Water as a catalytic switch in the oxidation of aryl alcohols by polymer incarcerated rhodium nanoparticles

Article abstract of DOI:10.1039/c7cy01006k

One of the major goals in the oxidation of organic substrates, and especially for alcohol oxidation, is the use of molecular oxygen as the oxidant under mild conditions. Here we report the synthesis and testing of Rh polymer incarcerated catalysts, using a metal so far not used for alcohol oxidation reactions, in which the catalytic activity towards aryl alcohol oxidation, for substrates like 1-phenylethanol and benzyl alcohol, is switched on by the addition of water as co-solvent in toluene. This is done by using air as oxidant at atmospheric pressure, in one of the mildest reaction conditions reported for this class of reaction. The promoting effect of water to higher conversions was observed also for rhodium over alumina supported catalysts, which were used as a benchmark allowing in all cases high conversion and selectivity to the ketone or the aldehyde within a short reaction time. The effect of water was explained as a medium capable to promote the oxidation of the alcohol to the ketone in a biphasic system assisted by phase transfer catalysis. This is particularly relevant for alcohols like 1-phenylethanol or benzyl alcohol that are not soluble in water at room temperature, and for which alternative oxidation routes are needed, as well as to switch on the catalytic activity of metal nanoparticles in a facile and green manner for the activation of molecular oxygen. Aliphatic alcohols like 1-octanol and 3-octanol were also tested, still showing Rh based catalysts as promising materials for this reaction if toluene only was used as solvent instead.

Full text of DOI:10.1039/c7cy01006k

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
Catalysis
Science &
Technology
View Article Online
View Journal
PAPER
Water as a catalytic switch in the oxidation of aryl
alcohols by polymer incarcerated rhodium
nanoparticles†
Cite this: DOI: 10.1039/c7cy01006k
Jack O. Weston,a Hiroyuki Miyamura,b Tomohiro Yasukawa,b Dedi Sutarma,a
Chloe A. Baker,a Preabjot K. Singh,a Mariela Bravo-Sanchez,c Naoko Sano,c
a
Peter J. Cumpson,c Yulia Ryabenkova,d Shū Kobayashi b and Marco Conte
*
One of the major goals in the oxidation of organic substrates, and especially for alcohol oxidation, is the
use of molecular oxygen as the oxidant under mild conditions. Here we report the synthesis and testing of
Rh polymer incarcerated catalysts, using a metal so far not used for alcohol oxidation reactions, in which
the catalytic activity towards aryl alcohol oxidation, for substrates like 1-phenylethanol and benzyl alcohol,
is switched on by the addition of water as co-solvent in toluene. This is done by using air as oxidant at at-
mospheric pressure, in one of the mildest reaction conditions reported for this class of reaction. The pro-
moting effect of water to higher conversions was observed also for rhodium over alumina supported cata-
lysts, which were used as a benchmark allowing in all cases high conversion and selectivity to the ketone
or the aldehyde within a short reaction time. The effect of water was explained as a medium capable to
promote the oxidation of the alcohol to the ketone in a biphasic system assisted by phase transfer catalysis.
This is particularly relevant for alcohols like 1-phenylethanol or benzyl alcohol that are not soluble in water
at room temperature, and for which alternative oxidation routes are needed, as well as to switch on the
catalytic activity of metal nanoparticles in a facile and green manner for the activation of molecular oxygen.
Aliphatic alcohols like 1-octanol and 3-octanol were also tested, still showing Rh based catalysts as promis-
ing materials for this reaction if toluene only was used as solvent instead.
Received 18th May 2017,
Accepted 18th July 2017
DOI: 10.1039/c7cy01006k
rsc.li/catalysis
In this context, a large number of catalytic systems is pres-
ent in the literature for the oxidation of alcohols to ketones
1. Introduction
The oxidation of alcohols to ketones is a very relevant oxida-
tion reaction in the area of catalysis, with applications rang-
ing from the manufacture of fine chemicals to mechanistic
studies,1,2 with ketones being important precursors for the
drug and food industries.3,4 On the other hand a major target
both in industry and academia is the use of molecular oxygen
as an oxidizing agent under mild conditions in order to
achieve energy-effective and sustainable oxidation processes.5
including: metal complexes,6 metal oxides,7 and supported
metal nanoparticles.8,9 Most of this recent research though is
devoted to solvent-free processes.10,11 Nevertheless, solvents
can play beneficial roles in promoting the solubility of oxygen
in the reaction media12 or by promoting the reactivity of
functional groups over metal or metal oxide interfaces.13
However, these studies are mainly centred on polarity effects.
In contrast, if a solvent could play a more active role for the
reaction, as well as being a green solvent like water, this
would be of extreme importance.
These premises, together with the aim to identify a more
general trend or rule to activate alcohols under mild condi-
tions, prompted us to investigate the use of rhodium polymer
incarcerated nanoparticles (Rh-PI) for the oxidation of alco-
hols like: 1-phenylethanol, benzyl alcohol, 1-octanol and
3-octanol. For this, we investigated the use of a solvent/co-sol-
vent system comprising toluene and water and the use of air
as oxidant under atmospheric pressure. In fact, besides the
industrial and fundamental relevance of alcohol oxida-
tion,14,15 the use of polymer incarcerated metal nanoparticles
has received limited attention so far for alcohol oxidation
a Department of Chemistry, University of Sheffield, Sheffield, S3 7HF, UK.
E-mail: m.conte@sheffield.ac.uk
b Department of Chemistry, Graduate School of Pharmaceutical Sciences, The
University of Tokyo, Hongō, Bunkyō-ku, Tokyo, 113-0033, Japan
c NEXUS, School of Mechanical & Systems Engineering, Newcastle University,
Newcastle upon Tyne, NE1 7RU, UK
d School of Engineering, Polymer IRC, University of Bradford, Bradford, BD7 1DP,
UK
† Electronic supplementary information (ESI) available: 1H-NMR spectra, GC/MS
chromatograms and fragmentation patterns, XPS data. Details of styrene based
copolymer synthesis, products characterization, partition constants, particle size
distribution. Schemes S1 and S2, Fig. S1–S41, and Tables S1–S7. See DOI:
10.1039/c7cy01006k
This journal is © The Royal Society of Chemistry 2017
Catal. Sci. Technol.

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
View Article Online
Paper
Catalysis Science & Technology
despite their interesting structural properties.16,17 These ma-
terials have the advantage of a ‘soft’ metal/support interac-
tion18,19 and therefore they are ideal for physical chemistry
studies aimed to represent a catalytic system where the cata-
lytic activity mostly originates from the metal nanoparticle
only, together with a high monodispersivity.20 Moreover, de-
spite Rh being a metal that has been proven to be efficient
for oxidative coupling reactions,21 it is mostly used for reduc-
tion reactions,22 and especially for the reduction of nitric ox-
ide in automotive exhaust systems. As such, it has been so
far neglected as a catalyst for alcohol oxidation, in favour of
metals like Pd, Pt and Au.23–25 Exceptions are the total
electro-oxidation of ethanol to CO2 in the presence of bime-
tallic electrodes comprising Rh/Pt alloys,26 the use of porphy-
rin complexes,27 and a recent theoretical study28 for the oxi-
dation of ethanol over Rh(111) surfaces assisted by water.
These studies led us to combine the use of co-solvents men-
tioned above to investigate if Rh-PI nanoparticles could actu-
ally be a catalytic system for alcohol oxidation, and to explore
the activity of this material against that of Rh over a standard
support like alumina, or hybrid materials comprising the
polymer cross linked with activated carbon.
[RhIJOAc)2]2 by NaBH4; (ii) a catalyst obtained by in situ reduc-
tion of [RhIJOAc)2]2 by NaBH4 (Experimental section and ESI†)
in the presence of both polymer and carbon black as support
(here denoted ad Rh-PI/CB), which includes cross-linking of
the polymer with activated carbon (see Experimental section
and ESI†); (iii) a catalyst obtained in the same way as Rh-PI/
CB but using alumina as support instead (Rh-PI/Al2O3); and
(iv) a commercial catalyst comprising Rh supported on alu-
mina only (Rh/Al2O3) used as a benchmark. These catalysts
form a set of materials ranging from a fully embedded metal
nanoparticle with soft metal support interactions (Rh-PI) to a
material presenting a metal/metal oxide interface (Rh/Al2O3)
that could also affect the catalytic activity.33
Catalytic tests were carried out at 100 °C for 4 h at atmo-
spheric pressure of air, using toluene as a solvent and a mo-
lar metal to substrate ratio of 1 : 100, and their activity is
reported in Table 1.
Under these reaction conditions both Al2O3 containing cat-
alysts are active towards the oxidation of 1-phenylethanol in
the range of 10% conversion, with acetophenone as the main
product with selectivity in excess of 97% (see ESI,† Fig. S1–
S4). However, ethylbenzene was also identified, ca. 2–4% se-
lectivity. The carbon mass balance was >98% in all cases,
and no trace of benzaldehyde was detected at GC/MS level,
(see ESI,† Fig. S5–S7), thus showing that decarboxylation does
not take place. The Rh-PI and Rh-PI/CB catalysts have a negli-
gible activity and a set of control tests (a to f, vide infra) was
carried out to confirm these results for solvent, autoxidation,
support, diffusion and impurities. (a) The use of Rh/Al2O3 in
presence of toluene only, showed that the catalyst is not able
to oxidize the solvent under these reaction conditions, and
therefore acetophenone is not formed by secondary reactions
involving toluene. (b) A test in the absence of any catalyst for
a toluene/alcohol reaction mixture led to less than 1% con-
version to acetophenone, thus showing negligible autoxida-
tion.34 In turn, even the very small activity detected for Rh-PI
and Rh-PI/CB can be ascribed to a reaction with the walls of
the reactor and thus the catalysts displaying no activity. (c) A
catalytic test carried out using only alumina showed little cat-
alytic activity, ca. 1.5% but with some ethylbenzene (ca. 30%
selectivity) thus suggesting that this support could promote a
disproportionation pathway.35 (d) It was also verified that at
100 °C the catalysts operate under kinetic control regime (i.e.
not diffusion limited by oxygen mass transfer), by exploring a
range of reaction temperatures from 80 to 120 °C. (e) Finally,
2. Results and discussion
2.1 Catalytic tests for 1-phenylethanol oxidation in the
presence of toluene as a solvent
Our investigation was initially focused on the oxidation of
1-phenylethanol to acetophenone due to its relevance in both
industrial applications, as a precursor for resins, adhesives
and drugs,29 as well as in physicochemical studies,30 as a
substrate to test novel catalysts by restricting the reactivity to
one CHOH group. Four different rhodium containing cata-
lysts were prepared and investigated towards the oxidation of
1-phenylethanol to acetophenone using oxygen from air as ox-
idant under mild conditions.
The catalysts tested were: (i) a catalyst comprising Rh
nanoparticles embedded in a polymeric matrix only31,32 (here
denoted ad Rh-PI) obtained using a styrene based copolymer
(Fig. 1, Experimental section and ESI,† Scheme S1) and reduc-
tion of a rhodium precursor, rhodiumIJII) acetate dimer
Table 1 Catalytic tests for the oxidation of 1-phenylethanol in toluene
by polymer incarcerated Rh nanoparticles and Rh supported on alumina.
Reaction conditions: T = 100 °C, P = atmospheric pressure, molar metal-
to-substrate M : S = 1 : 100, reaction time t = 4 h
Selectivity (%)
Conversion
Fig. 1 Copolymer units (1, 2 and 3) of the styrene based polymer used
to prepare the Rh polymer incarcerated nanoparticles, Rh-PI. The
metal precursor was [RhIJOAc)2]2 which was reduced by NaBH4. The
same polymer and precursor were used to prepare Rh-PI/CB and Rh-
PI/Al2O3, but with the in situ reduction of the metal and crosslinking of
the polymer in presence of the second support.
Catalyst
(%)
Acetophenone
Ethylbenzene
Rh-PI
<1
<1
10
100
100
94
0
0
6
2
Rh-PI/CB
Rh-PI/Al2O3
Rh/Al2O3
11
98
Catal. Sci. Technol.
This journal is © The Royal Society of Chemistry 2017

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
View Article Online
Catalysis Science & Technology
Paper
GC/MS analysis of the starting materials identified trace
amounts (ca. 2%) of 1-cyclohexylethanol and cyclohexyl
methyl ketone for 1-phenylethanol, and o-xylene for toluene
(see ESI,† Fig. S8–S11), however these are not considered to
affect our results. (f) Leaching tests were also carried out (see
sections 2.2 and 2.4), and these also confirmed the activity is
due to Rh nanoparticles.
and the C–Hα to the hydroxyl group39 of the alcohol. Many
reports suggest the presence of an alkoxide intermediate in
the presence of a base40,41 (Scheme 1).
In this case, a stabilization of the transition state by a pro-
ton carrier like water should accelerate the reaction. A recent
computational study on Rh[111] surfaces in the oxidation of
ethanol and glycerol,28 showed that water could promote the
dissociation of the O–H bond, which would accelerate the re-
action, and slightly inhibit the dissociation of the C–H bond
of the hydrogen alpha to the hydroxyl group. The net result
of these two effects would be an enhanced catalytic activity
promoted by water. However, it should be noted that ethanol
and glycerol are fully miscible in water, unlike our substrate
and products, and as such our system is necessarily biphasic.
Therefore, we explored the effect of water using a water and
toluene mixture with an initial 2 : 1 volume ratio of these two
solvents. The results of these catalytic tests are reported in
Table 2.
These results are quite surprising, as under these reaction
conditions water not only increased the activity of Rh-PI/
Al2O3 or Rh/Al2O3 catalysts, but also switched on the activity
of Rh-PI and Rh-PI/CB catalysts, turning them to materials
even more active than those containing alumina, with conver-
sions in the range of 50% (see ESI,† Fig. S12–S15). It is well
known that a change in reaction medium can change the ac-
tivity of a catalyst,42 however to the best of our knowledge
this is one of the most dramatic media-dependent reactivity
effect observed in the area of nanoparticles. Even more im-
portantly it is obtained by the ‘simple’ addition of water into
the catalytic system, and under very mild reaction conditions.
Moreover, this also shows that in spite of the catalysts' sur-
face composition (vide infra, section 2.6.2), the metal surface
is clearly accessible to the substrate for Rh-PI and Rh-PI/CB
when water is present.
These initial set of data and control tests, would suggest
that Rh-PI/Al2O3 and Rh/Al2O3 are active by virtue of a metal
oxide to metal nanoparticle interface induced by alumina.
Interface effects are very common in heterogeneous catalysis
and can strongly promote the catalytic activity of a reaction.36
Conversely, as this metal oxide-to-nanoparticle interface is
absent for Rh-PI and Rh-PI/CB these catalysts would not be
active for the oxidation of the alcohol under these reaction
conditions. Another factor that could, in principle, explain
the difference in activity between these catalysts could also
be the lack of swelling of the polymeric matrix for Rh-PI and
Rh-PI/CB, and thus the lack of diffusion of the reactants over
the Rh surface. However, these catalysts can swell well with
organic solvents such as toluene, THF or DCM. In fact, PI cat-
alysts comprising the same copolymer, although using gold
as active metal were capable to oxidize alcohols37 including
1-phenylethanol to acetophenone in the presence of toluene
as a solvent, as well as aldehydes to carboxylic acids in the
presence of toluene as the sole solvent.38 Therefore, we con-
sider toluene as an appropriate solvent to carry out the reac-
tion also for Rh-PI and such differences in activity could then
be ascribed to the presence of this particular metal.
2.2 Catalytic tests for 1-phenylethanol oxidation in the
presence of toluene and water as a co-solvent, water as
catalyst switch
In view of the apparent lack of activity for Rh-PI and Rh-PI/
CB in toluene, we considered the possibility of catalytic tests
using a reaction medium by adding water as a co-solvent,
with water acting as a possible proton carrier for the oxida-
tion reaction. It should be noted that as 1-phenylethanol and
acetophenone are largely immiscible in water at room tem-
perature, water cannot be used alone as a solvent for high
alcohol concentrations, and toluene has also to be present
to solubilize the substrate, the products, and to swell the
polymer.
This experimental evidence prompted us to study the ori-
gin of the effect of water as a catalytic switch in the most sys-
tematic manner and extending our study to other alcohols
(see section 2.7). At first we experimentally ruled out the ef-
fect of leaching, as well as assessed the role of the dispropor-
tionation pathway in this process in the first instance (vide
infra). Due to the large amount of Rh3+ present in our cata-
lysts, (section 2.6.2) and the known possibility of supported
metal systems to be subjected to leaching effects,43 the reac-
tion mixtures were characterized via ICP-MS. A negligible
metal leaching for all the catalysts was detected with values
of <0.1% for Rh-PI, Rh-PI/Al2O3 and Rh/Al2O3, and a metal
leaching of 1.3% for Rh-PI/CB (relative amount with respect
The oxidation of an alcohol to a ketone is formally a
sequential dehydrogenation via abstraction of two H atoms
(either as charged species or radicals) from the O–H group
Scheme 1 One of the accepted reaction mechanisms for alcohol oxidation via alkoxide formation. In this case by considering deprotonation of
the hydroxyl functional group, followed by hydrogen abstraction, as hydride, of the H in alpha to the OH group in a step mediated by the metal
(changes in metal oxidation state not shown). The alkoxide can also bind to the metal.40 An alternative mechanism involving H˙ radical abstraction
by atomic oxygen adsorbed over the catalyst surface is also possible.37,38
This journal is © The Royal Society of Chemistry 2017
Catal. Sci. Technol.

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
View Article Online
Paper
Catalysis Science & Technology
Table
2 Catalytic tests for the oxidation of 1-phenylethanol to
proportionation route. It is possible to calculate the contribu-
tion to the formation of the ketone by these two reactions by
splitting the contribution of these two routes in terms of
yields (full details in ESI† and Table S1). We observe that the
amount of acetophenone obtained from the direct oxidation
route is always dominant for all four catalysts and in excess
of 70% with respect to the disproportionation route (Table
S1†), except for Rh-PI/CB where acetophenone has a nearly
equal contribution from direct oxidation and disproportion-
ation (ca. 50% each). Conversely (Table 2 and Table S1†) the
two catalysts containing alumina promote the disproportion-
ation route to a much lesser extent (2 to 3 times less, ca.
15%) than Rh-PI and Rh-PI/CB. Hydrated γ-Al2O3 presents a
surface basicity with pKa values in the range of 8,45 and stud-
ies on the surface basicity of Au–Pd/TiO2 catalysts for the dis-
proportionation reaction of benzyl alcohol46 also reported an
inhibition of the disproportionation pathway by increased ba-
sicity of the catalyst or the reaction media. It is possible that
this same parameter can affect the product distribution also
in the case of Rh and Al2O3.
This set of catalytic data proves that water is capable to
trigger a set of reactions different in nature (oxidation and
disproportionation), and it also provides some ground for wa-
ter to be a proton carrier for the direct oxidation route of
1-phenylethanol to acetophenone. Yet for this reaction, and
to some extent also for the disproportionation reaction, we
would like to emphasize that the effect of water is not as ob-
vious as it may sound. In fact, both the polymer and the acti-
vated carbon have hydrophobic surfaces,47 thus the presence
of water was not expected to lead to a such large increase in
conversion. Moreover, water is also a by-product of the oxida-
tion reaction (see Schemes 1 and 2) and this should also dis-
courage this high increase in conversion. These apparently
counter-intuitive factors prompted us to consider the exis-
tence of a partition effect for the alcohol between the organic
and the aqueous phase and how this could affect our results.
acetophenone in water/toluene by polymer incarcerated Rh nanoparticles
and Rh supported on alumina. Reaction conditions: T = 100 °C, P = at-
mospheric pressure, M : S = 1 : 100, t = 4 h. Water : toluene volume ratio
2 : 1. Carbon mass balance >98%
Selectivity (%)
Conversion
Catalyst
(%)
Acetophenone
Ethylbenzene
Rh-PI
48
55
22
36
77
66
86
87
23
34
14
13
Rh-PI/CB
Rh-PI/Al2O3
Rh/Al2O3
to the total metal loading). We do not consider this amount
for Rh-PI/CB to be significant with respect to the other cata-
lysts,44 and clearly the results of the catalytic tests in Table 2
do not correlate with the negligible amount of Rh in solution.
Nevertheless, in order to rule out any possible effect induced
by solubilized species even in very small amount like in our
case, a control test using a soluble RhCl3·xH2O salt was car-
ried out. This test was also performed in toluene/water me-
dia, and used an amount of Rh equal to the total amount of
Rh present in the catalyst. The test did not show any catalytic
activity towards alcohol oxidation. We can conclude that the
catalytic activity originates from the metal nanoparticles.
Furthermore, the addition of water not only switched on
an oxidation route of alcohol to ketone for all catalysts
(Scheme 2(a)), but it also led to the unexpected formation
of ethylbenzene, which is
a
reduction product of
1-phenylethanol instead. This can be explained by dispropor-
tionation of two molecules of alcohol to give acetophenone
and ethylbenzene (Scheme 2(b)). Disproportionation of aryl
alcohols, has been reported for palladium or gold catalysts45
under solvent free conditions and pressure of O2 or Ar and
thus possibly suggesting the existence of both an aerobic and
anaerobic disproportionation pathway for this complex reac-
tion.46 Though, to the best of our knowledge this parallel re-
action pathway has not been reported for rhodium catalysts
under mild oxidizing conditions with air.
2.3 Toluene and water equilibria, partition constants
In view of the existence of this secondary pathway, the to-
tal amount of acetophenone would then be the consequence
of two contributions: the direct oxidation route and the dis-
As water and toluene are not miscible, and 1-phenylethanol
and acetophenone have a limited solubility in water at room
temperature, 19 and 5.5 g dm−3 respectively,48,49 the reaction
mixture is effectively a microemulsion under our reaction
conditions with droplets of toluene dispersed in water.
However, the catalytic conversion is calculated by
analysing the organic phase only (see Experimental section
and ESI†), from the amounts of 1-phenylethanol and
acetophenone in toluene (ethylbenzene can be neglected in
this data treatment as virtually insoluble in water). Yet de-
spite the low solubility of 1-phenylethanol and acetophenone
in water, this could potentially alter the observed conversion
values in the water/toluene mixture because of a partition of
the alcohol between water and toluene. That is the alcohol
could solubilize in water and lead to higher conversion values
than those we actually observe because of this latent partition
effect.
Scheme 2 (a) Direct oxidation of 1-phenylethanol to acetophenone
and (b) disproportionation of two reagent molecules to acetophenone
and ethylbenzene.
Catal. Sci. Technol.
This journal is © The Royal Society of Chemistry 2017