Hydrogen transfer processes mediated by supported iridium oxide nanoparticles

Article abstract of DOI:10.1002/cctc.201300253

Homogeneous iridium catalysts have demonstrated exceptional catalytic activity for a number of hydrogen transfer reactions. Herein, we demonstrate the synthesis of a heterogeneous iridium catalyst supported on nanoparticulate cerium oxide and investigate its application for the aerobic oxidation of benzyl alcohol and the Meerwein-Ponndorf-Verley transfer hydrogenation of cyclohexanone. Along with the optimisation of the activity of the catalyst, the kinetic parameters have been examined to unravel the elementary reaction steps mediated by this catalyst and further rationalise the observed structure-activity relationships. Both spectroscopic and catalytic investigations suggest that iridium oxide nanoparticles, most likely Ir₂O₃, mediate these reactions via the formation of metal hydroxide species, which are subsequently reoxidised with either a molecular oxygen or a ketone. In contrast to many other metal- or metal oxide-based catalysts, this catalyst can perform the selective oxidation of alcohols in the absence of a base, at mild temperatures and at a low metal loading. An idea in transit: Iridium oxide, supported on nanoparticulate cerium oxide, is reported to catalyze the oxidative dehydrogenation of alcohols via β-hydride elimination. The catalyst surface can be reoxidized with either a molecular oxygen or a ketone, which completes the transfer hydrogenation cycle.

Full text of DOI:10.1002/cctc.201300253

CHEMCATCHEM
FULL PAPERS
FULL PAPERS
An idea in transit. Iridium oxide, sup-
ported on nanoparticulate cerium oxide,
is reported to catalyze the oxidative de-
hydrogenation of alcohols via b-hydride
elimination. The catalyst surface can be
reoxidized with either a molecular
C. Hammond, M. T. Schꢀmperli,
S. Conrad, I. Hermans*
&& – &&
Hydrogen Transfer Processes
Mediated by Supported Iridium Oxide
Nanoparticles
oxygen or a ketone, which completes
the transfer hydrogenation cycle.
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 0000, 00, 1 – 9
&
1
&
ÞÞ
These are not the final page numbers!

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
DOI: 10.1002/cctc.201300253
Hydrogen Transfer Processes Mediated by Supported
Iridium Oxide Nanoparticles
Ceri Hammond, Martin T. Schꢁmperli, Sabrina Conrad, and Ive Hermans*[a]
Homogeneous iridium catalysts have demonstrated exception-
al catalytic activity for a number of hydrogen transfer reac-
tions. Herein, we demonstrate the synthesis of a heterogeneous
iridium catalyst supported on nanoparticulate cerium oxide
and investigate its application for the aerobic oxidation of
benzyl alcohol and the Meerwein–Ponndorf–Verley transfer hy-
drogenation of cyclohexanone. Along with the optimisation of
the activity of the catalyst, the kinetic parameters have been
examined to unravel the elementary reaction steps mediated
by this catalyst and further rationalise the observed structure–
activity relationships. Both spectroscopic and catalytic investi-
gations suggest that iridium oxide nanoparticles, most likely
Ir2O3, mediate these reactions via the formation of metal hy-
droxide species, which are subsequently reoxidised with either
a molecular oxygen or a ketone. In contrast to many other
metal- or metal oxide-based catalysts, this catalyst can perform
the selective oxidation of alcohols in the absence of a base, at
mild temperatures and at a low metal loading.
Introduction
In recent years, considerable attention has been paid to the
synthesis and exploration of new catalytic materials or com-
plexes based on late transition metals, such as Au, Pd, Ru and
Pt. The interest in these elements often lies in their considera-
ble redox potential, their large number of available oxidation
states, and their ability to coordinate and activate a number of
important substrates, such as dioxygen, hydrogen, alkanes, al-
cohols and olefins. Thus, these metals are the crucial species in
many well-known homogeneous catalysts and can efficiently
catalysed processes such as olefin metathesis (e.g. Ru),[1]
carbon–carbon coupling (e.g. Pd)[2] and the selective oxidation
of methane (e.g. Pt).[3] In contrast to many of the aforemen-
tioned transition metals, the chemistry of Ir has received rela-
tively little attention. This is in spite of its promising activity
and selectivity for a number of important chemical transforma-
tions: (de)hydrogenations,[4] alkane activation,[5] preferential CO
oxidation,[6] and more recently steam reforming,[7] (electro)cata-
lytic water oxidation,[8] amongst others.[9] Furthermore, in con-
trast to the increasing interest in the synthesis of monometal-
lic[10] and bimetallic[11] heterogeneous catalysts based on, for
example, Au, Pd, Pt and Ru, few reports describe the synthesis
and application of a heterogeneous Ir-based catalyst for selec-
tive chemical transformations and even fewer describe the use
of a heterogeneous Ir catalyst for liquid phase reactions.
an important role in the functionalisation of molecules. Despite
their pre-eminence, they remain some of the most environ-
mentally problematic transformations on an industrial scale,
often because of the nature of the oxidative species: bulky
and/or toxic inorganic salts of Cr6+ and Mn7+ are highly atom
inefficient, and oxidants such as alkyl hydroperoxides or Dess–
Martin periodinane co-produce stoichiometric quantities of
other organic species.[12] In the context of sustainable chemis-
try, developing catalysts that can perform selective oxidation
processes with ‘green’ and/or environmentally benign oxi-
dants, such as hydrogen peroxide or dioxygen, is a key chal-
lenge for the future.
Many selective oxidation processes proceed via the activa-
tion of CÀH bonds and the subsequent formation of metal hy-
dride or metal hydroxide species. Such species are often the
key species formed in many Ir-catalysed transformations,[13]
and we reasoned that if they could be formed on a heteroge-
neous catalyst, the catalyst may prove to be a suitable material
for the selective oxidative dehydrogenation of alcohols to alde-
hydes/ketones, a key transformation in synthetic chemistry.
Considering these factors, we decided to investigate the po-
tential activity and selectivity of various metal oxide-supported
Ir catalysts and evaluate their potential as heterogeneous cata-
lysts for aerobic oxidations.
Selective oxidation processes are one of the cornerstones of
the chemical industry at all levels of the value chain and play
Results and Discussion
[a] Dr. C. Hammond, M. T. Schꢀmperli, S. Conrad, Prof. Dr. I. Hermans
Institute for Chemical and Bioengineering
ETH Zurich
Material synthesis and catalyst evaluation
Given that a number of noble metal and metal oxide nanopar-
ticles have recently shown considerable catalytic activity for
the aerobic oxidation of alcohols, we decided that a solid cata-
lyst composed of supported iridium (oxide) nanoparticles
could prove to be a suitable material from which to begin our
Wolfgang-Pauli-Strasse 10, CH-8093 Zurich (Switzerland)
Fax: (+41)44-633-42-58
E-mail: hermans@chem.ethz.ch
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/cctc.201300253.
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 0000, 00, 1 – 9
&
2
&
ÞÞ
These are not the final page numbers!

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
investigations. Our preliminary studies therefore focused on
the deposition of Ir, in the form of H2IrIVCl6, onto various high-
surface area metal oxides, which are suitable as catalytic sup-
port materials. Although impregnation is not the most suitable
technique for the preparation of highly dispersed metal (oxide)
nanoparticles, this scalable methodology enabled the rapid
activity of the catalyst.[16] However, given the inapplicability of
TEM for this catalyst composition (see below), it is not possible
at this stage to confirm this. Secondly, the additional washing
method used for DP may lead to a reduction in the chloride
(ClÀ) concentration of the catalyst. Because ClÀ ions facilitate
1) the agglomeration of various metal (oxide) particles[17] and
2) poison catalytic active sites,[18] their absence from the DP
catalyst in comparison to the impregnated catalyst is of impor-
tance (Table S1). Nevertheless, the precise detrimental role of
ClÀ ions is currently the subject of further investigation.
screening of
a number of metal–support combinations
(Table 1). From these preliminary studies, we observed that
Table 1. Effect of the support and preparation method on the catalytic
activity of nanoparticulate IrxOy for the aerobic oxidation of benzyl
alcohol.[a]
Notably, the optimal activity for catalysts prepared by using
DP methodology was achieved at lower metal loadings, with
0.5 wt% Ir supported on CeO2 and calcined at 2008C (hence-
forth labelled 0.5Ir/CeO2(DP200C)), leading to the best activity
levels (Table 1, entry 6). Further improvement in catalytic activi-
ty could be achieved by varying the catalyst pre-treatment
method. For instance, although 0.5Ir/CeO2(DP200C) lead to
a 33.4% yield of benzaldehyde in 3 h, the reduction (in H2) of
the same catalyst at 4008C lead to a 54.8% yield of benzalde-
hyde in the same time (Figure 1a). However, the performance
Entry
Ir loading[b]
Support
Preparation method
Yield[c]
1
2
3
4
5
6
7
n.a.
2.5
2.5
2.5
1.5
0.5
0
no catalyst
TiO2
MgO
CeO2
CeO2
n.a.
0
<1
<1
16.8
21.4
33.4
8.4
impregnation
impregnation
impregnation
deposition–precipitation
deposition–precipitation
n.a.
CeO2
CeO2
[a] Reaction conditions: 3 h, 908C, 0.2m benzyl alcohol in toluene,
0.5 mol% metal (substrate/metal=200); [b] wt% metal; [c] calculated as
(moles of benzaldehyde produced/initial moles of benzyl alcohol)ꢂ100.
All catalysts were pre-treated in air at 4008C for 3 h, except entry 6
(2008C, 3 h, air). For entry 7, an equivalent amount of CeO2 to that used
in entry 6 was used as catalyst.
cerium oxide (CeO2) was the most suitable support material,
with a catalyst comprising 2.5 wt% Ir/CeO2 selectively (ꢀ98%)
oxidising benzyl alcohol to benzaldehyde at a rather modest
yield of 16.8% in 3 h after calcination at 4008C for 3 h
(entry 4). The unique ability of this catalyst composition to me-
diate this reaction could be due to a number of factors. Firstly,
CeO2 can activate and oxidise alcohols to a small extent, and
examination of the analogous reaction catalysed by an excess
of CeO2 alone revealed that this could contribute around one
half of the overall activity of the 2.5 wt% Ir/CeO2 catalyst. Fur-
thermore, CeO2 is a well-known oxygen pump that can readily
donate its lattice oxygen atoms to oxidation reactions. The re-
duced metal oxide may subsequently be re-oxidised by O2,
and this facile O2 transport could lead to significant improve-
ments in the catalytic system. Previous work has demonstrated
that CeO2, and particularly nanoparticulate CeO2, is one of the
most optimal support materials for Au-catalysed alcohol oxida-
tion for this reason.[14]
Figure 1. Temporal evolution of benzaldehyde with 0.5Ir/CeO2(DP400R): a) stan-
dard reaction; b) reaction performed with an added radical scavenger
(1.0 mol% 2,6-di-tert-butyl-p-cresol); c) hot filtration experiment; the catalyst
was removed from the hot reaction solution after 5 min. Further points rep-
resent the activity of the supernatant solution. Reaction conditions: 180 min,
908C, 0.2m benzyl alcohol in toluene, 0.5 mol% Ir, 1 bar (100 kPa) O2.
of this optimal catalyst is masked by a severe deactivation
effect, which occurs after approximately 30 min of the reac-
tion; as shown in Figure 1, the activity of this catalyst decreas-
es significantly at a benzaldehyde yield of approximately 50%.
In contrast, the less active catalysts, though possessing a lower
reaction rate, did not suffer such extreme deactivation.
After identifying Ir/CeO2 as the most suitable catalyst com-
position, we focused our attention on the optimisation of the
activity of this catalyst. We first investigated alternative meth-
ods of preparation and found that a deposition–precipitation
(DP) methodology, through which the cationic metal precursor
is deposited onto the support at high pH,[15] lead to improve-
ments in catalytic activity in comparison to the standard im-
pregnation route (Table 1, entry 5). This increase in activity may
tentatively be ascribed to two factors. Firstly, it is well known
that a DP methodology typically produces smaller nanoparti-
cles with a narrower particle size distribution in comparison to
a standard impregnation route, which may positively affect the
Nevertheless, the initial turnover frequency of this catalyst
(0.5Ir/CeO2(DP400R)) is more than three times higher than that
achieved by the other pre-treated catalysts in the first 10 min
of the reaction, which emphasises the beneficial effect of re-
ductive pre-treatment on the activity of the catalyst. Although
these TOFs are far lower than those previously observed for
the solvent-free aerobic oxidation of benzyl alcohol with
mono- and bimetallic Au, Pd and Au–Pd catalysts,[19,10b] the re-
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 0000, 00, 1 – 9
&
3
&
ÞÞ
These are not the final page numbers!

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
actions herein are performed in a solvent, at ambient pressures
and in the absence of base and are similar in magnitude to
those obtained for the aerobic oxidation with other precious
metal-based catalysts under similar conditions (Figure 2).[10c,14a]
particles between 1 and 2 nm in size. Although this does not
immediately confirm the nature of the Ir species in the CeO2-
supported materials, it suggests that Ir is nanoparticulate, es-
pecially given the more suitable preparation method used in
that case (e.g. DP). Furthermore, the lack of any XRD reflections
arising from Ir (not shown) suggests that any Ir species present
in the material are below Æ10 nm in size, given the rough usa-
bility limit of the technique.
More insight regarding the nature of the Ir species in the
active catalysts was, however, provided by various spectro-
scopic methods. Although UV/Vis analysis revealed little usable
information, given the broad absorption arising from the CeO2
support, Raman spectroscopy revealed that in addition to the
main peak arising from CeO2 (at 465 cmÀ1), the as-synthesised
and pre-treated catalysts had a weak broad signal at Æ550–
600 cmÀ1 (Figure S1). Such a signal has previously been attrib-
uted to IrÀO bonds,[20] and its presence in all the materials sug-
gests that Ir is not in metallic form. XPS analysis subsequently
confirmed this, as no signal attributable to Ir0 was detected
even after reduction at 4008C. Instead, the XPS spectra of all
samples consisted exclusively of Ir3+ and Ir4+ cations (Figur-
es S2 and S3).
Figure 2. Initial turnover frequency observed for 0.5Ir/CeO2 before and after
various pre-treatment methods. Reaction conditions: 10 min, 908C, 0.2m
benzyl alcohol in toluene, 0.5 mol% Ir, 1 bar (100 kPa) O2.
A linear correlation between the Ir3+/Ir4+ ratio and the initial
turnover frequency of each catalyst is illustrated in Figure 4.
To rationalise the observed trend in reactivity, we subse-
quently investigated the pre-treated series of catalysts with
a combination of spectroscopic and microscopic techniques.
Because of the low Z-contrast of Ir and Ce, we were unable to
determine any useful particle size and/or morphological data
from the most active catalysts by using scanning TEM (Fig-
ure 3a and b). However, analysis of the 2.5Ir/MgO(IMP200C) cata-
lysts revealed that even though Ir was prepared through im-
pregnation, it was present in nanoparticulate form, with most
Figure 4. Straight line relationship observed between the initial turnover
frequency and the Ir3+/Ir4+ ratio for 0.5Ir/CeO2 after various pre-treatments.
Reaction conditions: 10 min, 908C, 0.2m benzyl alcohol in toluene, 0.5 mol%
Ir. Reaction conditions: 180 min, 908C, 0.2m benzyl alcohol in toluene,
0.5 mol% Ir, 1 bar (100 kPa) O2.
This correlation suggests that Ir3+ is the key active component
of the catalyst. The observation that the catalytic activity of
the Ir-containing materials correlates so well with the fraction
of Ir3+ appears to be unusual, because one would expect Ir4+
species to be the stronger oxidant for this reaction. However,
Ir3+ species appear to be critical intermediates in homogene-
ous Ir chemistry, as Ir1+–Ir3+ cycles are crucial for activity for
transfer hydrogenations, hydroformylations and hydrogena-
tions.[4,13] Nevertheless, in the absence of direct mechanistic in-
formation of the reaction pathway mediated by these catalysts,
it is difficult to conclude that the same chemistry is at play.
Figure 3. Scanning TEM analysis of 2.5Ir/CeO2(IMP200C) (a,b) and 2.5Ir/
MgO(IMP200C) (c,d).
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 0000, 00, 1 – 9
&
4
&
ÞÞ
These are not the final page numbers!

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
The presence of cationic Ir species and the presence of IrÀO
bonds in the active catalysts suggest that the reactive species
on the surface is nanoparticulate iridium oxide (IrxOy). Given
the oxidation states present, it is likely that IrO2 (Ir4+) and the
sesquioxide Ir2O3 are the species present in the sample and
that catalytic activity correlates with the fraction of Ir2O3.
Notably, little Ce3+ was detected in any of the catalysts even
after high-temperature reduction in H2. This is a crucial obser-
vation because Ce3+ is a critical component of various CeO2-
supported Au catalysts for aerobic oxidation reactions, as it fa-
cilitates the formation of various metal hydroperoxy species,
which can subsequently lead to catalytic activity.[14b] Its ab-
sence in these highly active catalysts suggests that some alter-
native active species is formed in the present system.
ation at the benzylic position lead to a 40–50% decrease in
the reaction rate and KIE values of 1.88 at 908C and 1.75 at
708C (Figure S4). These relatively large KIE values confirm that
the cleavage of the CÀH bond and the subsequent formation
of a transient carbocation is rate limiting. In addition, an Arrhe-
nius expression was obtained from these experiments, which
indicated that the reaction proceeded with an activation barri-
er of 55.0 kJmolÀ1 and a pre-factor of 6.7ꢂ107 sÀ1, values that
are in line with those of similar supported metal catalysts
under comparable reaction conditions (Figure S5).[10c,14a]
To gain further insight into the reaction mechanism, we de-
termined the effect of the reactant concentrations on the ini-
tial rate of activity. These experiments demonstrate that al-
though the reaction is first order in [benzyl alcohol] (Figure 6)
Mechanistic investigations
We next turned our attention to the elucidation of the elemen-
tary reaction mechanism. We first excluded the possibility that
the oxidative mechanism was radical based, as the addition of
a known radical scavenger (2,6-di-tert-butyl-p-cresol, hence-
forth called ‘trap’) leads to no change in the initial rate of oxi-
dation for 0.5Ir/CeO2(DP400R) (Figure 1b). We thus focused on the
effect that substitution at the para position of benzyl alcohol
had on the initial rate of activity (Hammett correlation). As il-
lustrated in Figure 5, substitution at the para position of
Figure 6. Effect of benzyl alcohol concentration on the initial rate of reaction
for 0.5Ir/CeO2(DP400R). Reaction conditions: 180 min, 908C, various concentra-
tions of benzyl alcohol in toluene, 0.5 mol% Ir, 1 bar (100 kPa) O2.
and [catalyst] (Figure S6), it is zero order in [O2], at least at O2
partial pressures of 0.2 bar and above (20 kPa; Figure 7). These
observations are indicative of a reaction mechanism involving
1) the cleavage of the benzylic CÀH bond by the catalyst,
Figure 5. Hammett plot depicting the effect of the substitution of the para
substituent of benzyl alcohol on the initial rate of oxidation. Reaction condi-
tions: 180 min, 908C 0.2m benzyl alcohol in toluene, 0.5 mol% Ir, 1 bar
(100 kPa) O2.
benzyl alcohol leads to substantial modifications in the rate of
reaction and a linear correlation (R2 =0.993) is observed be-
tween the logarithms of rate constants against the Hammett
coefficient (s) of each substituent (Hammett plot). The nega-
tive slope signifies that 1) the reaction rate is favoured by elec-
tron-donating substituents in the para position and 2) the tran-
sition state likely possesses cationic character. This immediately
suggests that the cleavage of the benzylic CÀH bond is rate
determining, a hypothesis subsequently confirmed by examin-
ing the kinetic isotope effect (KIE) of the reaction. Monodeuter-
Figure 7. Effect of O2 partial pressure on the initial rate of the reaction for
0.5Ir/CeO2(DP400R). Reaction conditions: 180 min, 908C, 0.2m benzyl alcohol in
toluene, 0.5 mol% Ir.
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 0000, 00, 1 – 9
&
5
&
ÞÞ
These are not the final page numbers!

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
which is rate determining and 2) a re-oxidation of the reduced
metal centre, which is non-rate determining. The build-up of
positive charge at the benzylic position signifies that the CÀH
cleavage occurs more quickly than the final formation of the
C=O group. Such a mechanism would, however, require the
formation of a metal hydride (MÀH) or metal hydroxide (MÀ
OH) intermediate, which would be formed via a b-hydride
elimination of the adsorbed alcoholate and rupture of the CÀH
bond. Considering the probable nature of the active site (i.e.
Ir2O3), it is likely that the intermediate species formed is IrÀOH,
which is similar to the intermediates formed during Ru-cata-
lysed aerobic oxidation.[21] We emphasise that the ability of this
catalyst to perform this oxidation with air at atmospheric pres-
sure is highly promising, as it avoids issues associated with
forming explosive mixtures of hydrocarbon vapours and O2
and requires little technical expertise.
Figure 8. Temporal evolution of cyclohexanol by 0.5Ir/CeO2(DP400R) under
various conditions: a) standard reaction. Reaction conditions: 60 min, 808C,
0.4m cyclohexanone in 2-butanol, 0.5 mol% Ir, N2 atmosphere (1 bar;
100 kPa); b) hot filtration experiment. Catalyst removed after 5 min of the
reaction; c) reaction performed in an O2 atmosphere (1 bar; 100 kPa).
Notably, the concentration versus activity profile observed
for benzyl alcohol is indicative of a Langmuir–Hinshelwood-
type mechanism, in which the free alcohol is in equilibrium
with the adsorbed, activated substrate, most likely an alcohol-
ate-type species. Therefore, increasing the alcohol concentra-
tion beyond a certain level does not lead to an increase in the
reaction rate, as the sites responsible for alcohol coordination
and oxidation are fully saturated.
(Figure 7). A hot filtration experiment, in which the catalyst
was removed at an intermediate stage of the reaction, con-
firmed that the MPV reaction is catalysed heterogeneously
(Figure 8b).
To determine 1) whether the mechanism proceeded via the
dehydrogenation of the alcohol and subsequent re-oxidation
of the reduced metal hydroxide intermediate with O2, or
2) whether adsorbed oxygen, either adsorbed direct by IrxOy, or
adsorbed and transported through CeO2, was responsible for
the abstraction of the CÀH hydride, we used the Meerwein–
Ponndorf–Verley (MPV) transfer hydrogenation of cyclohexa-
none (Scheme 1), which as a useful probe reaction.
Despite the negligible yield of cyclohexanol in an O2 atmos-
phere (Figure 8c), some conversion of 2-butanol to 2-butanone
was indeed observed. The lack of cyclohexanol yield in this re-
action is due to the faster re-oxidation of the IrÀOH species by
O2 in comparison to a ketone; thus, the hydroxide species
formed is removed from the catalyst surface before it reacts
with cyclohexanone. Our observations during this reaction also
provided some idea regarding the nature of the iridium oxide
species on the catalyst surface; when the reaction was per-
formed in an inert atmosphere (i.e. in which reaction is ob-
served), the catalyst powder had a distinctive blue colour. In
contrast, in an O2 atmosphere, the powder was dark brown. Of
the two oxides of Ir, iridium(III) oxide (Ir2O3) is blue and iridiu-
m(IV) oxide (IrO2) is brown.
Scheme 1. Meerwein–Ponndorf–Verley reduction of cyclohexanone with 2-
butanol. Reaction conditions: 60 min, 808C, 0.4m cyclohexanone in 2-buta-
nol, 0.5 mol% Ir.
Catalyst deactivation
Our final series of investigations focused on the severe deacti-
vation process observed for the most active alcohol oxidation
catalyst 0.5Ir/CeO2(DP400R) (Figure 1). We focused initially on the
reuse of the catalyst to identify whether the catalyst was fully
deactivated or whether the loss in activity observed was a fea-
ture of the reaction mechanism. However, the second use of
a given charge of the catalyst leads to negligible levels of ben-
zaldehyde formation, which suggests that the catalyst was de-
activated. We therefore focused on the possible formation of
minor quantities of benzoic acid, given the non-negligible loss
in carbon balance during a reaction (Æ5%) and the known
ability of carboxylic acids to poison various hydrogenation/oxi-
dation catalysts. As shown in Figure 9, even by adding small
amounts of benzoic acid to the initial reaction solution, a con-
siderable decrease in activity is achieved, with the maximal
Upon performing the MPV reaction in an inert atmosphere
(nitrogen), quantitative conversion to cyclohexanol was ach-
ieved in short reaction times (Figure 8a). In contrast, upon per-
forming the reaction in air, no cyclohexanol was observed (Fig-
ure 8c). The ability of the reaction to proceed in the absence
of O2 demonstrates that physisorbed oxygen is not responsible
for the cleavage of the CÀH bond. We thus propose that a typi-
cal dehydrogenation mechanism is in place, in which the metal
hydroxide species, formed solely through the interaction of
the alcohol and the catalyst, are subsequently re-oxidised by
oxygen (in the case of aerobic oxidation) or by the unsaturated
substrate (cyclohexanone, in the case of the MPV reaction).
This agrees well with the zero-order dependence in O2
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 0000, 00, 1 – 9
&
6
&
ÞÞ
These are not the final page numbers!

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
Figure 10. Reuse ability of 0.5Ir/CeO2(DP400R) without and with an NaOH wash-
ing method. Reaction conditions: 180 min, 908C, 0.2m benzyl alcohol in tol-
uene, 0.5 mol% Ir, 1 bar (100 kPa) O2.
Figure 9. Effect of benzoic acid addition on the catalytic activity of 0.5Ir/
CeO2(DP400R). Reaction conditions: 180 min, 908C, 0.2m benzyl alcohol in tolu-
ene, 0.5 mol% Ir, 1 bar (100 kPa) O2. Benzoic acid added at various initial
concentrations.
boxylic acid), whether this undesirable increase in oxidation
state is due to the formation of carboxylic acid or whether the
two features are independent of one another.
yield of 0.5Ir/CeO2(DP400R) decreasing to zero after the addition
of only approximately 2.5 mol% of benzoic acid (relative to
benzyl alcohol). However, even though benzoic acid was delib-
erately added to the reaction solution, it could not be detected
by using GC–FID–MS. It is likely that once it is formed it quickly
coordinates to the catalyst surface and thus inhibits the coordi-
nation and subsequent reaction of the alcohol substrate.
Conclusions
We have shown that by supporting iridium on CeO2, a highly
active and selective heterogeneous catalyst for various transfer
hydrogenations (aerobic alcohol oxidation and Meerwein–
Ponndorf–Verley hydrogenation) can be prepared. The activity
of this catalyst is related both to the properties of the support,
that is the ability of CeO2 to transport oxygen, and the iridium
oxide centre. Key spectroscopic and catalytic investigations
suggest that the major active species is nanoparticulate Ir2O3,
which are preferentially formed through the high-temperature
(4008C) reduction of the as-prepared catalyst.
The kinetic parameters and mechanistic features of this
system are consistent with a b-hydride elimination mechanism
and indicate that the ability of this catalyst to mediate 1) the
aerobic oxidation of alcohols and 2) the Meerwein–Ponndorf–
Verley transfer hydrogenation of ketones is related to its ability
to form metal hydroxide species, which are subsequently trans-
ferred to either a molecular oxygen or a ketone, respectively.
We emphasise that this catalyst can perform these reactions at
mild temperatures (<1008C) in the absence of stoichiometric
quantities of a base (i.e. base-free) and with a molecular
oxygen or air at ambient pressure. Finally, we have demon-
strated that the catalyst suffers from a severe deactivation pro-
cess, which is caused by the radical-based byproduction of
benzoic acid. The coordination of the carboxylate to the metal
oxide centre significantly decreases its activity for this reaction,
though this can be limited by adding trace quantities of a radi-
cal scavenger to the initial reaction solution.
To rationalise the formation of benzoic acid, we reasoned
that the over-oxidation of benzaldehyde to benzoic acid could
proceed via free-radical intermediates. Given the positive effect
of the addition of the radical trap in terms of both catalyst de-
activation and carbon balance (Figure 1b), it seems highly
likely that benzoic acid forms through the free-radical (aut)oxi-
dation of benzaldehyde,[22] given that the addition of a known
radical scavenger minimises deactivation and leads to carbon
balance of Æ100%. By adding only 1 mol% of trap (relative to
the starting benzyl alcohol concentration), a 25% increase in
benzaldehyde yield can be achieved, which demonstrates the
effect of oxygen-centred radicals on mediating the formation
of benzoic acid and causing catalyst deactivation. It is likely
that this deactivation through acid formation is the reason
why many investigations concerning alcohol oxidation with
(noble) metal-based catalysts are performed in
a basic
medium; although leading to the formation of the correspond-
ing carboxylic acid through hemiacetal formation, neutralisa-
tion of the acid through formation of the sodium adduct re-
moves its coordinative ability, thus freeing the metal from the
undesired carboxylate coordination. Upon treating the deacti-
vated catalyst with a dilute solution of NaOH (0.1m, 20 mLgÀ1
of the catalyst), full catalyst activity was restored (Figure 10).
XPS analysis demonstrated that the spent catalyst, that is
0.5Ir/CeO2(DP400R) after the reaction, had a significantly lower
fraction of Ir3+. Given that Ir3+ is the critical component of the
catalyst (see above), the formation of Ir4+ during the reaction
could also contribute to catalyst deactivation. Nevertheless, it
is impossible to confirm at this stage whether the formation of
Ir4+ precedes the formation of carboxylic acid (i.e. whether the
oxidation of Ir3+ to Ir4+ is responsible for the formation of car-
We anticipate that the ability of this catalyst to mediate
these transformations will open the possibility of further devel-
oping heterogeneous iridium-based catalysts for a number of
other crucial transformations, such as hydrogenation, hydrofor-
mylation and CÀH activation.
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 0000, 00, 1 – 9
&
7
&
ÞÞ
These are not the final page numbers!

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
Experimental Section
Acknowledgements
Catalyst synthesis
We acknowledge ETH Zurich for financial support (project ETH-17
09-3) and are indebted to Dr. Frank Krumeich for the TEM
measurements.
Various metal oxide-supported Ir catalysts were prepared through
the wet impregnation of the relevant metal oxide by an appropri-
ate amount of aqueous solution of hexachloroiridic acid (H2IrCl6,
3.7 mgIr mLÀ1). The solvent was evaporated at 908C and the resul-
tant paste dried in an oven at 1108C for 16 h. Analogous materials
were also prepared by using a DP methodology. In this case, an
aqueous solution containing the appropriate amount of CeO2 and
H2IrCl6 was stirred at RT. The pH was subsequently raised to 10
with NaOH (0.1 M), and the solution was stirred for 24 h at ambient
temperature. The solid catalyst was filtered under vacuum, and the
catalyst was washed thoroughly with distilled water (2 LgÀ1 of cat-
alyst) before drying in an oven at 1108C for 16 h. The dried cata-
lysts were subsequently pre-treated in a sealed combustion fur-
nace in a flow of air (for calcination) or H2 (for reduction) for 3 h at
the required temperature described in the text (ramp rate:
208CminÀ1).
Keywords: elimination · heterogeneous catalysis · iridium ·
oxidation · hydrogenation
[1] a) M. S. Sanford, J. A. Love, R. H. Grubbs, J. Am. Chem. Soc. 2001, 123,
6543–6554; b) T. M. Trnka, R. H. Grubbs, Acc. Chem. Res. 2001, 34, 18–
29; c) R. H. Grubbs, S. Chang, Tetrahedron 1998, 54, 4413–4450; d) E. L.
Dias, S. T. Nguyen, R. H. Grubbs, J. Am. Chem. Soc. 1997, 119, 3887–
3897; e) A. Fꢁrstner, Angew. Chem. 2000, 112, 3140–3172; Angew. Chem.
Int. Ed. 2000, 39, 3012–3043.
[2] a) N. Miyaura, A. Suzuki, Chem. Rev. 1995, 95, 2457–2483; b) N. Miyaura,
T. Yanagi, A. Suzuki, Synth. Commun. 1981, 11, 513–519.
[3] a) N. F. Gol’dshleger, V. V. Es’kova, A. E. Shilov, A. A. Shteinman, Russ. J.
Phys. Chem. 1972, 46, 785; b) R. A. Periana, D. J. Taube, S. Gamble, H.
Taube, T. Satoh, H. Fujii, Science 1998, 280, 560–564; c) C. Hammond, S.
Conrad, I. Hermans, ChemSusChem 2012, 5, 1668–1686.
[4] a) J. Choi, A. H. R. MacArthur, M. Brookhart, A. S. Goldman, Chem. Rev.
2011, 111, 1761–1779; b) I. Gçttker-Schnetmann, P. White, M. Brookhart,
J. Am. Chem. Soc. 2004, 126, 1804–1811; c) R. H. Crabtree, M. F. Mellea,
J. M. Mihelcic, J. M. Quirk, J. Am. Chem. Soc. 1982, 104, 107–113; d) R.
Crabtree, Acc. Chem. Res. 1979, 12, 331–338; e) R. Ahuja, B. Punji, M.
Findlater, C. Supplee, W. Schinski, M. Brookhart, A. S. Goldman, Nat.
Chem. 2011, 3, 167–171; f) R. Kawahara, K.-i. Fujita, R. Yamaguchi, J. Am.
Chem. Soc. 2012, 134, 3643–3646; g) C. Cheng, M. Brookhart, J. Am.
Chem. Soc. 2012, 134, 11304–11307; h) S. Park, D. Bezier, M. Brookhart,
J. Am. Chem. Soc. 2012, 134, 11404–11407.
[5] a) B. A. Arndtsen, R. G. Bergman, Science 1995, 270, 1970–1973; A. M.
Voutchkova, R. H. Crabtree, J. Mol. Catal. A 2009, 312, 1–6; b) J. Y. Cho,
M. K. Tse, D. Holmes, R. E. Maleczka, M. R. Smith, Science 2002, 295,
305–308; c) J. K. Hoyano, A. D. McMaster, W. A. G. Graham, J. Am. Chem.
Soc. 1983, 105, 7190–7191.
[6] a) J. Lin, B. Qiao, J. Liu, Y. Huang, A. Wang, L. Li, W. Zhang, L. F. Allard, X.
Wang, T. Zhang, Angew. Chem. 2012, 124, 2974–2978; Angew. Chem.
Int. Ed. 2012, 51, 2920–2924; b) F. Morfin, S. Nassreddine, J. L. Rousset,
L. Piccolo, ACS Catal. 2012, 2, 2161–2168; c) L. Piccolo, S. Nassreddine,
F. Morfin, Catal. Today 2012, 189, 42–48.
[7] a) W. Cai, F. Wang, C. Daniel, A. C. van Veen, Y. Schuurman, C. Descorme,
H. Provendier, W. Shen, C. Mirodatos, J. Catal. 2012, 286, 137–152; b) F.
Wang, W. Cai, H. Provendier, Y. Schuurman, C. Descorme, C. Mirodatos,
W. Shen, Int. J. Hydrogen Energy 2011, 36, 13566–13574; c) W. Du, Q.
Wang, D. Saxner, N. A. Deskins, D. Su, J. E. Krzanowski, A. I. Frenkel, X.
Teng, J. Am. Chem. Soc. 2011, 133, 15172–15183.
Catalyst characterisation
The as-synthesised and pre-treated catalysts were characterised by
using a combination of scanning TEM, XRD, XPS and Raman spec-
troscopy techniques. XPS analysis was performed by using a PHI
5000 VersaProbe spectrometer (ULVAC-PHI, Inc.) equipped with
a 1808C spherical capacitor energy analyser and a multi-channel
detection system with 16 channels. The XPS spectra were acquired
at a base pressure of 5ꢂ10À8 Pa by using a focused scanning mon-
ochromatic AlKa source (1486.6 eV) with a spot size of 200 mm. The
instrument was run in the FAT analyser mode, with electrons emit-
ted at 458 to the surface normal. Pass energy used was 187.85 eV
for survey scans and 46.95 eV for detailed spectra. Charge neutrali-
sation was applied throughout the analysis by using both a cool
cathode electron flood source (1.2 eV) and low-energy Ar+ ions
(10 eV). Data were analysed with CasaXPS software. Powder XRD
was performed by a PANalytical X’Pert PRO X-ray diffractometer
equipped with a CuKa radiation source (40 kV and 40 mA) and a Ni
filter. Raman spectra were recorded by using a Renishaw’s InVia
Raman microscope. Samples were irradiated with the 514.5 nm line
of an Ar-ion laser at 20 mW power.
[8] a) J. F. Hull, D. Balcells, J. D. Blakemore, C. D. Incarvito, O. Eisenstein,
G. W. Brudvig, R. H. Crabtree, J. Am. Chem. Soc. 2009, 131, 8730–8733;
b) D. B. Grotjahn, D. B. Brown, J. K. Martin, D. C. Marelius, M.-C. Abadjian,
H. N. Tran, G. Kalyuzhny, K. S. Vecchio, Z. G. Specht, S. A. Cortes-Llamas,
V. Miranda-Soto, C. van Niekerk, C. E. Moore, A. L. Rheingold, J. Am.
Chem. Soc. 2011, 133, 19024–19027.
[9] a) H. Shi, O. Y. Gutierrez, G. L. Haller, D. Mei, R. Rousseau, J. A. Lercher, J.
Catal. 2013, 297, 70–78; b) H. Shi, X. Li, G. L. Haller, O. Y. Gutierrez, J. A.
Lercher, J. Catal. 2012, 295, 133–145; c) J. Moran, A. Preetz, R. A. Mesch,
M. J. Krische, Nat. Chem. 2011, 3, 287–290; d) I. Piras, R. Jennerjahn, R.
Jackstell, A. Spannenberg, R. Franke, M. Beller, Angew. Chem. 2011, 123,
294–298; Angew. Chem. Int. Ed. 2011, 50, 280–284.
[10] a) A. S. K. Hashmi, G. J. Hutchings, Angew. Chem. 2006, 118, 8064–8105;
Angew. Chem. Int. Ed. 2006, 45, 7896–7936; b) K. Mori, T. Hara, T. Mizu-
gaki, K. Ebitani, K. Kaneda, J. Am. Chem. Soc. 2004, 126, 10657–10666;
c) K. Yamaguchi, N. Mizuno, Angew. Chem. 2002, 114, 4720–4724;
Angew. Chem. Int. Ed. 2002, 41, 4538–4540.
[11] a) M. Sankar, N. Dimitratos, P. J. Miedziak, P. P. Wells, C. J. Kiely, G. J.
Hutchings, Chem. Soc. Rev. 2012, 41, 8099–8139; b) D. M. Alonso, S. G.
Wettstein, J. A. Dumesic, Chem. Soc. Rev. 2012, 41, 8075–8098; c) J. A.
Lopez-Sanchez, N. Dimitratos, N. Glanville, L. Kesavan, C. Hammond,
Catalyst testing
The aerobic oxidation of benzyl alcohol was performed in a 50 mL
round bottom flask sealed with a balloon of O2 or air. The reactor
contained the desired amount of the catalyst (typically correspond-
ing to 0.5 mol% Ir) and the reactant solution (0.2m benzyl alcohol
in toluene; total volume: 5 mL). Reactions were performed be-
tween 70 and 1108C, but typically at 908C for 3 h. High-pressure
experiments were performed in a Parr autoclave reactor (100 mL)
lined with Teflon (working volume: 70 mL). Product analysis and
quantification was performed by using GC–FID analysis (30 m FFAP
column) using a biphenyl internal standard. The MPV transfer hy-
drogenation of cyclohexanone was also performed in a 50 mL
round bottom flask equipped with a reflux condenser and kept
under a N2 atmosphere. The vessel was charged with the reactant
solution (0.4m cyclohexanone in 2-butanol) and the appropriate
amount of the catalyst (corresponding to 0.5 mol% Ir). Reactions
were performed at 808C and quantified against a biphenyl internal
standard by using GC–FID analysis (30 m FFAP column).
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 0000, 00, 1 – 9
&
8
&
ÞÞ
These are not the final page numbers!

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
J. K. Edwards, A. F. Carley, C. J. Kiely, G. J. Hutchings, Appl. Catal. A 2011,
391, 400–406.
[12] a) F. Cavani, N. Ballarini, S. Luciani, Top. Catal. 2009, 52, 935–947; b) F.
Cavani, J. H. Teles, ChemSusChem 2009, 2, 508–534; c) R. A. Sheldon,
Chem. Commun. 2008, 3352–3365.
[19] D. I. Enache, J. K. Edwards, P. Landon, B. Solsona-Espriu, A. F. Carley, A. A.
Herzing, M. Watanabe, C. J. Kiely, D. W. Knight, G. J. Hutchings, Science
2006, 311, 362–365.
[20] a) Y. B. Mo, I. C. Stefan, W. B. Cai, J. Dong, P. Carey, D. A. Scherson, J.
Phys. Chem. B 2002, 106, 3681–3686; b) R. K. Kawar, P. S. Chigare, P. S.
Patil, Appl. Surf. Sci. 2003, 206, 90–101.
[13] T. Suzuki, Chem. Rev. 2011, 111, 1825–1845.
[14] a) A. Abad, A. Corma, H. Garcia, Chem. Eur. J. 2008, 14, 212–222; b) A.
Abad, P. Concepcion, A. Corma, H. Garcia, Angew. Chem. 2005, 117,
4134–4137; Angew. Chem. Int. Ed. 2005, 44, 4066–4069.
[15] a) F. Moreau, G. C. Bond, A. O. Taylor, J. Catal. 2005, 231, 105–114; b) F.
Moreau, G. C. Bond, A. O. Taylor, Chem. Commun. 2004, 1642–1643.
[16] F. Moreau, G. C. Bond, Catal. Today 2006, 114, 362–368.
[17] H.-S. Oh, J. H. Yang, C. K. Costello, Y. M. Wang, S. R. Bare, H. H. Kung,
M. C. Kung, J. Catal. 2002, 210, 375–386.
[21] a) N. Mizuno, K. Yamaguchi, Catal. Today 2008, 132, 18–26; b) K. Yama-
guchi, N. Mizuno, Chem. Eur. J. 2003, 9, 4353–4361.
[22] I. Hermans, J. Peeters, L. Vereecken, P. A. Jacobs, ChemPhysChem 2007,
8, 2678–2688.
Received: April 4, 2013
[18] Y. Huang, A. Wang, L. Li, X. Wang, T. Zhang, Catal. Commun. 2010, 11,
1090–1093.
Published online on && &&, 0000
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 0000, 00, 1 – 9
&
9
&
ÞÞ
These are not the final page numbers!