Catalytic cleavage of lignin β-O-4 link mimics using copper on alumina and magnesia-alumina

Article abstract of DOI:10.1039/c3gc37056a

Copper on γ-alumina and on mixed magnesia-alumina, Cu/MgO-Al ₂O₃, catalyse the hydrodeoxygenation (HDO) of β-O-4 lignin-type dimers, giving valuable aromatics. The typical selectivity to phenol is as high as 20%. By changing the support's acidity we can modify the dispersion of copper. Interestingly, more HDO occurs with larger copper agglomerates than with finely dispersed particles. The presence of copper also increases the selectivity of the HDO cleavage. Three different pathways are hypothesized for the reaction on the catalyst surface. Thus, copper activates ketones more and especially more selective towards cleavage than their corresponding alcohols. DFT calculations of bond dissociation energies correlate well with this experimental observation. Excitingly, ethylbenzene is formed in proportional amounts to phenol, showing that these catalysts can reduce the oxygen content of lignin-type product streams. Considering its low price and ready availability, we conclude that copper on alumina is a promising alternative catalyst for lignin depolymerization.

Full text of DOI:10.1039/c3gc37056a

Green Chemistry
View Article Online
View Journal
PAPER
Catalytic cleavage of lignin β-O-4 link mimics using
copper on alumina and magnesia–alumina†
Cite this: DOI: 10.1039/c3gc37056a
Zea Strassberger, Albert H. Alberts, Manuel J. Louwerse, Stefania Tanase* and
Gadi Rothenberg*
Copper on γ-alumina and on mixed magnesia–alumina, Cu/MgO–Al2O3, catalyse the hydrodeoxygena-
tion (HDO) of β-O-4 lignin-type dimers, giving valuable aromatics. The typical selectivity to phenol is as
high as 20%. By changing the support’s acidity we can modify the dispersion of copper. Interestingly,
more HDO occurs with larger copper agglomerates than with finely dispersed particles. The presence of
copper also increases the selectivity of the HDO cleavage. Three different pathways are hypothesized for
the reaction on the catalyst surface. Thus, copper activates ketones more and especially more selective
towards cleavage than their corresponding alcohols. DFT calculations of bond dissociation energies corre-
late well with this experimental observation. Excitingly, ethylbenzene is formed in proportional amounts
to phenol, showing that these catalysts can reduce the oxygen content of lignin-type product streams.
Considering its low price and ready availability, we conclude that copper on alumina is a promising
alternative catalyst for lignin depolymerization.
Received 11th September 2012,
Accepted 18th January 2013
DOI: 10.1039/c3gc37056a
www.rsc.org/greenchem
Introduction
Woody biomass holds the key for reducing our dependency on
fossil carbon sources.1,2 The majority of nonfood-derived
biomass is lignocellulose, a giant matrix of hemicellulose, cellu-
lose and lignin. Currently, two approaches are used for con-
verting this into liquid fuels and bulk chemicals: acid-
catalyzed treatment and high-temperature pyrolysis/
gasification.3–5 But both of these routes share the same
problem: the breaking down of the valuable functional groups
in lignin and cellulose. Since 2000, several methods have been
published regarding transforming cellulose to more valuable
chemicals such as glucose, sorbitol and HMF.6–8 However,
there is still no viable process for converting lignin into chemi-
cals, even though some industrial sectors produce massive
amounts of lignin as a by-product.9 The pulp and paper indus-
try, for example, produced in 2008 over 40 million tons of
lignin, 95% of which was simply burned to generate energy.10
Burning lignin is wasteful, because it is the most abundant
natural resource of aromatic compounds. Fig. 1 shows a
typical lignin fragment. Phenylpropane monomers have been
identified as the major units of typical lignin (it is a random
Fig. 1 A typical lignin fragment (lignin is a random biopolymer) showing four
β-O-4 linkages highlighted by dashed rectangles.
Van’t Hoff Institute for Molecular Sciences, University of Amsterdam, Science Park
biopolymer), linked together through C–O bonds of α- and
β-arylalkyl ethers.11,12 The β-O-4 linkages (highlighted dashed
frames in Fig. 1) account for roughly 50% of all the linkages in
lignin.13,14 Cleaving them selectively would give smaller frag-
ments, while preserving the aromatic groups.
904, 1098 XH Amsterdam, The Netherlands. E-mail: s.grecea@uva.nl,
g.rothenberg@uva.nl; http://hims.uva.nl/hcsc; Fax: +31 20 525 560
†Electronic supplementary information (ESI) available: 1H NMR spectra of 1 and
2, and surface area measurements (BET) of both catalysts. See DOI:
10.1039/c3gc37056a
This journal is © The Royal Society of Chemistry 2013
Green Chem.

View Article Online
Paper
Our aim, thus, is finding a selective catalytic alternative for
Green Chemistry
The X-ray diffraction patterns of Cu/γ-Al2O3 and CuO/
converting lignin into high-value aromatics. The challenge is not γ-Al2O3 show broad peaks for the alumina, indicating small
so much in depolymerising lignin, but rather in finding a catalyst crystallites. However, the copper peaks are very sharp, so the
that will selectively cleave the β-O-4 linkages while preserving the clusters are rather large (Fig. 2). These reflect the formation
aromaticity. Here we will show that Cu/γ-Al2O3 can do this job.
and agglomeration of crystalline Cu and CuO.30 Conversely,
the XRD patterns of Cu/MgO–Al2O3 reveal much more dis-
persed copper particles, showing broad Cu(0) and CuO peaks.
Note that Feng et al. reported that the presence of MgO can
lead to a wider dispersion of the metal on the surface of the
support.31,32
Using TPR data (see Fig. 3) to approximate the reduced
metal content by measuring the hydrogen consumption, we
calculated that there was 10.6 wt% of copper on the alumina
and 10.4 wt% of copper on the mixed support, magnesia–
alumina. Nonetheless, comparing the two graphs shows two
main differences: the peak shape and the temperature of
reduction. For alumina we can identify two distinguished
peaks and a temperature of reduction ranging from 230 to
250 °C. This corresponds to the reduction of Cu(II). The first
peak is assigned to CuO clusters, whilst the second one is due
to dispersed CuO particles on the γ-Al2O3 surface.33
Results and discussion
To start with, we selected appropriate β-O-4 linkage analogues.
These must be similar enough to the real lignin so that the
results are relevant, yet still simple enough for carrying out
meaningful experiments on laboratory scale. While previous
research focused on small monomers and alcohol–ether-type
dimers,15–17 we opted for ketone–ether dimers, which are also
important products of lignin depolymerisation.18,19 Thus, we
selected 2-phenoxy-1-phenylethanone 1 and 2-phenoxy-1-phe-
nylethanol 2.20 These were synthesized following the procedure
of Britt et al. (eqn (1) and (2)).21
In the case of Cu/MgO–Al2O3, the XRD pattern showed
finely dispersed copper particles but the reduction tempera-
ture was higher than for CuO/γ-Al2O3 (see Fig. 3). This indi-
ð1Þ cates a stronger interaction between copper oxide and the
MgO–Al2O3 support, making the CuO particles less accessible
ð2Þ
For catalysts, we focused on copper and copper oxides sup-
ported on γ-Al2O3 and MgO–Al2O3. Apart from their economic
and environmental advantages over heavy/noble metals,22
copper catalysts were also reported for hydro-deoxygenation
reactions.23–25 Allegrini et al.26 showed the potential of reduc-
tive deoxygenation using copper on different supports, where
aromatic ketones were fully deoxygenated to their methylene
analogues. More recently, Sittisa et al.27,28 reported that
surface copper interacts preferentially with carbonyl groups
rather than with aromatic rings. This was explained in terms
of the preferred adsorption mode on Cu, η1(O)-carbonyl, and
the relatively weak interaction of copper with carbon–carbon
double bonds.27 Therefore, we reasoned that copper sites
might help to retain the aromaticity of our products.
As supports, we used the acidic/basic combination of mag-
nesia–alumina, details of which are published elsewhere,29
and pure γ-alumina. We thus synthesised CuO/γ-Al2O3 and
CuO/MgO–Al2O3.
Fig. 2 X-ray diffraction patterns of CuO/γ-Al2O3 and Cu/γ-Al2O3 (top) and
CuO/MgO–Al2O3 and Cu/MgO–Al2O3 (bottom).
Green Chem.
This journal is © The Royal Society of Chemistry 2013

View Article Online
Green Chemistry
Paper
Table 1 Bond dissociation energies (BDEs) of 1 and 2 for pathways A and
B. Energies are given in kJ mol−1
Reactant
Pathway
BDE
A
B
265
253
A
B
284
201
Fig. 3 Temperature programmed reduction (TPR) plots for CuO/γ-Al2O3 (gray
line) and CuO/MgO–Al2O3 (black line) using hydrogen as a reducing agent.
for reduction. This is in line with the temperature needed for
calcination: 500 °C (2 °C min−1) for 4 h.34
Interestingly, the alcohol 2 is less reactive than its ketone
analogue. Other studies on several β-O-4 models reported a
reduction of the bond dissociation energy (BDE) for oxidized
species compared to their alcohol analogues.35,36 We calcu-
lated BDEs with DFT for our reactants as well, showing that
indeed the ether bond is weaker in the ketone than in the
alcohol (Table 1). As BDEs only explain thermal non-catalytic
cleavage, we also studied the cleavage of protonated molecules
(the cleavage is catalysed by alumina, the activity of which is
usually explained by its acidity). However, these calculations
were severely hindered by reorganisation of bonds, and there-
fore no numbers are reported here. Nevertheless, we do
observe that both for the ketone and the alcohol the ether
bond (pathway B) is activated by protonation of the ether. In
the case of the alcohol, though, an immediate reorganisation
occurs resulting in oligomerisation rather than cleavage. This
fits strikingly well with the oligomerisation observed
experimentally.
ð3Þ
In a typical reaction (eqn (3)), a solution of 2-phenoxy-1-
phenylethanone 1 or 2-phenoxy-1-phenylethanol 2 in ethyl
acetate was stirred under 25 bar of H2 at 150 °C in a stainless
steel autoclave with 2 wt% catalyst without prior reduction of
the catalyst (see the detailed procedure in the Experimental
section). To also investigate the role of Cu(0), the catalyst was
pre-reduced in H2 for 2 h at 300 °C (the reduction temperature
was selected from TPR experiments, see Fig. 3). Reaction pro-
gress was monitored by GC. Note that obtaining a quantitative
mass balance in HDO reactions is particularly difficult.16 Here
we succeeded in quantifying typically 85–95% with careful cali-
bration and rigorous low-temperature quenching of the reac-
tion mixture.
The main products were monoaromatics, (see Tables 2 and
3) plus typically 3–18% of oligomers. When using the ketone 1,
we also observed a significant amount of reduction to the
alcohol 2. Using Cu/γ-Al2O3 and Cu/MgO–Al2O3 gives phenol
and ethylbenzene as the main monoaromatic products. This
indicates that the cleavage occurs mostly at the C–O–aryl bond
(pathway B) and that HDO is the main reaction route. Introdu-
cing basic sites on the support hinders HDO (the selectivity to
ethylbenzene drops considerably; see for example the last two
entries in Tables 2 and 3).
A side effect increasing the difference in reactivity between
the alcohol and ketone may be the three-dimensional shape of
the molecules. Modelling the ketone and the corresponding
alcohol in the gas phase, we see that the ketone is planar,
which allows both oxygens to adsorb simultaneously. The
alcohol, however, is twisted (see Fig. 4) making adsorption
more difficult. Constraining the alcohol into a planar shape
costs 12 kJ mol−1
.
Nichols et al.37 reported the oxidation of the alcohol dimer
2 to the ketone dimer 1 via a well-known Ru-dehydrogenative
equilibrium. Under our conditions (substrate 2 and 25 bar of
hydrogen pressure), this equilibrium is not observed, as
neither dimer 1 nor acetophenone 4 were detected by GC
analysis. To rule out the influence of the thermal cleavage in
the absence of hydrogen, we ran a series of control exper-
iments with argon pressure. These reactions gave less than 5%
conversion. The main products were high-molecular-weight
oligomers, with no oxidation of dimer 2 to dimer 1. Thus, at
150 °C an external hydrogen source is required to cleave the
β-O-4 linkage via hydrogenolysis.
A series of blank experiments starting from 1, both in the
absence of any catalyst and with only the oxide support (but
This journal is © The Royal Society of Chemistry 2013
Green Chem.

View Article Online
Paper
Green Chemistry
Table 2 Product distribution for the conversion of 2-phenoxy-1-phenylethanone 1
4
5
6
7
8
9
Oligomers
2
3
Catalyst
Reactant
Conv.%
Mass balance %
Yielda %
No catalyst
γ-Al2O3
4.71
38.7
30.0
96.7
99.1
97.7
95.3
91.4
89.0
88.1
94.4
81.5
0.0
0.0
0.0
45.2
50.3
24.5
3.3
0.0
0.5
0.6
1.8
5.6
5.1
0.0
0.0
1.2
4.3
1.9
0.0
0.0
0.0
6.6
2.6
9.1
8.5
0.0
0.0
0.0
0.0
2.2
0.6
0.6
0.0
0.0
0.0
0.0
0.0
0.9
7.8
5.6
6.1
18.1
13.6
18.8
12.2
17.3
9.5
0.4
0.6
0.7
0.5
0.0
MgO–Al2O3
CuO/γ-Al2O3
b
b
CuO/MgO–Al2O3
Cu/γ-Al2O3
c
14.1
21.9
c
Cu/MgO–Al2O3
82.3
87.5
34.4
10.2
1.6
0.1
4.3
0.0
4.3
5.3
16.3
a Yields determined by GC analysis (chlorobenzene is the external standard). b Standard reaction conditions: 0.120 mg reactant 1 in 10 mL EtOAc;
25 bar H2; 150 °C; 21 h, 2 wt% catalyst (amount of copper relative to 1) and with a minimal TON of 48–50. Catalyst was used without prior
reduction and no inert conditions during reaction process. c Standard reaction conditions: 0.120 mg reactant 1 in 10 mL EtOAc; 25 bar H2;
150 °C; 21 h, 2 wt% catalyst (amount of copper relative to 1) and with a minimal TON of 42–48. Prior to the experiments, all catalysts were
reduced at 300 °C under a flow of H2 for 2 h, the solvent was purged for 2 h with N2 and autoclaves were purged thrice.
Table 3 Product distribution for the conversion of 2-phenoxy-1-phenylethanol 2
5
8
9
Oligomers
3
Catalyst
Reactant
Conv. %
Mass balance %
Yielda %
No catalyst
γ-Al2O3
<2
98.0
89.5
94.8
95.6
87.5
93.0
<1
2.5
5.0
3.6
10.9
21.7
0.0
0.4
0.4
0.4
0.0
0.1
0.0
1.5
3.9
9.7
4.1
2.4
0.0
0.0
0.0
2.4
0.0
<1
25.7
23.4
16.8
20.6
52.5
18.2
11.8
8.7
2.6
5.2
MgO–Al2O3
CuO/γ-Al2O3
b
b
CuO/MgO–Al2O3
Cu/γ-Al2O3
c
19.1
c
Cu/MgO–Al2O3
47.0
95.0
24.7
0.1
1.3
8.3
9.1
a Yields determined by GC analysis (chlorobenzene is the external standard). b Standard reaction conditions: 0.120 mg reactant 2 in 10 mL EtOAc;
25 bar H2; 150 °C; 21 h, 2 wt% catalyst (amount of copper relative to 2) and with a minimal TON of 8–10. Catalyst was used without prior
reduction and no inert conditions during reaction process. c Standard reaction conditions: 0.120 mg reactant 2 in 10 mL EtOAc; 25 bar H2;
150 °C; 21 h, 2 wt% catalyst (amount of copper relative to 2) and with a minimal TON of 23–26. Prior to the experiments, all catalysts were
reduced at 300 °C under a flow of H2 for 2 h, the solvent was purged for 2 h with N2 and autoclaves were purged thrice.
Fig. 4 DFT-optimised 3D structures of the ketone 1 (top left) and the alcohol 2
(top right) and a possible approach of these to the catalyst surface (bottom).
no copper), showed no traces of the alcohol 2. This affirms
that the reduction of 1 to 2 requires the copper active site.
For both 1 and 2, we envisage a two-step process. To illus-
trate this hypothesis, we propose the following mechanism for
dimer 1 (Scheme 1). Alumina-catalysed hydrogenolysis of the
dimer’s C–O(aryl) bond (pathway B) occurs first, giving aceto-
phenone and phenol. This is followed by copper-catalysed
hydrodeoxygenation of the carbonyl group to the
Scheme 1 Reaction pathways for the β-O-4 cleavage of dimer 1.
Green Chem.
This journal is © The Royal Society of Chemistry 2013

View Article Online
Green Chemistry
Paper
corresponding ethylbenzene. The presence of benzaldehyde
and anisole can be explained by the hydrogenolysis of the
Conclusion
OC–CH2O(aryl) bond (pathway A). Then, benzaldehyde is Copper particles supported on γ-alumina catalyse the scission of
rapidly oxidized by air to benzoic acid (the latter is indeed lignin-type β-O-4 linkages under HDO conditions, yielding
absent under inert conditions, confirming that oxygen is phenol and ethylbenzene in substantial amounts. Using magne-
needed38). If both bonds A and B are cleaved, methane may sia–alumina as the support increases the dispersion of copper,
form. When Cu or CuO is present, for dimer 1 the hydrogen- yet lowers the selectivity towards HDO. For industrial appli-
ation of the ketone can occur as a third pathway. Dimer 2 can cations, the catalyst price/performance ratio is the key criterion,
also split under influence of the alumina, but this only leads and the main challenge in this case is catalyst stability. Working
to oligomerisation. When metallic copper is present, dimer 2 with real lignin depolymerisation feeds means dealing with
can be split into phenol and 8, explaining the smaller amount sulphur in the feedstock, as well as rapid deactivation by coke for-
of dimer 2 in the product mixture when starting from 1 and mation and potential poisoning by water. Importantly, our sup-
using the Cu/Al2O3 catalyst.
ported copper catalysts are cheap and readily available. As such,
Excitingly, up to 22% yield of ethyl benzene was identified they open a practical route for decreasing the oxygenated content
using a reduced copper catalyst and removing air/oxygen from of lignin depolymerisation streams while keeping the aromatic
the reaction mixture (see the experimental procedure for both rings intact, a key hurdle for efficient biomass conversion.41
1 and 2). This is in agreement with the two-step route outlined
above. The experiments gave roughly proportionate yields of
phenol and ethylbenzene for both 1 and 2. Looking at the XRD
Experimental section
Materials, instrumentation and computational methods
pattern, we conclude that HDO occurs more readily with larger
copper agglomerates than on finely dispersed sites.30 The
highest selectivity towards phenol and ethylbenzene is Unless otherwise noted, all chemicals were purchased from
obtained using plain alumina as a support, which also has the commercial sources and used as received. The MgO–Al2O3 was
largest copper clusters. In contrast, when we used unreduced a gift sample from Eurosupport.42
catalysts, copper is involved mainly in the reduction of 1 to 2,
X-Ray diffraction (XRD) patterns were obtained with a Mini-
with only traces of HDO. In the absence of HDO, some phenol Flex II diffractometer using Ni-filtered CuKα radiation. The
is still formed by the first reaction step, hydrogenolysis of the X-ray tube was operated at 30 kV and 15 mA. Temperature pro-
C–O aryl bond B. In this case, products 4, 5 and 7 are formed grammed reduction (TPR) was carried out using hydrogen on
in higher amounts.
an instrument equipped with a thermal conductivity detector
To understand the stability of our catalyst, we performed a (TCD). The H2-TPR studies were carried out for all catalysts.
series of recycling and leaching tests with the unreduced cata- Samples of ca. 100 mg were loaded into a quartz U-tube
lyst. These were run with the unreduced samples for practical reactor and were pre-treated in N2 (40 ml min−1) at 473 K for
reasons, but we expect similar results with reduced ones. The 30 min. After cooling to ambient temperature, the gas stream
conversion remains constant after filtering the catalyst out, was switched to 5% H2/N2 flowing at 40 ml min−1. The
showing that the catalyst does not leach into solution. samples were heated at 10 K min−1 to 1000 K, during which
However, in the recycling experiments the conversion dropped the hydrogen consumption was determined quantitatively by
to 44% after the first cycle, indicating catalyst deactivation. TCD. Surface area measurements were performed by the BET
Twigg and Spencer39 studied the deactivation of supported method using N2 at 77 K on a Thermo Scientific Surfer instru-
copper metal catalysts for different hydrogenation ment. The samples were dried in a vacuum (1 × 10−3 mbar) for
reactions. They highlighted four main causes: (i) coke 3 h at 200 °C prior to the measurement.
formation, (ii) sintering of copper particle, (iii) change of
DFT calculations were performed with the ADF package,43
oxidation of copper and (iv) finally catalyst poisoning (with using the rPBE functional44 and a DZP basis set. Calculations
chlorine or sulfur compounds or adsorbed byproducts/ were done on isolated molecules. BDEs were calculated by
products on the catalyst). Under our reaction conditions, comparing the energies of the starting molecules and the iso-
poisoning of the catalyst with sulfur or chlorine is unlikely. lated radicals formed. For the protonated species, linear tran-
Because we have organic hydrogenation reactions, coking is sits were performed, slowly breaking the bonds instead of
more likely. The reaction temperatures are too low to involve comparing only the end energy. In this manner, the calcu-
sintering. Rao et al.40 studied the deactivation of several lations allow for the possibility of oligomerisation.
copper catalysts in the hydrogenation of aromatic ketones
Gas chromatography analyses were run on an Interscience
and aldehydes. They reported that the catalyst deactivation GC-8000 gas chromatograph with 14% cyanopropylphenyl and
occurs via different pathways: coke formation and/or poison- 86% dimethyl polysiloxane capillary column (Rtx-1701, 30 m;
ing of the catalyst (byproduct adsorption), or a change in 25 mm ID; 1 μm df). Samples were diluted in 1 ml MeOH. GC
the oxidation state of the copper during the reaction. Consider- conditions: isotherm at 50 °C (2 min); ramp at 2 °C min−1 to
ing the close chemical similarity with our reaction products, 70 °C; ramp at 70 °C min−1 to 140 °C; ramp at 10 °C min−1 to
similar deactivation processes may occur in the conversion 280 °C; isotherm at 260 °C (2 min). Products were identified by
of 1 and 2.
comparing their retention times to those of authentic samples.
This journal is © The Royal Society of Chemistry 2013
Green Chem.

View Article Online
Paper
Green Chemistry
Procedure for catalyst synthesis
The reactors were cooled down to room temperature using an
ice bath. Liquid samples were analysed by GC using chloroben-
zene as an external standard.
Example:10 wt% CuO/γ-Al2O3. 1.0 g of γ-Al2O3 was added to a
sol. of Cu(NO3)2·6H2O (0.38 g, 1.6 mmol) in 20 ml water and
stirred for 2 h at 25 °C. The liquid was evaporated overnight on
an oil bath at 60 °C, yielding a light green powder. This was
dried at 120 °C for 24 h and then calcined in air at 500 °C
(2 °C min−1) for 4 h. The analogous 10 wt% CuO/MgO–Al2O3
was prepared as above starting from 1.0 g of MgO–Al2O3. BET
analysis data of all catalysts are given in the ESI.†
Procedure for recycling and leaching tests
Recycling and leaching tests were carried out in a six-parallel
stainless steel 75 ml autoclave.
For the recycling test, 2 wt% of catalyst (copper weight rela-
tive to starting material using 10 wt% CuO/γ-Al2O3) was added
to a solution of 2-phenoxy-1-phenylethanone 1 (0.240 g,
1.12 mmol) in 20 ml of EtOAc. The autoclave was pressurized
Procedure for synthesising 2-phenoxy-1-phenylethanone 1
This is a modification of a previously published procedure.21,45 with 25 bar H2 and heated to 150 °C for 21 h. Then, the reac-
Bromoacetophenone (9.0 g, 45 mmol) and phenol (5.0 g, tors were cooled down to room temperature using an ice bath.
53 mmol) were dissolved in 200 ml DMF, mixed with KOH The catalyst was filtered out and placed in a desiccator over-
(3.0 g, 53 mmol) and stirred overnight at room temperature. night. The recycled catalyst was then reused for an extra 21 h
The product was then extracted with H2O and Et2O, dried over reaction using 2-phenoxy-1-phenylethanone
1
(0.120 g,
Na2SO4 and recrystallized from ethanol (yellowish powder, 0.56 mmol) in 10 ml of EtOAc with 2 wt% of catalyst (copper
86 mol% pure product yield based on bromoacetophenone). weight relative to starting material). Liquid samples were ana-
1H NMR (DMSO) δ 5.58 (s, 2H), 6.93–6.98 (m, 3H), 7.27–7.31 lysed by GC using chlorobenzene as an external standard.
(m, 2H), 7.56–7.60 (m, 2H), 7.68–7.72 (m, 1H), 8.02–8.04 (m,
2H), see ESI.†
For the leaching test, the autoclave was pressurized with 25
bar H2 and heated to 150 °C for 3 h. Then, the reactors were
cooled down to room temperature using an ice bath. The cata-
lyst was filtered out and the reaction mixture (without catalyst)
was charged again with 25 bar H2 and heated to 150 °C for
20 h. Liquid samples were analysed by GC using chloroben-
zene as an external standard.
Procedure for synthesising 2-phenoxy-1-phenylethanol 2
This is a modification of a previously published procedure.21,45
A solution of 2-phenoxy-1-phenylethanone 1 (2.5 g, 11 mmol)
in methanol (100 ml) was treated with small portions of
sodium borohydride (5.5 mmol) and stirred for 2 h. A satu-
rated solution of ammonium sulfate (200 ml) followed by
CHCl3 (200 ml) was added to the reaction mixture. The
organic layer was separated, washed with water (2 × 100 ml),
Acknowledgements
dried and recrystallized from ethanol (fine white needles, This research was performed within the framework of the
1
73 mol% pure compound yield based on 1). H NMR (DMSO) CatchBio program. The authors gratefully acknowledge the
δ 4.01–4.02 (m, 2H), 4.90–4.94 (m, 1H), 5.63–5.64 (d, 1H), support of the Smart Mix Program of the Netherlands, Ministry
6.90–6.94 (m, 3H), 7.25–7.47 (m, 7H), see ESI.†
of Economic Affairs, and the Netherlands Ministry of Edu-
cation, Culture and Science.
Procedure for catalytic hydrodeoxygenation (HDO)
Experiments were carried out in a six-parallel stainless steel
75 ml autoclave. In a typical experiment without an inert
atmosphere and without pre-reducing the catalysts (results in
Tables 2 and 3), 2 wt% of catalyst (copper weight relative to
starting material) was added to a solution of 2-phenoxy-1-phe-
nylethanone 1 (0.120 g, 0.56 mmol) in 10 ml of EtOAc. The
autoclave was pressurized with 25 bar H2 and heated to 150 °C
for 21 h. Then, the reactors were cooled down to room temp-
erature using an ice bath. Liquid samples were analysed by GC
using chlorobenzene as an external standard.
Notes and references
1 E. Dorrestijn, L. J. J. Laarhoven, I. W. C. E. Arends and
P. Mulder, J. Anal. Appl. Pyrolysis, 2000, 54, 153–192.
2 A. Séguin, Curr. Opin. Environ. Sustainability, 2011, 3,
90–94.
3 J. R. Regalbuto, Science, 2009, 325, 822–824.
4 Z. Ma, E. Troussard and J. A. van Bokhoven, Appl. Catal., B.,
2012, 423–424, 130–136.
In a second set of experiments (results shown in Tables 2
and 3), we used inert purging and the catalysts were pre-
reduced as follows: 1.0 g of catalyst was heated at 300 °C under
a 40 ml min−1 H2 for 2 h. The solvent, EtOAc, was purged for
2 h with nitrogen. 2 wt% of catalyst (copper weight related to
starting material) was added to an autoclave containing a solu-
tion of 2-phenoxy-1-phenylethanone 1 (0.120 g, 0.56 mmol) in
10 ml EtOAc. The autoclaves were flushed thrice with argon,
then pressurized to 25 bar H2 and heated to 150 °C for 21 h.
5 J. Zakzeski, P. C. A. Bruijnincx, A. L. Jongerius and
B. M. Weckhuysen, Chem. Rev., 2010, 110, 3552–3599.
6 H. Lawford and J. Rousseau, Appl. Biochem. Biotechnol.,
2002, 98–100, 429–448.
7 G. W. Huber, J. N. Chheda, C. J. Barrett and J. A. Dumesic,
Science, 2005, 308, 1446–1450.
8 H. Kobayashi, Y. Ito, T. Komanoya, Y. Hosaka, P. L. Dhepe,
K. Kasai, K. Hara and A. Fukuoka, Green Chem., 2011, 13,
326–333.
Green Chem.
This journal is © The Royal Society of Chemistry 2013

View Article Online
Green Chemistry
Paper
9 R. J. A. Gosselink, E. de Jong, B. Guran and A. Abacherli, 27 S. Sitthisa and D. Resasco, Catal. Lett., 2011, 141, 784–791.
Ind. Crops Prod., 2004, 20, 121–129.
10 M. Kleinert and T. Barth, Chem. Eng. Technol., 2008, 31,
736–745.
28 S. Sitthisa, T. Sooknoi, Y. Ma, P. B. Balbuena and
D. E. Resasco, J. Catal., 2011, 277, 1–13.
29 Z. Strassberger, S. Tanase and G. Rothenberg, Eur. J. Org.
Chem., 2011, 5246–5249.
11 J. Gierer, Wood Sci. Technol., 1980, 14, 241–266.
12 R. Vanholme, B. Demedts, K. Morreel, J. Ralph and 30 P. W. Park and J. S. Ledford, Appl. Catal., B, 1998, 15,
W. Boerjan, Plant Physiol., 2010, 153, 895–905.
13 E. Adler, Wood Sci. Technol., 1977, 11, 169–218.
14 F. S. Chakar and A. J. Ragauskas, Ind. Crops Prod., 2004, 20,
131–141.
221–231.
31 Z. Wang, H. Wan, B. Liu, X. Zhao, X. Li, H. Zhu, X. Xu, F. Ji,
K. Sun, L. Dong and Y. Chen, J. Colloid Interface Sci., 2008,
320, 520–526.
15 S. Jia, B. J. Cox, X. Guo, Z. C. Zhang and J. G. Ekerdt, Ind. 32 J.-T. Feng, X.-Y. Ma, D. G. Evans and D.-Q. Li, Ind. Eng.
Eng. Chem. Res., 2010, 50, 849–855. Chem. Res., 2011, 50, 1947–1954.
16 A. L. Jongerius, R. Jastrzebski, P. C. A. Bruijnincx and 33 W.-P. Dow, Y.-P. Wang and T.-J. Huang, Appl. Catal., A,
B. M. Weckhuysen, J. Catal., 2012, 285, 315–323. 2000, 190, 25–34.
17 H. Kawamoto, S. Horigoshi and S. Saka, J. Wood Sci., 2007, 34 K. M. Lee and W. Y. Lee, Catal. Lett., 2002, 83, 65–70.
53, 168–174.
35 S. Kim, S. C. Chmely, M. R. Nimlos, Y. J. Bomble,
T. D. Foust, R. S. Paton and G. T. Beckham, J. Phys. Chem.
Lett., 2011, 2, 2846–2852.
36 A. Beste and A. C. Buchanan, J. Org. Chem., 2011, 76,
2195–2203.
37 J. M. Nichols, L. M. Bishop, R. G. Bergman and
J. A. Ellman, J. Am. Chem. Soc., 2010, 132, 12554–12555.
38 J. R. Pound, J. Phys. Chem., 1930, 35, 1496–1497.
18 R. J. A. Gosselink, W. Teunissen, J. E. G. van Dam, E. de
Jong, G. Gellerstedt, E. L. Scott and J. P. M. Sanders, Biore-
sour. Technol., 2012, 106, 173–177.
19 K. Takeno, T. Yokoyama and Y. Matsumoto, BioResources,
2012, 7, 99–111.
20 T. vom Stein, T. Weigand, C. Merkens, J. Klankermayer and
W. Leitner, ChemCatChem, 2013, 5, 439–441.
21 P. F. Britt, A. C. Buchanan, M. J. Cooney and 39 M. V. Twigg and M. S. Spencer, Appl. Catal., A, 2001, 212,
D. R. Martineau, J. Org. Chem., 2000, 65, 1376–1389.
22 J. E. Tilton and G. Lagos, Resour. Policy, 2007, 32, 19–23.
23 T. T. Pham, L. L. Lobban, D. E. Resasco and
R. G. Mallinson, J. Catal., 2009, 266, 9–14.
161–174.
40 R. Rao, R. T. Baker and M. A. Vannice, Catal. Lett., 1999,
60, 51–57.
41 P. Gallezot, ChemSusChem, 2008, 1, 734–737.
24 M. V. Bykova, D. Y. Ermakov, V. V. Kaichev, 42 http://www.eurosupport.nl/
O. A. Bulavchenko, A. A. Saraev, M. Y. Lebedev and 43 G. te Velde, F. M. Bickelhaupt, E. J. Baerends, C. Fonseca
V. A. Yakovlev, Appl. Catal., B, 2012, 113–114, 296–307.
Guerra, S. J. A. van Gisbergen, J. G. Snijders and T. Ziegler,
25 V. Dundich, S. Khromova, D. Ermakov, M. Lebedev,
J. Comput. Chem., 2001, 22, 931–967.
V. Novopashina, V. Sister, A. Yakimchuk and V. Yakovlev, 44 B. Hammer, L. B. Hansen and J. K. Nørskov, Phys. Rev. B:
Kinet. Catal., 2010, 51, 704–709. Condens. Matter, 1999, 59, 7413–7421.
26 F. Zaccheria, N. Ravasio, M. Ercoli and P. Allegrini, Tetra- 45 P. H. Kandanarachchi, T. Autrey and J. A. Franz, J. Org.
hedron Lett., 2005, 46, 7743–7745.
Chem., 2002, 67, 7937–7945.
This journal is © The Royal Society of Chemistry 2013
Green Chem.