et al. / Journal of Catalysis 231 (2005) 41–47
47
Table 3
Oxidation of cinnamyl alcohol to cinnamaldehyde by N O catalysed by Ru–Co (1:1) oxide
a
2
Entry
Catalyst
Oxidant
Conversion
(%)
Selectivity
(%)
TOF
(h
−1
)
b
1
2
3
4
5
RuO · CoO(OH)/Al O -n
N O (10 bar) + air
99
97
20
18
6
100
98
100
100
100
9.9
9.7
2.0
1.8
0.6
2
2
3
2
b
RuO · CoO(OH)/Al O -n
N
(10 bar) + air
2
2
3
2
c
RuO · CoO(OH)/Al O -n
N O (10 bar)
2
N O (10 bar)
2
2
2 3
c
RuO · CoO(OH) · 3H O
2
2
c,d
RuO · CoO(OH) · H O
N O (10 bar)
2
2
2
a
◦
Reactions were carried out in 50-ml stirred autoclave in toluene (5 ml) at 110 C for 2 h; 1.0 mmol alcohol, alcohol/Ru = 20:1 mol/mol. Turnover
frequencies (TOF) were defined as mol alcohol reacted per mol Ru and per hour.
b
The autoclave was pressurised with N O or N to 10 bar without removing air.
To remove air, the autoclave was pressurised with N O and vented three times then finally pressurised with N O to 10 bar.
2
2
c
2
2
d
◦
The catalyst was pre-treated at 110 C/0.5 Torr for 2 h.
dant (E0 = 1.77 V vs NHE). It is also a clean oxidant—only
N2 forms as a by-product. But N2O is a quite inert mole-
cule and a poor ligand; hence oxidation with N2O is very
difficult to realise. Very few catalysts are known to activate
N2O for selective oxidation. The best-known catalyst is Fe-
zeolite for the gas-phase oxidation of benzene to phenol by
N2O [3]. Some Ru complexes [22] and polyoxometalates
[23] catalyse alcohol oxidation by N2O in a homogeneous
solution.
Acknowledgments
Financial support from Johnson Matthey Catalysts, Quest
International, and EPSRC (grant GR/R53760) is gratefully
acknowledged. Thanks are due to Mr. M. Kett (Johnson
Matthey Catalysts) for the XPS measurements.
References
Ru–Co oxide was found to catalyse the oxidation of sat-
urated and unsaturated primary alcohols to aldehydes. Some
representative results for the oxidation of cinnamyl alcohol
are shown in Table 3. The reaction was carried out in an au-
toclave at 110 ◦C in toluene solution under a N2O pressure of
10 bar. In first experiments, no attempt was made to remove
air from the autoclave. Under such conditions, a cinnamalde-
hyde yield of almost 100% was obtained in 2 h (entry 1).
Then the blank experiments with N2 instead of N2O were
performed, which showed that this result was mainly due to
oxidation by the remaining O2 from air (entry 2). When the
air was thoroughly removed, only 20% alcohol-to-aldehyde
conversion was observed (entry 3).1 The oxidation by N2O,
like the oxidation by O2, was enhanced by the presence of
hydration water in the Ru–Co catalyst (entry 5), which sug-
gests the participation of RuIV hydroxo species in the reac-
tion. Unsaturated alcohols were oxidised chemoselectively
by N2O; for example, 9-decenol gave 9-decenal without mi-
gration of the double bond. N2O was found to selectively
oxidise t-Bu(Ph)CHOH to t-Bu(Ph)CO, indicating that the
oxidation is a two-electron process, like the oxidation by O2.
Therefore, the oxidation of alcohols by N2O catalysed by
Ru–Co oxide exhibits a close similarity to the correspond-
ing oxidation by O2. However, N2O is much less efficient
than O2, which is not unexpected. To make the oxidation by
N2O synthetically useful, much more active catalysts are re-
quired.
[1] R.A. Sheldon, I.W.C.E. Arends, A. Dijksman, Catal. Today 57 (2000)
157.
[2] T. Mallat, A. Baiker, Chem. Rev. 104 (2004) 3037.
[3] G.I. Panov, A. Uriate, M.A. Rodkin, V.I. Sobolev, Catal. Today 41
(1998) 365.
[4] A.B. Leontiev, O.A. Fomicheva, M.B. Proskurina, N.S. Zefirov, Usp.
Khim. 70 (2001) 107.
[5] T. Naota, H. Takaya, S.I. Murahashi, Chem. Rev. 98 (1998) 2599.
[6] T. Matsushita, K. Ebitani, K. Kaneda, Chem. Commun. (1999) 265.
[7] K. Yamaguchi, K. Mori, T. Mizugaki, K. Ebitani, K. Kaneda, J. Am.
Chem. Soc. 122 (2000) 7144.
[8] H. Ji, T. Mizugaki, K. Ebitani, K. Kaneda, Tetrahedron Lett. 43 (2002)
7179.
[9] K. Yamaguchi, N. Mizuno, Angew. Chem. Int. Ed. 41 (2002) 4538.
[10] P. Davey, I.V. Kozhevnikov, M. Musawir, G. Kelly, J.R. Anderson, WO
2004030812, 2002.
[11] M. Musawir, P.N. Davey, G. Kelly, I.V. Kozhevnikov, Catal. Commun.
(2003) 1414.
[12] T.L. Stuchinskaya, I.V. Kozhevnikov, Catal. Commun. 4 (2003) 417.
[13] T.L. Stuchinskaya, I.V. Kozhevnikov, Catal. Commun. 4 (2003) 609.
[14] F.A. Cotton, G. Wilkinson, Advanced Inorganic Chemistry, fifth ed.,
Wiley, New York, 1998.
[15] J.M. Fletcher, W.E. Gardner, B.F. Greenfield, M.J. Holdoway, M.H.
Rand, J. Chem. Soc. A (1968) 653.
[16] M. Matsumoto, N. Watanabe, J. Org. Chem. 49 (1984) 3435.
[17] J.J. Sunol, M.E. Bonneau, L. Roue, D. Guay, R. Schulz, Appl. Surf.
Sci. 158 (2000) 252.
[18] Y. Zhang, D. Wei, S. Hammache, J.G. Goodwin, J. Catal. 188 (1999)
281.
[19] E. van Steen, G.S. Sewell, R.A. Makhothe, C. Micklethwaite, H. Man-
stein, M. de Lange, C.T. O’Connor, J. Catal. 162 (1996) 220.
[20] I.V. Kozhevnikov, V.E. Tarabanko, K.I. Matveev, Kinet. Katal. 22
(1981) 619.
[21] M. Hasan, M. Musawir, P.N. Davey, I.V. Kozhevnikov, J. Mol. Catal.
A 180 (2002) 77.
[22] K. Hashimoto, Y. Kitaichi, H. Tanaka, T. Ikeno, T. Yamada, Chem.
Lett. (2001) 922.
1
The results reported earlier [13] showed higher yields of aldehydes and
[23] R. Ben-Daniel, R. Neumann, Angew. Chem. Int. Ed. 42 (2003) 92.
ketones. These were obtained in the presence of traces of air.