Acceptorless Alcohol Dehydrogenation Catalysed by Pd/C

Article abstract of DOI:10.1002/cssc.201901313

Although the selective oxidation of alcohols to carbonyl compounds is a critical reaction, it is often plagued by several challenges related to sustainability. Here, the continuous, acceptorless dehydrogenation of alcohols to carbonyl compounds over heter

Full text of DOI:10.1002/cssc.201901313

Accepted Article
Title: Acceptorless alcohol dehydrogenation catalysed by Pd/C
Authors: Guillermo Nicolau, Giulia Tarantino, and Ceri Hammond
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: ChemSusChem 10.1002/cssc.201901313
Link to VoR: http://dx.doi.org/10.1002/cssc.201901313
A Journal of
www.chemsuschem.org

10.1002/cssc.201901313
ChemSusChem
FULL PAPER
Acceptorless alcohol dehydrogenation catalysed by Pd/C
Guillermo Nicolau, Giulia Tarantino and Ceri Hammond*[a]
Abstract: Although the selective oxidation of alcohols to carbonyl
compounds is a critical reaction, it is often plagued by several
challenges related to sustainability. Here, we demonstrate the
continuous, acceptorless dehydrogenation of alcohols to carbonyl
et. al. have recently demonstrated that acceptorless
dehydrogenation can also be used as a starting point to achieve
a broader range of organic reactions, following further reaction
of the generated carbonyl compounds with nucleophiles, such
compounds over heterogeneous catalysts in the absence of oxidants, as amines and terminal alkenes (so-called one pot synthesis via
bases or acceptor molecules. In addition to improving selectivity and
atom efficiency, the absence of an acceptor results in the co-
production of molecular H2, a clean energy source, and permits
dehydrogenation to proceed at >98% selectivity at turnover
frequencies values amongst the highest in the literature. Moreover,
excellent durability is observed during continuous operation over 48
h, reaching space-time-yields of 0.683 g(product) mL-1 h-1, better than
the state of the art by over two orders of magnitude. Alongside these
breakthroughs, we also identify the basic kinetic parameters of the
reaction, allowing some of the elementary reaction steps to be
identified.
acceptorless dehydrogenative coupling).4c
Due to the potential of this approach, several attempts to
develop catalysts capable of mediating the reaction have been
made. Much focus has been placed upon homogeneous
catalysts. However, these typically require several additives to
be used (thus decreasing atom efficiency), or require the use of
scare metals and/or expensive ligand complexes for high
performance to be achieved. Amongst potential heterogeneous
catalysts, supported noble metal nanoparticles (Cu, Ag, Au)
exhibit high levels of performance, with the best catalytic
performance being achieved by Ag nanoparticles supported on
hydrotalcite, which achieves a turnover frequency (TOF) of
around 2000 h-1.5c Although displaying lower activities (TOF <
200 h-1), examples of supported Ni,5i,5m Co,5h Pt,5j Re5f and
Ru5a,5b catalysts have also been reported in the literature.
Nevertheless, despite these previous breakthroughs, alcohol
dehydrogenation over heterogeneous catalysts still suffers from
several major drawbacks, including low intrinsic activity (i.e. low
TOFs) and a lack of information on their long term stability.
Moreover, the reactors employed for this reaction to date
generally prohibit time on line measurement of the H2 produced
during the chemical reaction, and therefore do not allow detailed
kinetic studies of the chemical process to be made, thus
preventing a detailed reaction mechanism from being identified.
Finally, the viability of this reaction to operate in a continuous
manner has yet to be studied, despite it being an essential
requirement for intensification purposes. Combined, these
disadvantages limit the favourability and sustainability of this
method, and make the rational improvement of catalytic activity
extremely challenging. Hence, several challenges remain to be
tackled.
Introduction
The synthesis of carbonyl compounds via alcohol oxidation is
one of the most important chemical reactions in organic
synthesis, due to the key role played by ketones and aldehydes
as building block or intermediates for a wide range of value-
added compounds, including fine chemicals and polymers.1
However, despite the importance of this process for synthetic
applications, alcohol oxidation is often performed by use of
stoichiometric equivalents of high molecular weight oxidising
agents.2 Evidently, use of such oxidants dramatically affects the
sustainability of these process, through the co-production of
stoichiometric amounts of waste, and by decreased atom
efficiency. Accordingly, much effort has been placed on
increasing the sustainability of this reaction over the last decade.
One approach involves the development of catalysts capable of
converting alcohols to the corresponding carbonyl compounds
using molecular O2 as a benign oxidant, i.e. removing the need
for inorganic and/or high molecular weight oxidants (oxidative
dehydrogenation).3 Another approach that has received far less
attention is the development of catalysts capable of catalysing
alcohol dehydrogenation under inert conditions, producing H2 as
by-product.4,5 Although far less explored, this last route
(acceptorless dehydrogenation) is especially desirable for a
series of reasons, the most notable of which include the parallel
production of H2, highly valuable as a sustainable source of
energy,6 and increased selectivity towards the desired carbonyl
compound, which is often readily overoxidised to the undesired
carboxylic acid in the presence of O2.3c Moreover, Gunanathan
Herein, we demonstrate the continuous, acceptorless
dehydrogenation of alcohols over 5 wt. % Pd supported on
activated carbon (henceforth, Pd/C),
a
readily available
heterogeneous catalyst known to be active for H2 production
from formic acid. We show this commercial catalyst is able to
quantitatively convert alcohols, such as 1-phenylethanol, to the
corresponding carbonyl compound with > 98 % selectivity under
optimised reaction conditions, whilst concurrently releasing
molecular H2 at a 1:1 molar ratio. The TOF values of the catalyst
reach up to 1475 h-1, amongst the highest in the literature.
Moreover, under continuous operational conditions, the catalyst
exhibits a space-time-yield (STY, amount of product produced
per volume of reactor per unit time) two orders of magnitude
higher than the best catalyst previously reported in the literature
(Ag supported on hydrotalcite),5c and is active for over 48 hours
on stream. We also report a novel type of reactor able to allow
time on line collection and measurements of the H2 produced
during reaction. By use of this reactor, detailed kinetic
investigation of the model reaction (1-phenylethanol
dehydrogenation to acetophenone) is achieved, providing
hitherto inaccessible insights into the reaction mechanism(s) of
the reaction.
[a]
G. Nicolau, Dr. G. Tarantino, Dr. C. Hammond
Cardiff Catalysis Institute
Cardiff University
Main Building, Park Place
Cardiff, CF10 3AT (UK)
E-mail: hammondc4@cardiff.ac.uk
Homepage: http://blogs.cardiff.ac.uk/hammond/
Supporting information for this article is given via a link at the end of
the document.((Please delete this text if not appropriate))
1
This article is protected by copyright. All rights reserved.

10.1002/cssc.201901313
ChemSusChem
FULL PAPER
Results and Discussion
deviation from linearity occurs above 0.1 mole % Pd. Thus, to
maintain kinetic integrity, 0.08 mol % was chosen as the amount
of catalyst for all further catalytic reactions. Additional
optimisation experiments were also performed (ESI Figure S2),
suggesting that stirring at rates above 250 rpm was also
essential to avoid mass transfer limitations.
Preliminary kinetic measurements.
Building on our previous work of formic acid dehydrogenation, in
which commercial palladium on carbon (Pd/C from Sigma
Aldrich, 5% wt.% Pd) exhibited the best catalytic performances
for H2 generation, we identified Pd/C to also be a suitable choice
of catalyst for the acceptorless dehydrogenation of 1-
phenylethanol to acetophenone. Kinetic studies were first carried
out using Pd/C as catalyst in N2 atmosphere, at conditions
comparable to various reports in the literature. Preliminary time
on line analysis, performed at a Pd:substrate ratio of 1:80 (1.25
mol % of Pd), showed that after 3 hours of reaction, over 90 %
of the substrate was converted into the desired product. Notably,
an acetophenone selectivity > 95 % was observed throughout
the reaction period, indicating that the dehydrogenation of 1-
phenylethanol can efficiently be catalysed by Pd/C in N2.
To ensure that the reaction is truly heterogeneously
catalysed, i.e. that no contribution to the reaction rate derives
from homogeneous species leached into the reaction mixture, a
“hot filtration” experiment was performed By filtering the catalyst
from the solution during the reaction, and analysing the residual
activity of the supernatant solution following removal of the
catalyst (ESI Figure S3). Following filtration of the solid catalyst,
no further changes to the solution (1-phenylethanol conversion
and acetophenone yield) were observed, indicating termination
of the reaction by removal of the catalyst and, hence,
demonstrating that the reaction is heterogeneously catalysed.
Acceptorless alcohol dehydrogenation is particularly
beneficial due to the co-production of H2, which is a valuablue
by-product. Therefore, accurate analysis of the gas produced
throughout 1-phenylethanol dehydrogenation is essential both to
perform thorough kinetic and mechanistic studies of alcohol
dehydrogenation, and confirm the formation of H2 as reaction
product, and confirm that the reaction is truly acceptorless.
However, under typical literature conditions i.e. in conventional
borosilicate batch reactors, technical challenges prevent
accurate collection and quantification of the gas produced during
the reaction. To overcome these problems, and, therefore, to
allow accurate analysis of the gaseous products, a novel reactor
was developed, by coupling a stainless steel reactor body with a
pressurised collection vessel with a tolerance of 8 bar (ESI
Figure S4). Equipped with this set up, the volume and
composition of the gas formed during reaction was periodically
monitored at three different temperatures, in a range 110-130 ˚C.
In all cases, the amount of gas collected during the first 10
minutes of reaction was in good agreement to the theoretical
amount of gas that should be produced based on the moles of 1-
phenylethanol converted (Table 1). This confirms that
appropriate time on line measurements can be achieved by
measuring the gas evolution during the reaction. In addition to
this, analysis of the liquid phase at the end of the reaction also
allows substrate conversion and product yields to also be
determined for all liquid products.
Scheme 1. General reaction scheme for 1-phenylethanol dehydrogenation to
acetophenone over Pd/C in N2.
Figure 1. Left. Conversion (X) of 1-phenylethanol, and yield (Y) of
acetophenone over time. Right. Initial (0.5 h) conversion (X) of 1-
phenylethanol as a function of metal molar ratio (moles Pd / moles 1-
phenylethanol x 100). Reaction conditions: 20 mL of a solution 0.2 M of 1-
phenylethanol in p-xylene, 120°C. Reaction carried out under N2 flow of 10
mL/min. (Left) 1.25 mol % Pd/Substrate molar ratio. (Right) Various amounts
of Pd/C (5 wt. %) in the range 0-1.25 mol %, reaction time 0.5 hours.
Table 1. Catalytic performances of Pd/C at different temperatures.[a]
T.
X
Y
S
C Bal.
Theor. vol.
Coll. vol.
(ºC) [a]
(%)[b]
(%)[c]
(%)[d]
(%)[e]
(mL)[f]
(mL)[g]
130
120
110
29.6
23.6
15.4
28.8
23.6
15.4
97.6
> 98
> 98
99.4
100
100
26.1
22.0
14.4
25
20
15
Interestingly constructing an Arrhenius plot by measuring the
initial rate of reaction (k) between 100-120 ˚C resulted in an
activation energy of only 31 kJ/mol being observed (ESI Figure
S1). Considering that the cleavage of a C(sp3)-H bond should be
involved in the reaction mechanism, the extremely low barrier
obtained indicates contribution from other factors, particularly
mass transfer. Therefore, further studies were performed to
ensure that the amount of catalyst employed was low enough to
carry out the reaction under the kinetic regime. Accordingly, the
amount of Pd, relative to phenylethanol (mole %) was varied
from to 0 to 1.25 mol % (Figure 1 (Right)). As can be seen,
although a linear relationship between initial conversion (at 0.5
h) and the quantity of Pd exists at low metal molar ratios,
[a] Reaction conditions: stainless steel reactor body with 100 mL
pressurised round bottom flask. 0.2 M of 1-phenylethanol in p-xylene, 0.08
mol % of Pd/substrate molar ratio, static N2 atmosphere, 10 mins. [b]
Conversion of 1-phenylethanol. [c] Yield of acetophenone. [d] Selectivity for
acetophenone. [e] Carbon balance. [f] Theoretical volume of gas. [g]
Collected volume of gas.
2
This article is protected by copyright. All rights reserved.

10.1002/cssc.201901313
ChemSusChem
FULL PAPER
In addition to allowing the kinetic parameters to be verified
from both gaseous and the liquid phases, this approach also
permits the acceptorless nature of the dehydrogenation reaction
to be confirmed. To do so, the gas produced during the
Table 3. Effect of N2 flow on 1-phenylethanol dehydrogenation
catalysed by Pd/C.[a]
Time (h)
0.16
2.5
X (%)[b]
29.6
Y (%)[c]
28.8
S (%)[d]
97.6
C Bal. (%)[e]
100
dehydrogenation of 1-phenylethanol was collected over
a
reaction time of 1 hour (43 mL of gas collected) and analysed
via mass spectrometry (Table 2). As can be seen, compositional
analysis of the collected gas indicated 24.4 mol % of H2 to be
present into the gas mixture. To ensure that the produced H2
51.0
40.8
80.0
100
18
85.2
58.8
69.0
100
[f]
derives from dehydrogenation of the substrate,
a control
2.5
72.7
71.5
98.4
100
experiment was performed, conducting the reaction under
typical reaction conditions, but without the substrate. As
expected, only trace amounts of H2 were detected (0.08 mol %),
indicating 1-phenylethanol to be the source of H2.
[a] Reaction conditions: stainless steel reactor body with 100 mL
pressurised round bottom flask. 0.2 M of 1-phenylethanol in p-xylene, 0.08
mol % of Pd/substrate molar ratio, static N2 atmosphere, 130 °C. [b]
Conversion of 1-phenylethanol. [c] Yield of acetophenone. [d] Selectivity for
acetophenone. [e] Carbon balance calculated including also the amounts of
observed by-products (styrene and ethylbenzene). [f] Reaction performed in
N2 flow (10 mL/min).
The relatively high molar percentage of H2 detected in the
gas phase, in addition to the good catalytic performances
exhibited by the catalyst, strongly indicates that acceptorless
alcohol dehydrogenation can be efficiently catalysed by Pd/C in
N2 atmosphere. Notably, the employment of an inert gas
dramatically reduces the over-oxidation problems typically
observed during aerobic alcohol oxidation,3c leading to higher
selectivity values for the carbonyl compounds. However,
performing the reaction for prolonged periods (18 hours) under
optimised reaction conditions resulted in a drop in selectivity
from 97.6 to 69.0 %, thus, suggesting the presence of side
reactions at elevated levels of conversion and low masses of
catalyst. In parallel to the drop of selectivity, the amount of
substrate converted at extended periods was lower than
anticipated based on the initial rate of reaction (from 29.6 %
conversion at 0.16 hours to only 51.0 % at 2.5 hours). This result
is in agreement to previous reports suggesting the reversibility of
the process,7 i.e. regeneration of the substrate following the
hydrogenation of acetophenone.
Figure 2. Left. Selectivity (S) for; i) acetophenone (squares), ii) ethylbenzene
(circles), and styrene (triangles) as a function of 1-phenylethanol conversion
(X). Right. Time on line profile of selectivity for i) acetophenone (squares), ii)
ethylbenzene (circles), and iii) styrene (triangles). Reaction conditions:
stainless steel reactor body with 100 mL pressurised round bottom flask. 20
mL of a solution 0.2 M of 1-phenylethanol in p-xylene, 0.08 mol % of
Pd/substrate molar ratio, static N2 atmosphere, 130 °C.
Table
2.
Gas
composition
analysis
for
1-phenylethanol
dehydrogenation with and without 1-phenylethanol.[a]
Gas Composition (mol %)
General reaction
N2
H2
O2
74.12
99.20
24.73
0.08
1.15
0.72
Without 1-phenylethanol[b]
Analysis of the liquid phase after reaction was also performed, in
order to allow the by-products produced at high conversion to be
identified. In addition to confirming the absence of benzoic acid,
this analysis revealed that under static N2 atmosphere, the
quantity of acetophenone lost at high conversion was
comparable to the amount of ethylbenzene formed (Figure 2).
The presence of ethylbenzene is in agreement to reactant
stability studies, performed by conducting the hydrogenation of
acetophenone and styrene in an autoclave filled with H2 (2 bar)
[a] Reaction conditions: stainless steel reactor body with 100 mL pressurised
round bottom flask. 0.2 M of 1-phenylethanol in p-xylene, 0.08 mol % of
Pd/substrate molar ratio, static N2 atmosphere,
phenylethanol.
1 hour. [b] no 1-
To study the negative effect of H2 accumulation on the reaction,
experiments were first performed in order to investigate the role
played by gas flow i.e. static N2 atmosphere was replaced with
N2 flow (Table 3), to remove H2 from the reaction mixture
following its formation. Notably, performing the reaction under N2
flow (10 mL/min) dramatically affected the reaction, increasing 1-
phenylethanol conversion and acetophenone yield, and boosting
acetophenone selectivity to 98.4 %, strongly indicating that
excessive quantities of H2 impact selectivity, especially at high
conversion.
and Pd/C as catalyst (Scheme 2). Following
1 hour of
acetophenone hydrogenation, a 1-phenylethanol yield of 71.3 %,
and ethylbenzene yield of 20.3 %, was detected. This indicates
that at high quantities of H2, the reverse hydrogenation,
accompanied by the formation of 1-phenylethanol occurs.
Although no styrene was found into the reaction mixture
(Scheme 2), it may still be involved in the reaction mechanism
as an intermediate with a short lifetime, i.e. this compound may
be formed from the dehydration of 1-phenylethanol, which
rapidly undergoes hydrogenation to ethylbenzene. Therefore,
3
This article is protected by copyright. All rights reserved.

10.1002/cssc.201901313
ChemSusChem
FULL PAPER
stability studies using styrene as substrate were also performed.
During these, quantitative conversion of styrene into
ethylbenzene was reached in only one hour of reaction (Scheme
3). Notably, the formation of ethylbenzene may also arise from
direct hydrogenolysis of 1-phenylethanol and/or acetophenone
catalysed by the Pd-containing catalyst.8 These side reactions
indicate the detrimental effect that excessive quantities of H2
exert on the reaction. Hence, working under conditions that
remove H2 from the reactor, and/or working at low conversion
conditions, is beneficial for the overall reaction selectivity. Based
on these observations, all further kinetic studies were performed
at short reaction times and lower conversion levels, to limit the
contribution of the side reactions to the kinetic parameters and
catalyst optimisation studies. Moreover, experiments were
performed under static conditions to continue to permit
measurement of the gaseous phase of the reaction.
Figure 3. Left. ln of [1-phenyletanol] over time at different temperatures (110-
130°C). Right. Arrhenius plot obtained in a temperature range of 90-130°C.
Reaction conditions: stainless steel reactor body with 100 mL pressurised
round bottom flask. 20 mL of a solution 0.2 M of 1-phenylethanol in p-xylene,
0.08 mol % of Pd/substrate molar ratio, static N2 atmosphere.
Scheme 2. Acetophenone hydrogenation catalysed by Pd/C. Autoclave
reactor (Parr® 5500) with 100 mL glass liner. Reaction conditions: 20 mL of a
0.2 M solution of acetophenone in toluene and 0.08 mol % of Pd/substrate
molar ratio, 130 ˚C. Reaction performed with 2 bars of H2.
Following these kinetic insights, more detailed attention was
paid to the mechanism of the reaction, since very little is known
about acceptorless dehydrogenation chemistry, especially over
heterogeneous catalysts. In order to identify the nature of the
reaction intermediate, a Hammett correlation was obtained,
using substituted benzyl alcohols as substrates, due to the
greater commercial availability of substituted benzylic alcohols.
As can be seen (Figure
4
(Left)), the acceptorless
dehydrogenation of benzyl alcohol and its p-substituted
analogues, 4-methylbenzyl alcohol and 4-chlorobenzyl alcohol,
indicates that electron donating substituents (σ < 1), such as -
CH3, lead to increased activity. On the other hand, electron
withdrawing groups (σ > 1), such as -Cl, exhibit a negative
impact on the reaction rate. The negative slope of the Hammett
plot strongly indicates the formation of a positively charged
reaction intermediate, suggesting the formation of a carbo-
cationic intermediate during the reaction.
Scheme 3. Styrene hydrogenation catalysed by Pd/C. Autoclave reactor
(Parr® 5500) with 100 mL glass liner. Reaction conditions: 20 mL of a solution
0.2 M of styrene in toluene and 0.08 mol % of Pd/substrate molar ratio, 130 ˚C.
Reaction performed with 2 bars of H2.
Additional mechanistic insights on the reaction mechanism
was obtained by deuteration of the benzylic H-atoms of the
substrate (KIE). As can be seen (Figure 4 (Right)), a lower
reaction rate (k) was observed when benzylic C-D bonds are
present over C-H bonds. Since KIEs greater than 1 usually
indicate that cleavage of that bond is involved in the rate
determining step of the reaction, this indicates that cleavage of
the benzylic C-H/D bond may be rate limiting. However, it is
notable that the observed KIE is lower than that typically
observed for reactions limited by the rate of C-H bond cleavage,
although they are in line with the values observed for
Scheme 4. Proposed reaction pathway and proposed intermediate species
for 1-phenylethanol dehydrogenation over Pd/C.
acceptorless
dehydrogenation
over
homogeneous
Pd
Having identified the optimal kinetic regime of the reaction, and
ensuring a negligible contribution of side reactions to the
reaction network, an accurate Arrhenius plot was obtained by
measuring the initial rate of the reaction (k) in the linearity range
true kinetic conditions, an activation energy of 98.8 kJ/mol was
determined, in better agreement to the nature of the reaction
(Figure 3). Extending the Arrhenius analysis over a larger range
of temperature demonstrates the barrier is consistent over a
temperature range of 90-130 ˚C.
complexes.3d,3e The relatively low KIE values observed in these
systems could be due to second order effects – where the H/D
substitution in question is not directly involved in the rate limiting
step – or the involvement of the H/D atom in a different rate
limiting step, such as elimination of the H/D atom from the Pd
surface to regenerate the active site. Nevertheless, the negative
Hammett correlation, coupled with a KIE > 1, indicates that the
acceptorless dehydrogenation reaction possesses some
similarities to classical beta-hydride elimination mechanisms,
often observed during aerobic alcohol oxidation.9
4
This article is protected by copyright. All rights reserved.

10.1002/cssc.201901313
ChemSusChem
FULL PAPER
Figure 5. Conversion (X) of 1-phenylethanol (red bars), yield (Y) of
acetophenone (blue bars) and selectivity (S) for acetophenone (black squares)
for: a) fresh Pd/C (i.e. not treated), Pd/C heated in 5 % H2 in Ar for 2 hours
(ramp rate 10 °C/min) b) at 300 °C, c) at 400 °C, and Pd/C heated under air
for 2 hours (ramp rate 10 °C/min) d) at 200 °C, and e) at 300 °C. Reaction
conditions: stainless steel reactor body with 100 mL pressurised round bottom
flask. 20 mL of a solution 0.2 M of 1-phenylethanol in p-xylene, 0.08 mol % of
Pd/substrate molar ratio, 130 ºC, static N2 atmosphere, 0.25 hours.
Figure 4. Left. Hammett plot obtained from substituted benzyl alcohol
dehydrogenation to acetophenone over Pd/C. Right. Reaction rate for benzyl
alcohol (squares) and α,α-deuterated benzyl alcohol (circles). Reaction
conditions: stainless steel reactor body with 100 mL pressurised round bottom
flask. 20 mL of a solution 0.2 M of substrate in p-xylene, 0.08 mol % of
Pd/substrate molar ratio, 0.16 hours, 130 ºC, static N2 atmosphere.
Pd/C Fresh
Pd/C air 300 °C
Following mechanistic and kinetic evaluation of the system,
preliminary catalyst optimisation studies were performed by
examining the effect of the oxidation state of the metal (Pd) on
the reaction rate. To do so, the catalyst was subjected to a
series of heat treatments in reducing atmosphere (5% H2 in Ar),
or oxidising conditions (air). Notably, although treatments in H2
at 400 °C did not affect the C support, analogous treatment in air
at 400 °C led to decomposition of the support (ESI Figure S5-6).
Therefore, the optimal temperatures were found to be 400 °C for
the H2 treatment and 300 °C for the air treatment. As shown in
Figure 5, although treating the catalyst in air for 2 hours at
300°C was beneficial for the catalytic performance of Pd/C (from
29.5 % conversion to 33.1 % following air treatment), a loss in
selectivity for acetophenone occurred (from 97.3 % to 87.6 %).
An opposite effect was observed calcining the catalyst under
reducing atmosphere, as lower conversion values were
achieved but higher acetophone selectivity observed (> 98 %).
However, we note these differences may also be (partially)
related to the natural conversion vs. selectivity relationship of the
reaction (Figure 2, Left).
Pd/C H2 400 °C
Binding Energy / eV
Figure 6. XPS analysis for: a) fresh Pd/C (i.e. not treated), b) Pd/C heated at
300 °C under air for 2 hours (ramp rate 10 °C/min), and c) Pd/C heated at
400 °C in 5 % H2 in Ar for 2 hours (ramp rate 10 °C/min).
To analyse the Pd speciation, XPS analysis was performed.
As expected, although a mix of oxidation states was found in the
untreated Pd/C catalyst (37.1
% % a
PdII, 62.9 Pd0),
predominance of PdII was found on Pd/C heated in air at 300 °C
(73.9 % PdII, 26.1 % Pd0) whilst mostly Pd0 was found on Pd/C
heated in H2 at 400 °C (1.7 % PdII, 98.3 % Pd0). Interestingly,
although large differences in Pd speciation were observed in this
series of catalysts, each of these materials exhibits high activity
for acceptorless 1-phenylethanol dehydrogenation. This may be
due to both oxidation states being active for this reaction, and/or
interconversion between PdII and Pd0 readily occurring during
To correlate the observed trends in selectivity and activity as
a function of pre-treatment, and hence with differences in Pd
speciation and material composition and structure, selected
catalysts were characterised via X-Ray Diffraction (XRD), X-ray
Photoelectron Spectroscopy (XPS), Transmission Electron
Microscopy (TEM) and Brunauer-Emmett-Teller (BET) analysis.
5
This article is protected by copyright. All rights reserved.

10.1002/cssc.201901313
ChemSusChem
FULL PAPER
the reaction itself. More detailed comparison of the initial rates of
each reaction, achieved by monitoring the gas evolution with
time, shows that lower initial rates (k) are observed for catalysts
possessing higher initial percentages of Pd0 (Table 4), indicating
that Pd0 may catalyse 1-phenylethanol dehydrogenation to a
lower degree than PdII. This is in contrast to the
dehydrogenation of formic acid over the same catalyst, where
metallic Pd0 has routinely been proposed to be the most active
form of Pd.10 Notably, XRD analysis is in good agreement to the
XPS analysis (ESI Figure S5), showing different distribution
between metallic Pd and Pd oxide to be present on these
catalysts, in analogy with the results displayed in Table 4.
Table 4. Correlation between amount of Pd0 and initial rate for 1-
phenylethanol dehydrogenation over Pd/C fresh (i.e. non-treated), and
following different heat treatments.
Figure 8. Particles size distribution for: a) fresh Pd/C (i.e. not treated), b) Pd/C
heated at 300 °C under air for 2 hours (ramp rate 10 °C/min), and c) Pd/C
heated at 400 °C in 5 % H2 in Ar for 2 hours (ramp rate 10 °C/min).
Catalyst
%Pd0 [a]
62.9
k x 10-3 (s) [b]
1.09
In addition to oxidation state, variation in activity may also
arise from several other parameters related to the material
surface and structure, as well as variations of the Pd particle
size. In fact, heat treatment often affects the particles size
distribution of supported nanoparticles, due to agglomeration
(sintering) of the particles.3c Accordingly, TEM analysis was also
performed on this series of catalysts. As can be seen (Figures
7,8), the untreated catalyst possesses a relatively narrow
distribution of Pd nanoparticles with a mean size of 3.5 nm, an
increase in the Pd particle size distribution was found for both
Pd/C Fresh
Pd/C H2 200 °C
Pd/C H2 400 °C
Pd/C air 200 °C
Pd/C air 300 °C
78.1
0.90
98.3
0.82
49.1
1.13
26.1
1.14
[a] Relative percentage of Pd0 on the overall Pd, measured via
deconvolution of XPS spectra [b] Initial reaction rate (ESI Figure S7)
heattreated samples. However, although there is
a net
Pd/C fresh
agglomeration of Pd in both heat treated catalysts, the extent of
agglomeration is comparable in both H2 and air atmospheres,
and is unlikely to explain the different kinetic capabilities of these
samples.
To further investigate structural changes potentially
occurring on the surface of the catalyst, BET analysis was also
performed. However, no major variation to the catalyst surface
area occurs following the heat treatments (ESI Table S1), with
all catalytic materials possessing surface areas of approximately
900 m2 g-1, regardless of the choice of heat treatment.
Pd/C air 300 °C
Alongside activity and selectivity, the durability of
a
heterogeneous catalyst also plays a key role in its potential
applicability. In order to perform an accurate investigation of the
durability of the material, and hence, gain preliminary
understanding of the scalability of the process, the catalytic
performance of Pd/C was studied under continuous flow
conditions using a Plug Flow Reactor (PFR), under otherwise
analogous conditions to the batch experiments. When employed
in continuous mode, Pd/C displayed good activity for 1-
phenylethanol dehydrogenation with acetophenone selectivity >
80 % (Figure 9 and ESI Figure S8) observed. Although some
losses in activity are observed over the first 24 h on stream, the
system reaches steady state conditions thereafter, indicating the
longer term durability of Pd/C for continuous operations. XPS
analysis of the catalyst after 48 h on stream indicates that the
relative content of Pd0 increases from 62.9 % (fresh sample) to
88.6 % during this period of operation, indicating that reduction
of Pd occurs in situ. This may be the reason for the initial loss of
activity over the first 24 h, prior to steady state being achieved,
since Pd0 has been shown to be less intrinsically active than PdII
Pd/C H2 400 °C
Figure 7. TEM images for: a) fresh Pd/C (i.e. not treated), b) Pd/C heated at
300 °C under air for 2 hours (ramp rate 10 °C/min), and c) Pd/C heated at
400 °C in 5 % H2 in Ar for 2 hours (ramp rate 10 °C/min).
6
This article is protected by copyright. All rights reserved.

10.1002/cssc.201901313
ChemSusChem
FULL PAPER
(Table 4). However, we note that several other phenomena,
such as poisoning, pore fouling and active site reorganisation,
may also contribute to the initial drop of activity. However, these
studies indicate that long term dehydrogenation is clearly
feasible, even in the absence of acceptors and bases, and that
the reaction is accompanied by the continuous production of
high purity H2. Notably, this preliminary result yields a maximum
space-time-yield (STY) of 0.683 g(product) mL-1 h-1 being achieved,
which is over two orders of magnitude higher than the ones
calculated for the best catalysts reported in the literature
(Ag/hydrotalcite) to date, indicating the high viability of Pd/C as a
heterogeneous catalyst for alcohol dehydrogenation. In addition
to 1-phenylethanol and substituted benzyl alcohols, preliminary
studies on the general applicability of the system were made by
Commercially available 5 wt. % Pd/C is demonstrated to be a
suitable catalyst for the acceptorless dehydrogenation of
alcohols, such as 1-phenylethanol, in inert atmosphere (N2). In
addition to displaying STY values over two orders of magnitude
higher than the ones found in literature, the acceptorless nature
of the reaction results in the co-production of molecular H2. The
intrinsic activity of the catalyst (TOF) was found to be amongst
the highest in the literature, reaching a maximum of 1475 h-1.
Full kinetic analysis of all the reaction products and kinetic
parameters, including the gaseous product, was achieved
following design of a novel batch reactor. Accurate kinetic
analyses were performed, and an activation energy of 98.8
kJ/mol was found, alongside a negative Hammett correlation
and a Kinetic Isotope Effect > 1, indicating that the reaction
possesses mechanistic similarities to beta-hydride elimination
aerobic oxidation mechanisms. Preliminary structure-activity
relationships, performed with kinetics, XPS, XRD, TEM and BET,
reveal Pd/C calcined in air at 300 °C to be the optimal catalyst
for acceptorless dehydrogenation. Experiments in the
continuous regime also the durability of the catalyst over 48
hours on stream, resulting in space-time-yields up to 0.683
g(product) mL-1 h-1, over two orders of magnitude higher than the
ones found in literature up to date. Alongside its high TOF, the
extremely high space-time-yield of the system clearly
demonstrate the favourability of the system for acceptorless
dehydrogenation.
substituting 1-phenylethanol for
a
smaller, non-aromatic
substrate, 2-butanol. Interestingly, Pd/C was also able to
continuously perform 2-butanol dehydrogenation to 2-butanone
in the continuous regime, albeit at slightly higher temperature
(200 °C). Alongside 1-phenylethanol and the substituted
benzylic alcohols, this substrate further indicates the general
suitability of Pd/C-catalysed dehydrogenation in continuous
operational mode.
Experimental Section
All substrates were obtained from commercial sources and used without
purification. Fresh Pd/C (5 wt. %) was obtained from Sigma Aldrich® and
used without further modifications.
Heat treatments: All heat treatments were performed in a tube furnace.
In a typical heat treatment, Pd/C was loaded into a calcination boat and
subjected to heat treatment at temperatures varying between 200 – 400
˚C at a ramp rate of 10 ˚C/min under flowing air. For reduction, an
identical procedure was performed albeit under 5% H2 in Ar.
Experimental procedure for 1-phenylethanol dehydrogenation in
batch
using
a
borosilicate
glass
flask:
Pd/C-catalysed
dehydrogenation reactions were performed at a range of temperatures
between 100-120 °C using a three neck 100 mL round bottom flask. The
flask was charged at room temperature with 20 mL of a solution of 1-
phenylethanol in p-xylene (0.2 M) and various amounts of Pd/C (5 Pd
Figure 9. Continuous 1-phenylethanol dehydrogenation reactions in PFR.
Relative performance for Pd/C-catalysed 1-phenylethanol dehydrogenation
(Xt/X0) over time on stream. For details see Experimental Section.
wt.
% from Sigma Aldrich®) in the range 0-1.25 mol % Pd/1-
phenylethanol molar ratio (0-106.8 mg of Pd/C). The round bottom flask
was equipped with a condenser and glass cap adaptors and connected
to a N2 line. The N2 flow was controlled using a mass flow controller. For
all reactions, the flask was purged with a N2 flow of 50 mL/min for 6 min.
The flow was then decreased to 10 mL/min of N2 and the flask was
introduced in the oil bath and pre heated at the desired temperature. The
reaction was then initiated by switching on the magnetic stirring (750
rpm). After the reaction, the flask was cooled down to room temperature,
the catalyst was recovered by filtration and each sample was prepared
by adding 100 µL of reaction solution to 900 µL of a solution of biphenyl
in toluene (0.01 M), which acts as external standard. The samples were
analysed by a Gas Chromatograph (GC) (Agilent 7820) equipped with a
25 m CP-Wax 52 CB capillary column and a Flame Ionisation Detector
(FID), with He as carrier gas (5 mL min-1). The concentrations of
acetophenone and 1-phenylethanol were obtained by previous GC
calibration with the respective standards.
Table 4. Comparitive performance of Pd/C and Ag/hydrotalcite.
Catalyst
Ag/hydrotalcite5c
Pd/C
Substrate
1-phenylethanol
1-phenylethanol
2-butanol[c]
STY (g(product) mL-1 h-1)[a]
2.2x10-3[b]
0.683
Pd/C
0.010
[a] STYs calculated at maximum conversion as grams of reactant converted
per mL (reactor volume), per hour. Volume of catalyst bed used as the
reactor volume [b] Only the liquid volume was used as the reactor volume.
[c] Reaction condition identical to the ones used for 1-phenylethanol, albeit
at 200 °C.
Experimental procedure for 1-phenylethanol dehydrogenation in
batch using a stainless steel reactor body with a pressurised
connection vessel (ESI Figure S4): Pd/C-catalysed dehydrogenation
reactions were performed at a range of internal temperatures between
Conclusions
7
This article is protected by copyright. All rights reserved.

10.1002/cssc.201901313
ChemSusChem
FULL PAPER
110-130 °C using a one neck 100 mL Ace round-bottom pressure flask
with an Ace-Thred 15 PTFE front-seal plug. The flask was charged at
room temperature with 20 mL of a solution of 1-phenylethanol in p-xylene
(0.2 M) and 7.2 mg of 5 wt. % Pd/C (0.08 mol % Pd). The flask was
connected to a stainless steel body reactor, placed into an oil bath,
connected to a burette and to a N2 line with a mass flow control (see ESI
Figure S4). For all reactions, the flask was purged with a flow of 50
mL/min of N2 for 10 min at room temperature. When H2 measurement
was required, the flow was then stopped for the reaction to be carried out
under static N2 atmosphere. For experiments with N2 flow, the reaction
was carried out using a N2 flow rate of 10 mL/min. The flask was
introduced into the oil bath and pre heated at the desired temperature for
15 min. The reaction was then initiated by switching on the magnetic
stirring (750 rpm). For kinetic studies, the gas produced during the
reactions was collected in the burette allowing the quantification of the H2
produced. Aliquots of gas sample were then collected and analysed by
Mass Spectrometry (MS). Analysis of the liquid samples was performed
analogously to the procedure described in the previous section (General
Transmission Electron Microscopy (TEM): Samples for examination
by TEM were prepared by dispersing the catalyst powders in high purity
ethanol using ultra-sonication. 40 µl of the suspension was dropped on to
a holey carbon film supported by a 300 mesh copper TEM grid before the
solvent was evaporated. The samples for TEM were then examined
using a JEOL JEM 2100 TEM model operating at 200 kV.
X-Ray Photoelectron Spectroscopy (XPS): XPS analysis was
performed on a Thermo Scientific K-Alpha+ spectrometer. Samples were
analysed using a monochromatic Al X-ray source operating at 72 W (6
mA x 12 kV), with the signal averaged over an oval-shaped area of
approximately 600 x 400 microns. Data was recorded at pass energies of
150 eV for survey scans and 40 eV for high resolution scan with a 1 eV
and 0.1 eV step size respectively. Charge neutralisation of the sample
was achieved using a combination of both low energy electrons and
argon ions (less than 1 eV) which gave a C(1s) binding energy of 284.8
eV. All spectra were analysed using CasaXPS (v2.3.17 PR1.1) using
Scofield sensitivity factors and an energy exponent of -0.6.
procedure for 1-phenylethanol dehydrogenation in batch using
a
borosilicate glass flask). Theoretical gas volumes was calculated from
the ideal gas equation (PV=nRT) considering n=moles of acetophone
produced, T = 22 ºC, P=1 atm.
Acknowledgements
CH gratefully appreciates the support of The Royal Society, for provision
of a University Research Fellowship (UF140207) and further research
funding through the Grand Challenge Research Fund (CHG\R1\170092).
Hot filtration: During the first part of the hot filtration experiment, a
general reaction with the solid catalyst, as described earlier was initiated.
After 5 min of reaction, the reaction mixture was withdrawn, and the solid
catalyst was removed by filtration. The filtered reaction mixture was then
added into another flask equipped with a magnetic stirrer, purged with a
flow of 50 mL/min of N2 for 10 min at room temperature and heated again.
The reaction was then continued by switching on the magnetic stirring,
this time in the absence of the solid catalyst (dotted line Figure S2). After
an appropriate length of time, the reaction solution was then analysed
again to determine any differences in substrate conversion or product
yield in the absence of the solid catalyst.
General procedure for 1-phenylethanol dehydrogenation in Plug
Flow Reactor (PFR): Continuous 1-phenylethanol dehydrogenation
reactions were performed in a home-made, tubular, stainless steel, PFR.
Reactant delivery (0.19 M of a solution 1-phenylethanol in toluene) was
performed by an HPLC pump. The catalyst, 5 wt. % Pd/C (0.08 g) was
placed in between two plugs of quartz wool and densely packed into a
1/4 inch stainless steel tube (3.8 mm ID), and a frit (0.5 mm) was placed
at the end of the bed to avoid any loss of material. A contact time of 0.45
min was employed. The reactor temperature was controlled by
immersion in an oil bath, and the pressure was controlled by means of a
backpressure regulator. Aliquots of the reaction solutions were taken
Keywords: alcohol oxidation • dehydrogenation • hydrogen
storage • heterogeneous catalysis • nanoparticles
[1]
a) R. A. Sheldon, J. K. Kochi in Metal-Catalyzed Oxidations of Organic
Compounds, Academic Press, New York, 1981; b) K. Banerji in Organic
Reaction Mechanisms (Ed.: A. C. Knipe), Wiley, Chichester, 2012, pp.
79–152;
[2]
a) D. G. Lee, U. A. Spitzer, J. Org. Chem. 1970, 35, 3589–3590; b) J.
Muzart, Chem. Rev. 1992, 92, 113–140.
[3]
a) K. Mori, T. Hara, T. Mizugaki, K. Ebitani, K. Kaneda, J. Am. Chem.
Soc. 2004, 126, 10657–10666; b) K. Yamaguchi, N. Mizuno, Chem. Eur.
J. 2003, 9, 4353–4361; c) T. Mallat, A. Baiker, Chem. Rev. 2004, 104,
3037–3058; d) B. A. Steinhoff, S. S. Stahl, Org. Lett. 2002, 4, 4179–
4181; e) J. A. Mueller, M. S. Sigman, J. Am. Chem. Soc. 2003, 125,
7005–7013.
[4]
a) S. Chakraborty, P. O. Lagaditis, M. Förster, E. A. Bielinski, N. Hazari,
M. C. Holthausen, W. D. Jones, S. Schneider, ACS Catal. 2014, 4,
3994−4003; b) R. Kawahara, K.-i. Fujita, R. Yamaguchi, J. Am. Chem.
Soc. 2012, 134, 3643−3646; c) C. Gunanathan, D. Milstein, Science
2013, 341, 249.
periodically from
a sampling valve placed after the backpressure
regulator, and were analysed in the same manner as the batch reactions.
Gas Chromatography: For liquid sample analysis, a GC (Agilent 7820)
equipped with a 25 m CP-Wax 52 CB column and an FID (at 250 °C),
was employed, using He (5 mL/min) as carrier gas. Quantification of the
analytes was carried out against biphenyl as external standard. For all
samples the following temperature programme was used: Hold at 40 °C
for 5 minutes, ramp at 20 °C/min to 210 °C and hold at 210 °C for 5
minutes.
Mass Spectrometry (MS): Gas samples were analysed by a mass
spectrometer (Hiden Analytical Quadrupole Gas Analyzer (QGA))
equipped with an inert quartz capillary with a consumption rate of 16
mL/min and Faraday electron multiplier detector capable of detecting
concentrations between 0.1 ppm to 100 %. QGA Professional Software
was used to quantify the partial pressures as a function of the M/Z ratio.
Powder X-Ray Diffraction (pXRD): pXRD patterns were recorded on a
PANalytical X’Pert PRO X-ray diffractometer, with a Cu Kα radiation
source (40 kV and 30 mA) and Ni filter. Diffraction patterns were
recorded between 5-80° 2θ at a step size of 0.0167° (time/step = 150 s,
total time = 1.5 h).
[5]
a) J. H. Choi, N. Kim, Y. J. Shin, J. H. Park, J. Park, Tetrahedron Lett.
2004, 45, 4607–4610; b) W. H. Kim, I. S. Park, J. Park, Org. Lett. 2006,
8, 2543–2545; c) T. Mitsudome, Y. Mikami, H. Funai, T. Mizugaki, K.
Jitsukawa, K. Kaneda, Angew. Chem. Int. Ed. 2008, 47, 138–141; d) T.
Mitsudome, Y. Mikami, K. Ebata, T. Mizugaki, K. Jitsukawa, K. Kaneda,
Chem. Commun. 2008, 4804–4806; e) W. Fang, J. Chen, Q. Zhang, W.
Deng, Y. Wang, Chem. – Eur. J. 2011, 17, 1247–1256; f) J. Yi, J. T.
Miller, D. Y. Zemlyanov, R. Zhang, P. J. Dietrich, F. H. Ribeiro, S.
Suslov, M. M. Abu-Omar, Angew. Chem. Int. Ed. 2014, 53, 833–836; g)
K. Shimizu, K. Sugino, K. Sawabe, A. Satsuma, Chem. – Eur. J. 2009,
15, 2341–2351; h) K. Shimizu, K. Kon, M. Seto, K. Shimura, H.
Yamazaki, J. N. Kondo, Green Chem. 2013, 15, 418–424; i) K. Shimizu,
K. Kon, K. Shimura, S. S. M. A. Hakim, J. Catal. 2013, 300, 242–250; j)
K. Kon, S. M. A. H. Siddiki, K. Shimizu, J. Catal. 2013, 304, 63–71; k) H.
Liu, H.-R. Tan, E. S. Tok, S. Jaenicke, G.-K. Chuah, Chem. Cat. Chem.
2016, 8, 968–975; l) M. Hosseini-Sarvari, T. Ataee-Kachouei, F. Moeini,
Mater. Res. Bull. 2015, 72, 98–105; m) H. Chen, S. He, M. Xu, M. Wei,
D. G. Evans, X. Duan, ACS Catal. 2017, 7, 2735−2743; n) F. Wang, R.
Shi, Z. Q. Liu, P. J. Shang, X. Pang, S. Shen, Z. Feng, C. Li, W. Shen,
ACS Catal. 2013, 3, 890−894; o) J. Muzart, Tethraedron, Tetrahedron
2003, 59, 5789–5816.
N2 physisorption analysis: The surface area of the catalysts was
measured using N2 physisorption. Each sample was evacuated for 3
hours at 120 °C using a Quantachrome Quadrasorb SI instrument.
Following evacuation, surface area analysis was performed and
subsequently calculated using BET theory over the range P/P0 = 0-1.
8
This article is protected by copyright. All rights reserved.

10.1002/cssc.201901313
ChemSusChem
FULL PAPER
[6]
[7]
[8]
a) J. A. Turner, Science 2004, 305, 972-974; b) X. Zou, Y. Zhang,
[9]
a) A. Abad. A. Corma. H. García, Chem. Eur. J. 2008, 14, 212–222; b)
O. Pàmies, J.-E. Bläckvall, Chem. Eur. J. 2001, 7, 5052-5058; c) M.
Gómez-Gallego, M. A. Sierra, Chem. Rev. 2011, 111, 4857-4963.
Chem. Soc. Rev. 2015,44, 5148-5180; c) S. Dunn, Int. J. Hydrogen
Energ. 2002, 27, 235-264.
a) E. Buckley, E. F. G. Herington, Trans. Faraday Soc. 1965, 61, 1618-
1625; b) H. Esaki, R. Ohtaki, T. Maegawa, Y. Monguchi, H. Sajiki. J.
Org. Chem. 2007, 72, 2143–2150.
[10] a) F. Sanchez, M. H. Alotaibi, D. Motta, C. E. Chan-Thaw, A.
Rakotomahevitra, T. Tabanelli, A. Roldan, C. Hammond, Q. He, T.
Davies, A. Villa, N. Dimitratos, Sustainable Energy Fuels, 2018, 2,
2705-2716; b) F. Sanchez, D. Motta, L. Bocelli, S. Albonetti, A. Roldan,
C. Hammond, A. Villa, N. Dimitratos, C 2018, 4, 26; c) F. Sanchez,
D. Motta, A. Roldan, C. Hammond, A. Villa, N. Dimitratos, Top. Catal.
2018, 61, 254-266.
a) A. P. G. Kieboom, J. F. De Kreuk, H. Van Bekkum, J. Catal. 1971, 20,
58-66; b) K. Hattori, H. Sajiki, K. Hirota, Tetrahedron, 2001, 57, 4817-
4824.
9
This article is protected by copyright. All rights reserved.

10.1002/cssc.201901313
ChemSusChem
FULL PAPER
Entry for the Table of Contents (Please choose one layout)
Layout 1:
FULL PAPER
The acceptorless dehydrogenation of
alcohols is achieved with
G. Nicolau, G. Tarantino, C. Hammond*
Page No. – Page No.
heterogeneous catalysts and
continuous reactors, resulting in the
continuous production of carbonyl
compounds and molecular hydrogen.
Acceptorless alcohol
dehydrogenation catalysed by Pd/C
Layout 2:
FULL PAPER
Author(s), Corresponding Author(s)*
((Insert TOC Graphic here; max. width: 11.5 cm; max. height: 2.5 cm))
Page No. – Page No.
Title
Text for Table of Contents
10
This article is protected by copyright. All rights reserved.