Selective conversion of 5-hydroxymethylfurfural to cyclopentanone derivatives over Cu-Al2O3 and Co-Al2O3 catalysts in water

Article abstract of DOI:10.1039/c7gc00315c

The production of cyclopentanone derivatives from 5-hydroxymethylfurfural (HMF) using non-noble metal based catalysts is reported for the first time. Five different mixed oxides containing Ni, Cu, Co, Zn and Mg phases on an Al-rich amorphous support were prepared and characterised (XRD, ICP, SEM, TEM, H₂-TPR, NH₃/CO₂-TPD and N₂ sorption). The synthesised materials resulted in well-dispersed high metal loadings in a mesoporous network, exhibiting acid/base properties. The catalytic performance was tested in a batch stirred reactor under H₂ pressure (20-50 bar) in the range T = 140-180 °C. The Cu-Al₂O₃ and the Co-Al₂O₃ catalysts showed a highly selective production of 3-hydroxymethylcyclopentanone (HCPN, 86%) and 3-hydroxymethylcyclopentanol (HCPL, 94%), respectively. A plausible reaction mechanism is proposed, clarifying the role of the reduced metal phases and the acid/basic sites on the main conversion pathways. Both Cu-Al₂O₃ and Co-Al₂O₃ catalysts showed a loss of activity after the first run, which can be reversed by a regeneration treatment. The results establish an efficient catalytic route for the production of the diol HCPL (reported for the first time) and the ketone HCPN from bio-derived HMF over 3d transition metals based catalysts in an environmental friendly medium such as water.

Full text of DOI:10.1039/c7gc00315c

View Article Online
View Journal
Green
Chemistry
Accepted Manuscript
This article can be cited before page numbers have been issued, to do this please use: R. Ramos Velarde,
A. Grigoropoulos, N. Perret, M. Zanella, A. P. Katsoulidis, T. D. Manning, J. B. Claridge and M. J.
Rosseinsky, Green Chem., 2017, DOI: 10.1039/C7GC00315C.
This is an Accepted Manuscript, which has been through the
Volume 18 Number
7 7 April 2016 Pages 1821–2242
Royal Society of Chemistry peer review process and has been
accepted for publication.
Green
Chemistry
Accepted Manuscripts are published online shortly after
Cutting-edge research for a greener sustainable future
www.rsc.org/greenchem
acceptance, before technical editing, formatting and proof reading.
Using this free service, authors can make their results available
to the community, in citable form, before we publish the edited
article. We will replace this Accepted Manuscript with the edited
and formatted Advance Article as soon as it is available.
You can find more information about Accepted Manuscripts in the
author guidelines.
Please note that technical editing may introduce minor changes
to the text and/or graphics, which may alter content. The journal’s
standard Terms & Conditions and the ethical guidelines, outlined
in our author and reviewer resource centre, still apply. In no
event shall the Royal Society of Chemistry be held responsible
for any errors or omissions in this Accepted Manuscript or any
consequences arising from the use of any information it contains.
ISSN 1463-9262
CRITICAL REVIEW
G. Chatel et al.
Heterogeneous catalytic oxidation for lignin valorization into valuable
chemicals: what results? What limitations? What trends?
rsc.li/green-chem

Please do not adjust margins
Page 1 of 13
Green Chemistry
View Article Online
DOI: 10.1039/C7GC00315C
Green Chemistry
ARTICLE
Selective conversion of 5-hydroxymethylfurfural to
cyclopentanone derivatives over Cu-Al2O3 and Co-Al2O3 catalysts
in water
Received 00th January 20xx,
Accepted 00th January 20xx
Rubén Ramos,a Alexios Grigoropoulos,a Noémie Perret,a Marco Zanella,a Alexandros P.
Katsoulidis,a Troy D. Manning,a John B. Claridgea and Matthew J. Rosseinsky*,a
DOI: 10.1039/x0xx00000x
www.rsc.org/
The production of cyclopentanone derivatives from 5-hydroxymethylfurfural (HMF) using non-noble metal based catalysts
is reported for the first time. Five different mixed oxides containing Ni, Cu, Co, Zn and Mg phases on an Al-rich amorphous
support were prepared and characterised (XRD, ICP, SEM, TEM, H2-TPR, NH3/CO2-TPD and N2 sorption). The synthesised
materials resulted in well-dispersed high metal loadings in a mesoporous network, exhibiting acid/base properties. The
catalytic performance was tested in a batch stirred reactor under H2 pressure (20 - 50 bar) in the range T = 140 - 180 °C.
The Cu-Al2O3 and the Co-Al2O3 catalysts showed a highly selective production of 3-hydroxymethylcyclopentanone (HCPN,
86 %) and 3-hydroxymethylcyclopentanol (HCPL, 94 %), respectively. A plausible reaction mechanism is proposed,
clarifying the role of the reduced metal phases and the acid/basic sites on the main conversion pathways. Both Cu-Al2O3
and Co-Al2O3 catalysts showed a loss of activity after the first run, which can be reversed by a regeneration treatment. The
results establish an efficient catalytic route for the production of the diol HCPL (reported for the first time) and the ketone
HCPN from bio-derived HMF over 3d transition metals based catalysts in an environmental friendly medium such as water.
Introduction
Emerging environmental concerns and declining availability of into cyclopentanone derivatives in water. These alicyclic
petroleum reserves have promoted worldwide interest in compounds are versatile raw materials for the synthesis of products
developing alternative technologies to convert sustainable biomass like fragrances, drugs, solvents, pesticides, polymers, etc.13,14
resources into value-added chemicals. Non-edible lignocellulosic Therefore, since they are currently mainly produced from
biomass can be efficiently transformed via acid-catalysed hydrolysis petrochemical resources (decarboxylation of adipic acid,15 oxidation
to C5 and C6 monosaccharides,1,2 which may be further processed of cyclopentene16 or cyclization of 1,6-hexanediol17) under harsh
(via acid-catalysed dehydration) to produce furanic compounds reaction conditions, the efficient manufacture of these commodity
such as furfural (C5) and 5-hydroxymethylfurfural (C6).3,4 Both furan chemicals from biomass-derived compounds would be of great
derivatives are regarded as promising biobased platform molecules, significance. Moreover, since production of HMF from sugars is
expected to play a critical role in future biorefineries.5,6,7
In this context, 5-hydroxymethylfurfural (HMF) can be processing of HMF in water is especially convenient.18,19
converted into useful bio-derived products that could serve as Although furfural conversion to cyclopentanone and
carried out in water, a renewable, safe and low cost solvent, further
building blocks in diverse sectors of the chemical industry, e.g.; 2,5- cyclopentanol has been previously described in aqueous medium
furandicarboxylic acid (FDCA),8 tetrahydrofuran 2,5-diyldimethanol (140 - 190 °C; 20 - 80 H2 bar),14,20 the analogous transformation for
(THFDM),9 2,5-dimethylfuran (DMF),10 levulinic acid11 and C6 linear HMF has been scarcely studied. Ohyama et al. recently reported the
alcohols12 (Figure 1). Due to the overfunctionalized nature of HMF, conversion of HMF to 3-hydroxymethylcyclopentanone (HCPN; 86
the formation of these compounds involves several chemical %) in water and under hydrogen pressure (140 °C; 30 - 80 H2 bar),
transformations (e.g. oxidation, hydrogenation, dehydration, using supported Au nanoparticles13 and Pt/SiO2 in combination with
hydrogenolysis, ring opening, etc.) which require the presence of acid-base metal oxides.21 By contrast, we reported the formation of
multifunctional heterogeneous catalysts.
One particularly challenging reaction is the conversion of HMF opposed to the use of noble metals catalysts.22
The conversion of HMF into alicyclic compounds proceeds
HCPN from HMF in water by using a 3d metal Ni-Al catalyst, as
through initial hydrogenation steps on metal particles, followed by
furan ring rearrangement catalysed by Lewis acidic sites. However,
This journal is © The Royal Society of Chemistry 2017
Green Chem., 2017, 00, 1-3 | 1
Please do not adjust margins

Please do not adjust margins
Green Chemistry
Page 2 of 13
View Article Online
DOI: 10.1039/C7GC00315C
ARTICLE
Journal Name
cyclopentanone derivatives from lignocellulose-derived HMF in
water.
Experimental
Catalyst preparation
Catalyst precursors were prepared by co-precipitation using the
urea-based method developed by Constantino et al.25 Firstly, a
calculated mass of urea (Sigma) was weighed into glass vials (27 ml)
and combined with distilled water using a Chemspeed Accelerator
SLT synthesis platform. Then, a mixture of 1.5 M aqueous solutions
of AlCl3·6H2O (Fluka;
≥ 99%) and the corresponding metal
precursors were dispensed into those vials using an Eppendorf
epMotion 5075PC pipetting robot. The precursors used as Ni, Cu,
Co, Mg, and Zn sources were: NiCl2·6H2O (Aldrich; ≥ 99.9%),
Cu(NO3)2·3H2O (Fluka), CoCl2·6H2O (Fluka), MgCl2·6H2O (Sigma-
Aldrich; ≥ 99%) and ZnCl2 (Aldrich; ≥ 99.99%), respectively. The
volumes of the solutions and the mass of urea were selected to set
the following molar ratios; nAl/(nMe+nAl) = 0.45 and nurea/(nMe+nAl) =
3.3. The vials were heated to 95 °C (1.5 °C min-1) and stirred under
reflux. After aging for 65 h, the solutions were cooled to room
temperature and filtered. In order to remove residual Cl, the
precipitated materials were left in suspension with NH4HCO3 (Alfa
Caesar) for 5 h. Subsequently, the samples were filtered, washed
with distilled water and dried at 2 °C min-1 to 120 °C for 5 h in static
air. Finally, the synthesised materials were calcined under air (75
cm3 min-1) at 500 °C for 5 h (heating ramp 5 °C min-1) and reduced
under H2 (100 cm3 min-1) at 500 °C (700 °C in case of Co containing
sample) for 5 h (heating ramp 5 °C min-1). The catalysts were then
passivated at room temperature for 3 h under a flow of 1% v/v
O2/N2 (100 cm3 min-1).
Figure 1. Catalytic transformation of HMF to bio-derived products
with potential industrial applications (HCPN and HCPL are the
targeted products of the present work).
some fundamental aspects of the reaction mechanism like the role
of metal sites on ring-opening/closure steps or the influence of
basic sites in condensation reactions remain unclear. Moreover, the
formation of the fully hydrogenated 3-hydroxymethylcyclopentanol
(HCPL) from HMF has not been reported to date. Thus, the
development of novel heterogeneous catalysts based on
inexpensive and abundant 3d transition metals, exhibiting water
stability and high activity/selectivity towards HCPN and/or HCPL
remains a challenging task.
In this regard, the synthesis of mixed oxides derived from
layered double hydroxides (LDH) has attracted considerable
attention as multifunctional catalyst precursors, due to their ability
to accommodate a large variety of redox-active divalent and
trivalent cations in a Brønsted base framework. The subsequent
calcination and reduction steps afford high metal loadings
homogenously dispersed in a mesoporous network, with good
thermal stability.23 Likewise, the synergetic effects between metal
and oxide phases may generate acid-base active sites, which are
expected to contribute to the efficient rearrangement of HMF
intermediates into alicyclic compounds. In this respect, the use of
different M2+ cations may allow tailoring of the catalytic properties
and enhancement of the selectivity toward the targeted products.24
In this work, five different mixed oxides containing Ni, Cu, Co,
Zn and Mg, on alumina amorphous support were prepared. The
synthesised materials were widely characterised and investigated as
catalysts for the one-pot conversion of HMF to cyclopentanone
derivatives in water. High selectivity towards either HCPN or HCPL
was observed depending on the catalyst and the operation
conditions applied. The obtained product distribution was
correlated with the hydrogenating ability and acid-base properties
of the catalysts. The effects of reaction temperature (140 - 180 °C)
and H2 pressure (20 - 50 bar) on the performance of the most
promising catalysts were also examined in terms of HCPN and HCPL
production. Equally, catalyst stability and reusability were evaluated
by catalytic and characterization studies of the used catalysts. This
approach may establish a consistent understanding of the design of
heterogeneous catalyst for the competitive production of
Catalyst characterization
The synthesised catalysts were characterised by X-Ray
diffraction on a Panalytical X’Pert Pro diffractometer with Co Kα1
radiation (λ = 1.7890 Å). Samples were scanned at 0.023° s-1 over
the range 10° < 2θ < 80° for phase identification using reference
standards. The mean metal crystallite size was calculated using the
Scherrer equation. Metal content of the catalysts was measured by
inductively coupled plasma optical emission spectroscopy (ICP-OES)
after digestion in acid (25 mg of sample in 10 ml HCl 37 %) and
dilution with water (1:10 v/v). Textural properties were measured
using nitrogen adsorption–desorption isotherms at 77 K with a
Micromeritics TRISTAR II instrument. Prior to the measurement, the
samples were outgassed at 120 °C under vacuum for 20 h. Specific
surface area was calculated using the BET equation in the range
0.05 < P/P0 < 0.2, whereas pore volume and pore size distribution
were determined by the BJH method.
Temperature programmed reduction (TPR) was measured using
a Quantachrome ChemBET 3000 unit; ca. 50 mg of sample were
loaded into a quartz cell and heated up to 750 °C at 5 °C min-1 under
a flow of 5% v/v H2/N2 (100 cm3 min-1). The acid and base features
of the catalysts were determined in a Micrometrics 2920 equipment
by temperature programmed desorption (TPD) of ammonia and
carbon dioxide, respectively. Previously, the samples were
outgassed under a stream of He (20 cm3 min−1) heating at 5 °C min−1
2 | Green Chem., 2017, 00, 1-3
This journal is © The Royal Society of Chemistry 2017
Please do not adjust margins

Please do not adjust margins
Green Chemistry
Page 3 of 13
Journal Name
View Article Online
DOI: 10.1039/C7GC00315C
ARTICLE
up to 550 °C. For NH3-TPD curves, the samples were cooled to 180 Results and discussion
°C and saturated under an ammonia stream (5% v/v NH3/He; 20 cm3
Characterization analysis
min−1) for 30 min. Subsequently, the physically adsorbed ammonia
was removed by flowing helium (20 cm3 min−1) for 30 min.
The synthesised hydrotalcite precursors are described by the
2-
Thereafter, the chemically adsorbed ammonia was desorbed by general formula [M2+1-xAl3+x(OH)2][(CO3 x/2]·mH2O, where M2+ are
)
heating to 550 °C with a rate of 10 °C min−1 in flowing He (20 cm3 the divalent cations: Ni2+, Cu2+, Co2+, Zn2+, and Mg2+. The structure
min−1). The CO2-TPD experiment was initiated by cooling the sample of these layered double hydroxides corresponds to brucite-like
to 50 °C followed by saturation with a flow of 5% v/v CO2/He (20 sheets, replacing a fraction of M2+ by Al3+ and conferring a positive
2-
cm3 min-1) for 30 min. Afterward, the physisorbed carbon dioxide layer charge. This charge is balanced by interlayer CO3 anions,
was removed by flowing He (20 cm3 min-1) for 30 min. Finally, the whereas intercalated H2O molecules provide hydrogen bonding
chemically adsorbed CO2 was desorbed by heating the sample to between adjacent brucite layers. Five different layered double
650 °C (heating rate of 10 °C min-1) in flowing He (20 cm3 min-1). The hydroxides (labelled M-Al-P) were prepared with the same M2+/Al3+
variation of ammonia and carbon dioxide concentration in the ratio (xAl = 0.45), as described in the experimental section.
effluent helium stream was recorded continuously using a thermal
conductivity detector (TCD).
The XRD patterns (2θ = 10 - 80 °) of the precursors prepared
with Ni, Co, Zn and Mg show the characteristic peaks of
IR spectra were collected in ATR mode using a FT-IR Perkin- hydrotalcite-like materials (Figure 2a).26 Well-defined diffraction
Elmer Spectrum 100 spectrometer (4 cm-1 resolution, 16 scans). peaks of the three basal (003), (006) and (009) planes, and the
Thermogravimetric analysis (TGA) were carried out on a Q600 TA reflections associated with the non-basal (110) and (113) planes can
Instruments; ca. 10 mg of sample were loaded into an alumina be identified. The latter peaks reveal good dispersion of the metal
microcrucible and heated to 800 °C at 10 °C min-1 in flowing N2 at cations in the hydroxide layers.27 These results are in agreement
100 cm3 min-1. Elemental analysis (C and H content) of the used with the work of Constantino et al.,25 reporting a well-crystallized
catalysts were carried out on a Thermo EA1112 Flash CHNS layered structure
Analyser. SEM images were taken using a Hitachi S-4800 Field-
Emission Scanning Electron Microscope. Prior to recording the SEM
(a)
Hydrotalcite
Malachite
*
o
micrographs, the samples were sputter-coated with Au to reduce
charging effects. TEM images were obtained with a JEOL 2100
transmission electron microscope. EDX mapping was performed
with a windowless EDAX EDX detector using the microscope in
STEM mode. Previous to the observation, the samples were
dispersed in acetone, stirred in an ultrasonic bath and deposited on
a carbon-coated Cu grid.
*
*
*
*
*
*
Mg-Al-P
*
*
*
*
*
*
*
*
*
*
Zn-Al-P
Co-Al-P
Cu-Al-P
*
*
*
*
*
Catalytic experiments and product analysis
o
o
The performance of the catalysts was studied in a high pressure
100 ml batch stirred reactor (Parr Instrument Co.) equipped with a
Parr 4848 acquisition interface. A glass liner was loaded with 45 ml
of an aqueous solution of HMF (0.04 M) and 0.02 - 0.06 g of
catalyst, and placed into the stainless steel reactor. The reactor was
then sealed, flushed with N2 and heated to the required reaction
temperature (140 - 180 °C). Once the desired temperature was
reached, the vessel was pressurized to 20 - 50 bar of H2 and the
stirring speed was increased to 600 rpm. After the reaction, the
product identification and quantification was carried out by GC-MS
(Agilent 6890N GC with a 5973 MSD detector) and GC (Agilent
7890A GC with a FID detector), respectively. Both instruments were
equipped with a DB-WAXetr capillary column (60 m, 0.25 mm i.d.,
0.25 μm). Standard reference compounds like HMF (Sigma; > 99 %),
furan-2,5-diyldimethanol (FDM; Manchester Organics), THFDM
(Ambinter) and HCPL (Ambinter) were used for identification and
calibration measurements. Identification of 1-hydroxyhexane-2,5-
o
o
o
*
*
*
Ni-Al-P
60
*
*
* *
10
20
30
40
50
70
80
2
θ
(b)
#
MgO
#
Mg-Al-C
#
x
ZnAl O
2
4
x
x
x
Zn-Al-C
Co-Al-C
Cu-Al-C
x
x
o
x
o
Co O
o
3
4
o
o
o
o
CuO
^
^
^
^
^
^
^
NiO
*
dione
(HHD),
4-hydroxy-4-hydroxymethyl-2-cyclopentenone
*
*
Ni-Al-C
*
(HHCPN) and HCPN was carried out via a combination of GC-MS and
NMR spectroscopy using a Bruker AVANCE III HD spectrometer.
Details concerning conversion, yield and selectivity calculations are
given elsewhere.22
20
30
40
50
60
70
80
2
θ
Figure 2. (a) XRD patterns of the as-synthesised precursor materials
and (b) XRD patterns of the calcined materials (air, 500 °C, 5 h).
This journal is © The Royal Society of Chemistry 2017
Green Chem., 2017, 00, 1-3 | 3
Please do not adjust margins

Please do not adjust margins
Green Chemistry
Page 4 of 13
View Article Online
DOI: 10.1039/C7GC00315C
ARTICLE
Journal Name
with the employed synthesis parameters (pH ≈ 8.5; 65 h; 95 °C).
However, one exception should be noted in the case of the Cu-Al-P
diffraction pattern, in which an essentially amorphous material is
observed with traces of malachite phase (Cu2(OH)2(CO3)). Several
authors have claimed the difficulty in obtaining pure Cu-Al
hydrotalcite type compounds due to the Jahn-Teller effect of Cu2+
ions (d9 electron configuration) which destabilises octahedral
geometry and favours other phases like malachite.28,29,30
(a)
Mg-Al-C
Zn-Al-C
625 °C
435 °C
Co-Al-C
Cu-Al-C
The as-synthesised precursor materials were calcined at 500 °C
356 °C
5 hours in air (heating rate 5
°C min-1) to obtain the
288 °C
for
corresponding mixed oxides. During calcination, H2O molecules and
carbonate anions are initially removed, followed by dehydroxylation
of the surface and partial oxidation of M2+ at higher temperatures.
Under these conditions, the double-layered structure collapses and
the metal ions coordinate with adjacent oxygen atoms and work as
577 °C
Ni-Al-C
100
200
300
400
500
600
700
Temperature (°C)
pillars forming an open mesoporous structure. As
a result,
calcination leads to metal-like oxides with an Al-rich amorphous
component (labelled M-Al-C). Figure 2b shows the XRD patterns of
the calcined materials. The samples prepared with Ni or Cu show
peaks characteristic of cubic NiO (JCPDS 70-0989) and monoclinic
CuO (65-2909), respectively. For Co-Al-C and Zn-Al-C the initial
(b)
#
MgO
#
Mg-Al O
#
2
3
x
x
ZnAl O
2 4
x
x
Zn-Al O
x
x
2
3
3
x
o
layered structures were converted to
a spinel-like phase of
composition Co3O4 (80-1535) and ZnAl2O4 (73-1961). It should be
noted that for Co-Al-C, the observed diffraction peaks may be also
ascribed to CoAl2O4, since it is not possible to distinguish between
these two phases by PXRD.31 For Mg-Al-C, the calcination treatment
resulted in the formation of MgO periclase (03-0998).
o
Co
Co-Al O
2
o
o
o
^
Cu
Ni
^
*
^
Cu-Al O
2
3
*
The reducibility of the calcined samples was determined by
hydrogen temperature-programmed reduction (Figure 3a). The H2-
TPR profiles of Zn-Al-C and Mg-Al-C do not show any peak
attributed to H2 consumption, suggesting that the oxide phases
formed in these samples are not reduced under the analysis
conditions (5% v/v H2/N2; 20 - 750 °C). For Ni-Al-C, the TPR curve
exhibits a single broad peak centered on 577 °C, which has been
previously ascribed to the reduction of Ni2+ to Ni0. In this respect, it
is well known that the introduction of Al3+ into the NiO lattice
enhances its stability and increases the reduction temperature.32
The reduction profile of Cu-Al-C presents two overlapping peaks at
288 and 356 °C, assigned to the reduction of highly and poorly
dispersed Cu oxide species on alumina, respectively.33,34 The
reduction profile of Co-Al-C shows two H2 consumption peaks in
two different temperature intervals. The low-temperature peak
(435 °C) is attributed to partial reduction of Co3O4 to CoO, whereas
the high-temperature peak (625 °C) is typically assigned to the
reduction of CoO to metallic cobalt.35,36 However, it should be
noted that both the temperature interval and the area of the H2-
TPR peaks can be also influenced by the interaction with the
alumina support and the metal crystallite sizes.37
*
Ni-Al O
2
3
20
30
40
50
60
70
80
2
θ
Figure 3. (a) H2-TPR curves (5 % v/v H2/N2) of the calcined
precursors and (b) XRD patterns of the reduced catalysts (H2,
500/700 °C, 5 h). Only Ni, Cu and Co are reduced to the respective
metallic phase under these conditions.
since only Co3O4 can be reduced below 800 °C.38 By contrast, Zn-
Al2O3 and Mg-Al2O3 catalysts show identical XRD pattern to the
calcined samples. This corroborates the H2-TPR results and confirms
that the reduction treatment does not affect these materials, as
expected from their redox chemistry.
The composition of the prepared catalysts was determined by
ICP and TGA analysis (Table 1). The Al molar fraction for the
samples containing Ni, Co and Zn lies in the range xAl = 0.42 - 0.52,
similar to the one introduced in the synthesis of the hydrotalcite
precursors (xAl = 0.45). On the other hand, the Al molar fraction of
the Mg-Al2O3 sample is higher (xAl = 0.60), which could be related to
the higher solubility of Mg(OH)2 than aluminium hydroxide in the
synthesis solution (pH ≈ 8.5).25 A higher final Al molar ratio is also
found in Cu-Al2O3 (xAl = 0.76) which is not unexpected since a
different precursor material was used (vide supra) and is indicative
of incomplete precipitation of Cu during the synthesis of that
precursor.
The calcined samples were reduced under a stream of H2 (100
cm3 min-1; 5 h) at a temperature determined by the respective H2-
TPR profile and then passivated (1% v/v O2/N2; 100 cm3 min-1; 20 °C
for 3 h). Figure 3b shows the XRD patterns of the final catalysts
after the reduction treatment (M-Al2O3). The samples containing Ni,
Cu and Co show peaks corresponding to cubic Ni (70-0989), Cu (03-
1005) and Co (01-1259) metallic phases. The disappearance of
cobalt spinel-like phase after the reduction step indicates the
prevalence of Co3O4 over CoAl2O4 in the calcined precursor Co-Al-C,
The prepared catalysts exhibit type IV N2 adsorption-desorption
isotherms with significant adsorption of N2 at intermediate relative
pressure (0.2 < P/P0 < 0.8) due to the filling of the mesopores and
4 | Green Chem., 2017, 00, 1-3
This journal is © The Royal Society of Chemistry 2017
Please do not adjust margins

Please do not adjust margins
Green Chemistry
Page 5 of 13
Journal Name
View Article Online
DOI: 10.1039/C7GC00315C
ARTICLE
Table 1. Physicochemical properties of the synthesised catalysts.
Textural properties(d)
Acid properties(e)
Base properties(f)
(c)
M(b)
dM
Composition(a)
Catalyst
SBET (m2/g) Vp (cm3/g) Dp (nm)
T(g) (°C) μmol NH3/g Strength μmol CO2/g
(wt%)
(nm)
Low
Medium
High
Low
Medium
High
Low
Medium
High
Low
Medium
High
31
41
115
130
116
-
32
35
196
45
60
Ni-Al2O3
Ni0.58Al0.42O1.40
50/46
26/24
44/45
29
9
52
28
5
98
75
96
95
77
0.236
0.180
0.198
0.170
0.127
10
10
9
276
269
276
260
266
252
99
Cu-Al2O3
Co-Al2O3
Zn-Al2O3
Cu0.24Al0.76O1.24
Co0.56Al0.44O1.48
Zn0.48Al0.52O1.59
236
109
7
93
25
26
46
Low
Medium
High
Mg-Al2O3 Mg0.40Al0.60O1.26
20
6
6
113
(a)
(b)
(c)
Based on ICP and TGA analysis. Metal content calculated from ICP/SEM-EDX measurements. Mean metal crystallite size
(d)
based on XRD (Scherrer equation).
BET surface area (SBET), pore volume (Vp) and pore diameter (Dp) calculated from N2
adsorption isotherms at 77 K. (e) Acid properties from ammonia TPD measurements. (f) Base properties from carbon dioxide TPD
measurements. (g) Maximum desorption temperature.
an increase in N2 uptake at P/P0 > 0.8 due to voids of the basic sites (130 and 116 μmol CO2 g-1, respectively) and most
interparticulate space (Figure SI1). The calculated textural importantly the complete absence of strong basic sites.
properties (Table 1) show that the incorporation of different M2+
Representative TEM images of the prepared catalysts are
presented in Figure SI2, suggesting that a good metal dispersion
cations does not significantly change the specific surface area (SBET
=
75 - 98 m2 g-1) and pore volume (Vp = 0.127 - 0.236 cm3 g-1) of the degree and homogeneity was achieved for all samples. The
samples. The calculated pore distributions, according to the BJH observed particle sizes are consistent with those calculated from
method, show average pore-diameter values in the mesoporous XRD patterns (dM, Table 1). Thus, the Cu-Al2O3 and Co-Al2O3
range (Dp = 6 - 10 nm), large enough to host the reactant and the materials clearly present a higher average particle diameter (52 and
intermediate molecules (HMF long axis ≈ 0.9 nm).39
28 nm, respectively) than the Ni-Al2O3 catalyst (9 nm). Likewise,
Figure shows the NH3 (4a) and CO2 (4b) temperature Figure SI2d reveals that the Zn-Al2O3 catalyst is mainly composed of
4
programmed desorption (TPD) of the synthesised catalysts after nanoparticles (5 nm) with spherical shape, similar to the
calcination and reduction treatments. The overall measured morphology reported for ZnAl2O4 phase.42
amount and strength of acid/basic sites are summarized in Table 1.
The NH3-TPD curves (Figure 4a) show a noticeably higher acid
(a)
(b)
character for Ni-Al2O3 and Co-Al2O3 catalysts (252 and 236 μmol
NH3 g-1, respectively). On the other hand, the Cu-Al2O3 sample
adsorbs the lowest amount of ammonia (99 μmol NH3 g-1), which
implies a lower number of active acid sites present. Regarding the
acid strength, no significant differences were observed in the
ammonia TPD curves, with the maximum NH3 desorption
temperature occurring within a narrow interval 260 - 276 °C.
Mg-Al O
Mg-Al O
2
3
2
3
Zn-Al O
Zn-Al O
2
3
2
3
Co-Al O
2
3
Co-Al O
2
3
The total amount of each type of basic sites was determined by
integration of the deconvoluted peaks from the TPD curves (Table
1). The CO2 TPD profiles of Ni-Al2O3, Co-Al2O3, Zn-Al2O3 and Mg-
Al2O3 (Figure 4b) are in accordance with the well-known basic
character of hydrotalcite-derived material, showing three different
peaks in the interval 95 - 400 °C associated with the adsorption of
bicarbonate anions (low-strength), bidentate carbonates (medium-
strength) and monodentate species (high-strength).40,41 For
hydrotalcite derived materials, the highest desorption contribution
corresponds to monodentate species (46 - 196 μmol CO2 g-1),
indicating the prevalence of high-strength basic sites on these
materials. On the contrary, the CO2-TPD profile of the Cu-Al2O3
catalyst reveals a considerably larger amount of weak and medium
Low-strength
basic sites
Cu-Al O
2
3
Medium-strength
basic sites
Cu-Al O
2
3
High-strength
basic sites
Ni-Al O
2
3
Ni-Al O
2
3
50 100 150 200 250 300 350 400 450 500 550 600 650
Temperature (°C)
200 250 300 350 400 450 500 550
Temperature (°C)
Figure 4. (a) NH3 (b) and CO2 temperature programmed desorption
(TPD) curves of the prepared catalysts after calcination (air, 500 °C,
5 h) and reduction (H2, 500/700 °C, 5 h) steps. Dash lines represent
the deconvoluted peaks associated with different basic strengths.
This journal is © The Royal Society of Chemistry 2017
Green Chem., 2017, 00, 1-3 | 5
Please do not adjust margins

Please do not adjust margins
Green Chemistry
Page 6 of 13
View Article Online
DOI: 10.1039/C7GC00315C
ARTICLE
Journal Name
5-Hydroxymethylfurfural conversion over mixed oxide catalysts in inactive after 6 h (HMF conversion = 1 %).
water
Within the group of catalytically active materials (i.e. Ni-Al2O3,
Cu-Al2O3 and Co-Al2O3) significant differences are observed in the
final product distribution under identical conditions. The highest
yield for HCPN is observed for the Cu-Al2O3 catalyst (79 %), followed
by the Ni-Al2O3 (59 %). Although both catalysts show similar low
yields for FDM (4 vs. 5 %), the employment of the Ni-Al2O3 catalyst
results in a significant amount of THFDM (19 %), via the complete
hydrogenation of the furan ring. By contrast, this step is totally
impeded in the case of the Cu-Al2O3 catalyst. Moreover, two more
compounds are detected at relatively lower yields, namely HHD (7
%) and HHCPN (5 %).
The reaction conditions were selected on the basis of our
previous findings regarding the hydrogenation of HMF with various
Ni-Al2O3 catalysts.22 We have already shown that formation of
HCPN does not take place below 140 °C, but instead the
hydrogenation of both the aldehyde group and the furan ring of
HMF is favoured (THFDM yield = 100 % at 80 °C). Therefore, batch
experiments at 140 °C and 20 bar of hydrogen were performed in
order to evaluate the activity of the synthesised catalysts and
determine the preferred reaction pathways for the production of
cyclopentanone derivatives. During the course of the reaction,
hydrogen was continuously supplied to maintain the pressure
constant at 20 bar, guaranteeing H2 excess in the course of the
reaction. The effects related to the solubility of the reactant and the
intermediates in the aqueous phase were minimized by using a low
concentration HMF solution (0.04 M). Mass transfer limitations and
internal diffusion resistance were avoided by ensuring small
catalytic particle size (< 100 μm) and high stirring speed (600
rpm).22 Likewise, the amount of catalyst (0.06 g) and overall
reaction time (360 min) were also adjusted to detect significant
changes in the obtained product distributions. The choice of water
as solvent is indispensable to promote furan ring rearrangement
into cyclopentanone derivatives, a transformation which does not
take place when other solvents are used instead of H2O such as
CH2Cl2, MeOH, 1,4-dioxane, n-BuOH or THF.14,20b
A similar response has been reported using furfural as a
substrate.43,44 Cu-based catalysts favour hydrogenation of the C=O
bond whilst leaving the furan ring intact, leading to furfuryl alcohol
as the main product. On the other hand, Ni-based catalysts are
capable of hydrogenating the C=C bonds of the furan ring, leading
to tetrahydrofuran compounds. Higher selectivity for the
hydrogenation of the furan ring could be also attributed to the
lower size of Ni particles (d ≈ 9 nm) compared to Cu-Al2O3 (d ≈ 52
nm) and Co-Al2O3 (d ≈ 28 nm), since hydrogenation of unsaturated
rings tends to decrease as particle size increases due to lower
substrate adsorption.45,46
The high yield of HCPN achieved with the Cu-Al2O3 catalyst is
accompanied by the absence of any additional hydrogenation of the
C=O group that would lead to HCPL. By contrast, when Co-Al2O3 is
used, HCPN yield decreases to 5 %, whereas HCPL increases to 61 %.
Notably, the key intermediate FDM is consumed at a slower rate
over the Co-Al2O3 catalyst (29 %).
Table 2 summarizes the product distributions obtained from the
conversion of HMF over the synthesised catalysts. The calculated
carbon mass balance was higher than 94 % in all experiments,
confirming the absence of any significant undetected fraction of by-
products. Complete conversion of HMF is achieved with the Ni-
Al2O3, Cu-Al2O3 and Co-Al2O3 catalysts. The presence of reduced
metal phases on these materials (Figure 3b) ensures the initial
hydrogenation of the HMF aldehyde group towards the respective
alcohol (FDM), which is the key intermediate for the subsequent
formation of cyclopentanone derivatives. On the contrary, the
absence of such reduced metal phases on the catalysts containing
Zn and Mg (Figure 3b) renders them practically inactive in
hydrogenation of HMF (maximum conversion < 11%, Table 2).
Commercially available γ-Al2O3 (Sigma Aldrich) is also catalytically
In addition to the compounds listed in Table 2, traces of other
by-products
derived
from
C-O
hydrogenolysis
(3-
methylcyclopentanol and 5-methyl-2-furylmethanol) or C-C bond
cleavage (cyclopentanone and furfuryl alcohol) were also identified
in the product mixture at negligible yields (< 1%). These results
confirm that higher temperatures and stronger acidity (typically
Brønsted acid sites) are usually required to promote C-O
hydrogenolysis and C-C cleavage, due to the higher activation
energy of these steps.44
Figure 5 depicts the yield of the main products as a function of
the reaction time for the active catalysts (Ni-Al2O3, Cu-Al2O3 and Co-
Al2O3). All the kinetic curves show a progressive consumption of
Table 2. Product distributions from the conversion of HMF over the synthesised catalysts.(a)
Selectivity (%)
HMF
Catalyst conversion (%)
FDM
100
4
THFDM
HHD
HHCPN
HCPN
HCPL
γ-Al2O3
Ni-Al2O3
Cu-Al2O3
Co-Al2O3
Zn-Al2O3
Mg-Al2O3
1
-
19
-
1
-
-
0
7
0
-
-
0
5
0
-
-
59
79
5
-
15
1
61
-
100
100
100
4
5
29
86
95
-
11
-
-
-
-
-
(a) Reaction conditions: 0.23 gHMF, 45 ml H2O, 0.06 gcat, 140 °C, 20 bar H2, 600 rpm, 6 h. Carbon mass balance > 94 %.
6
| Green Chem., 2017, 00, 1-3
This journal is © The Royal Society of Chemistry 2017
Please do not adjust margins

Please do not adjust margins
Green Chemistry
Page 7 of 13
Journal Name
View Article Online
DOI: 10.1039/C7GC00315C
ARTICLE
HMF (concentration close to zero at the end of the experiment) Reaction scheme for the formation pathway of cyclopentanone
directly related to the selective hydrogenation of the aldehyde derivatives from HMF
group forming FDM as the first intermediate. In all cases, the yield
Based on the analysis of the kinetic curves, a reaction scheme is
of FDM reaches a maximum at short reaction times (60-120 min),
proposed to describe the major routes involved in the conversion of
followed by a continuous decrease as it is converted to products.
HMF to HCPN and HCPL. Firstly, the hydrogenation of the aldehyde
The rate of hydrogenation of HMF to FDM is significantly faster in
C=O bond takes place yielding FDM. This hydrogenation step is
the case of the Co-Al2O3 catalyst, reaching almost complete HMF
catalysed by metal species, and compared to furan ring
conversion after 1 h (98 %). On the contrary, the subsequent
hydrogenation is favoured by the absence of steric hindrance.46 The
conversion of FDM is significantly slower compared to Ni-Al2O3 and
subsequent conversion of FDM may proceed through: (i)
Cu-Al2O3 (Figure 5c).
hydrogenation of the furan ring, (ii) ring rearrangement or (iii) ring-
For all three catalysts tested, the ensuing consumption of FDM
opening (Figure 6). Complete hydrogenation of the furan ring yields
is accompanied by the formation of HCPN, detected only after FDM
THFDM, which is the end product of this pathway. This was
yield reaches a maximum (Figure 5). As previously discussed, HCPN
separately verified in a control experiment using THFDM as the
is the final major product for Cu-Al2O3 (79 %), whereas for Co-Al2O3
substrate (Table SI1; entry 1). No conversion of THFDM was
HCPN is further reduced to HCPL (61 %) via hydrogenation of the
observed after 6 h over Ni-Al2O3, conclusively demonstrating that
ketone C=O group. Importantly, the yield of HCPL shows
a
the formation of HCPN and HCPL proceeds via ring rearrangement
or ring opening of the key intermediate FDM.
pronounced increase once HMF is completely consumed, indicative
of the competitive hydrogenation between the aldehyde C=O group
of HMF and the ketone C=O group of HCPN (figure 5c). On the other
hand, Ni-Al2O3 results in the formation of both cyclopentanone
derivatives (HCPN; 59 % and HCPL; 15 %), along with THFDM (19 %).
The continuous increase of THFDM indicates a higher stability of
this compound, as a result of the difficulty in opening saturated
furan rings.44 By contrast, THFDM is practically not observed at any
stage of the reaction when Cu-Al2O3 or Co-Al2O3 are used as
catalysts.
The formation of HHCPN by ring rearrangement requires the
presence of water and acid sites (pathway ii, figure 6). The need of
the latter was proved by performing additional control experiments
using FDM as reactant (Table SI1; entries 2 - 4). HHCPN is not
formed in the absence of catalytically active sites (blank
experiment). By contrast, the experiments performed with γ-Al2O3
and Ni-Al2O3 (N2 atmosphere) produced HHCPN (selectivity 66 %
and 17 %, respectively), indicating that the presence of acid sites
promotes this step. On the other hand, when non-acidic
commercially available Ru/C catalyst was employed (Table SI1;
entry 5), HHCPN or HCPN were not detected and the sole product
identified was the fully hydrogenated THFDM (98 %). These results
are contrary to the reported furfural ring rearrangement, which
occurs spontaneously in hot water, without the need for catalyst or
hydrogen.20b,20c The subsequent transformation of HHCPN into
HCPN is catalysed by the combination of acid and metal sites, via
rapid deoxygenation (dehydration and/or C-O hydrogenolysis) and
hydrogenation steps (no intermediates were detected).
Finally, it should be underlined that HHCPN and HHD are
detected during the course of the reaction for all catalysts tested.
However, their concentration is decreasing once HCPN is formed
confirming their intermediate nature. Moreover, the yield of these
compounds reaches significant values only for Cu-Al2O3, indicating
that ring rearrangement and ring closure steps take place at a lower
rate. These findings suggest that two independent reaction
pathways occur in parallel depending on the type of the catalyst
employed. Contrary to the high selectivity towards THFDM
observed with Ni-Al2O3 at 80 °C,22 when the temperature is raised
to 140 °C rearrangement or opening of the furan ring of FDM is
favoured resulting in the targeted cyclopentanone derivatives.
The production of HHD from HMF in the presence of water has
been previously described to proceed through the acid-catalysed
rehydration of FDM followed by ring opening and hydrogenation of
the formed intermediates.47,48 Moreover, HHD may subsequently
undergo base-catalysed intramolecular aldol condensation followed
100
HMF
(a) Ni-Al O
(b) Cu-Al O
(c) Co-Al O
3
2
3
2
3
2
FDM
90
80
70
60
50
40
30
20
10
0
THFDM
HHD
HHCPN
HCPN
HCPL
0
60
120
180
240
300
360
60
120
180
240
300
360
60
120
180
240
300
360
Time (min)
Time (min)
Time (min)
Figure 5. Time evolution of HMF conversion and yield of products over the (a) Ni-Al2O3, (b) Cu-Al2O3 and (c) Co-Al2O3 catalysts (T = 140 °C;
PH2 = 20 bar).
This journal is © The Royal Society of Chemistry 2017
Green Chem., 2017, 00, 1-3 | 7
Please do not adjust margins

Please do not adjust margins
Green Chemistry
Page 8 of 13
View Article Online
DOI: 10.1039/C7GC00315C
ARTICLE
Journal Name
THFDM
(i) Hydrogenation M/H2
Deoxygenation
Hydrogenation
Hydrogenation
(ii) Ring rearrangement
Hydrogenation
Acid sites
Acid sites
M/H2
M/H2
M/H2
HHCPN
HCPN
HCPL
FDM
HMF
(iii) Ring opening Acid sites
Hydrogenation M/H2
Aldol condensation
Hydrogenation
Basic sites
M/H2
HHD
Figure 6. Proposed reaction scheme for the conversion of HMF into HCPN and HCPL.
by hydrogenation, producing HCPN.13 In order to validate this step, HCPN. Equally, as the temperature and the pressure are increased,
an additional control experiment was carried out over the Co-Al2O3 the yields of FDM and HHCPN decrease due to an enhanced
catalyst starting from a solution containing HHD/HCPN (35/59 mol conversion rate. Thus, for T > 160 °C and PH2 > 35 bar a highly
%) (Table SI1; entry 6). After 6 h, complete conversion of HHD into selective conversion of HMF into HCPN and HCPL is observed.
HCPN (3%) and mainly HCPL (82%) was observed, proving that HHD
According to Table 3, the highest yield of HCPN (86 %) is
is converted to final products via most likely an aldol condensation obtained at 180 °C and 20 bar H2. Nevertheless, excessive
reaction (pathway iii, figure 6). In this sense, the higher yield of HHD intensification of pressure and temperature leads to a decrease of
observed for the Cu-Al2O3 catalyst (Table 2), can be correlated with HCPN production due to the hydrogenation of the C=O ketone bond
the absence of high-strength basic sites in this material (Figure 4b), towards HCPL. For example, the reaction carried out at 180 °C and
which are known to promote aldol condensation reactions.41
50 bar H2 presents the lowest yield of HCPN (67 %) together with
The final production of HCPL rather than HCPN is strongly the highest production of HCPL (29 %). These results reveal that,
influenced by the hydrogenation ability of the catalyst. Therefore, unlike the hydrogenation of the furan ring, the hydrogenation of
the highest selectivity for HCPL shown by the Co-Al2O3 catalysts is the C=O ketone bond is not totally impeded over the Cu-Al2O3
associated with the superior interaction of Co particles with the catalysts and its rate can be tailored by varying the reaction
C=O ketone bond.20c Overall, this mechanistic study suggests that conditions.
both ring rearrangement and ring opening of FDM run in parallel
Influence of temperature and H2 pressure on HCPL production
and their relative rate strongly depends on the acid/base properties
of the catalyst (figure 6, ii and iii).
Concerning the production of HCPL, a similar study was
performed (T = 140 - 180 °C and PH2 = 20 - 50 bar) using the Co-
Al2O3 catalyst, which showed the highest selectivity to HCPL at 140
Influence of temperature and H2 pressure on HCPN production
The variation of the yield of HCPN at higher temperatures and °C and PH2 = 20 bar (61 %, Table 2). Table 4 lists the yields of the
H2 pressures was tested using the Cu-Al2O3 catalyst. Hydrogen main products obtained from the conversion of HMF over the Co-
pressure was varied between 20 and 50 bar ([H2] ≈ 0.421 - 0.942 Al2O3 catalyst under different reaction conditions.
cm3 g-1).49 A maximum temperature of 180 °C was selected in order
to keep the reactants and products in the liquid phase and avoid
Table 3. Yield of products from the conversion of HMF over the Cu-
undesirable C-C bond cleavage and coke formation, making the
Al2O3 catalyst. HMF conversion = 100 %.(a)
whole process less energy-efficient. Table
3 summarizes the
Yield (%)
T
(°C)
P
(bar)
Entry
obtained product distributions from HMF conversion over the Cu-
Al2O3 catalyst after 6 h of reaction at different temperatures and H2
pressures.
FDM HHCPN HHD
HCPN HCPL
1
2
3
4
5
6
7
8
140
160
180
140
160
180
140
160
180
5
1
0
4
0
0
5
0
0
5
6
0
2
0
0
0
0
0
7
5
4
0
0
0
0
0
0
79
82
86
81
85
76
76
71
67
1
20
35
50
3
HMF is fully converted to products and, importantly, THFDM is
not detected in any run. This fact confirms the selective attraction
of the Cu reduced metal phase to the carbonyl group and the strong
repulsion towards the furan ring which is not hydrogenated even at
higher H2 pressures.50 Consequently, a high selectivity towards
HCPN is observed, varying between 67 % and 86 %. HHD is only
observed in the final products mixture when the reaction is run at
low pressure (20 bar), suggesting a faster consumption of this
compound as hydrogen pressure increases. This can be related to
the hydrogenation step required for the conversion of HHD to
6
11
13
22
16
24
29
9
(a)
0.23 gHMF, 45 ml H2O, 0.06 gcat, 600 rpm, 6 h. Carbon mass
balance > 95 %.
8
| Green Chem., 2017, 00, 1-3
This journal is © The Royal Society of Chemistry 2017
Please do not adjust margins

Please do not adjust margins
Green Chemistry
Page 9 of 13
Journal Name
View Article Online
DOI: 10.1039/C7GC00315C
ARTICLE
When the temperature or the H2 pressure is raised above Characterization of used catalysts and reusability studies
standard conditions (T > 140 °C and PH2 > 20 bar), the carbon mass
The concentration of Cu and Co in the supernatant at the end of
balance drops below 95%, suggesting that a non-negligible fraction
the reaction (140 °C; 20 bar H2) was found less than 0.06 ppm
of products is not detected, in addition to the compounds already
(<0.01 wt % of metal content) according to ICP analysis. Moreover,
described (Figure 6). Most likely, oligomerization of reactive
in a typical hot filtration test, the Cu-Al2O3 and Co-Al2O3 catalysts
intermediates like FDM and HHCPN is also taking place, forming
were filtered off after 20 min (HMF conversion = 26 % and 66 %,
heavier organic molecules which are hard to detect by the
respectively), and the supernatant was left to react under the same
employed GC methods. Oligomerisation reactions are usually
conditions for additional 6 h. The analysis of the product mixture
promoted by the presence of acid sites at higher temperatures.7 In
after removing the catalyst did not show any additional HMF
this respect, the higher acidity showed by the Co-Al2O3 catalyst
conversion. Both results prove the absence of any metal leaching
(Figure 4a) lies in agreement with the augmented production of
into the liquid phase and confirm the heterogeneous nature of the
these heavier by-products.
catalytic system.
Nevertheless, a high yield of HCPL is still obtained when the
Samples of the used Cu-Al2O3 and Co-Al2O3 catalysts were
temperature is increased due to complete consumption of FDM.
recovered after reaction (T = 140 °C; P = 20 bar H2) by filtration and
Thus, the yield of HCPL varies within a narrow range of 84 - 89 % for
dried at room temperature. The TG/DTG curves of the spent Cu-
T > 160 °C after 6 h reaction. If the reaction is carried out at T = 140
Al2O3 sample (Figure SI5a) show a total loss of approximately 28 %.
°C, the formation of HCPL is slightly lower (61 - 77 %), since FDM is
In addition, the elemental microanalysis of the spent catalysts
not completely converted. Increasing the hydrogen pressure also
(Figure SI5a inset) reveals a low content of C (2.23 % w/w) together
has a significant impact, reducing the yield of FDM from 29 % to 3 %
with a relative high content of H (1.98 % w/w). Therefore, the TGA
as the pressure increases from 20 to 50 bar. This fact corroborates
measured weight loss cannot be solely attributed to organic
the existence of hydrogenation steps during the conversion of FDM
compounds deposited on the surface of the catalyst during the
into HCPN and HCPL.
reaction. The XRD pattern of the spent Cu-Al2O3 catalyst shows new
In order to improve the production of HCPL by avoiding the
additional peaks apart from the Cu metallic phase (Figure 7a) which
formation of undesirable oligomers, two additional experiments
have been ascribed to the formation of AlO(OH) boehmite structure
were carried out. The amount of catalyst added in the reaction
(JCPDS 83-1506). This point is corroborated by the FT-IR spectrum
medium was decreased from 0.06 g to 0.02 g and the reaction time
(Figure SI5b) which also exhibits the characteristic vibrational
was extended to 48 h (entries 10 and 11). The high carbon mass
modes of boehmite.51 More evidence of the formation of nano-
balance obtained in both experiments (> 97 %) suggests that the
AlO(OH) are provided by TEM (Figure 7b), detecting wire-like fibres
formation of oligomers practically does not occur. This fact is likely
(length = 15 - 25 nm; width = 2 - 5 nm) related with random
associated to the lower concentration of acid sites in the reaction
orientations of boehmite layers.
medium (0.02 vs. 0.06 g of catalyst), thus decreasing the extent of
acid-catalysed oligomerization steps.7 The highest yield of HCPL in
this work was obtained at 140 °C (reached 94 %, 48 h). However,
when the reaction was carried out at 180 °C, HCPL yield significantly
decreased (60 %, 48h). According to the kinetic curves presented in
Figure SI4, the hydrogenation of HCPN to HCPL is totally impeded
after 24 h and the Co-Al2O3 catalyst is deactivated. To this end, a
detailed investigation concerning the deactivation process of the
Cu-Al2O3 and the Co-Al2O3 catalysts was undertaken.
Table 4. Yield of products from the conversion of HMF over the Co-
Al2O3 catalyst. HMF conversion = 100 %.(a)
Yield (%)
T
(°C)
P
(bar)
Carbon mass
balance (%)
Entry
FDM THFDM HCPN HCPL
1
2
140
160
180
140
160
180
140
160
180
140
180
29
0
1
1
1
1
1
1
2
1
1
0
0
5
3
61
85
84
76
88
89
77
87
88
94
60
96
89
91
97
90
93
85
89
90
99
97
20
35
3
0
5
4
13
0
10
1
5
6
0
4
7
3
4
8
50
20
0
1
9
0
2
10(b)
11(b)
0
5
0
38
Figure 7. (a) XRD pattern and (b) TEM images of the used Cu-Al2O3
catalyst.
(a) 0.23 gHMF, 45 ml H2O, 0.06 gcat, 600 rpm, 6 h.
(b) 0.23 gHMF, 45 ml H2O, 0.02 gcat, 600 rpm, 48 h.
This journal is © The Royal Society of Chemistry 2017
Green Chem., 2017, 00, 1-3 | 9
Please do not adjust margins

Please do not adjust margins
Green Chemistry
Page 10 of 13
View Article Online
DOI: 10.1039/C7GC00315C
ARTICLE
Journal Name
Thus, we can conclude that the Cu-Al2O3 catalyst is partially bands of layered double hydroxide materials,54 which can be
hydrated in the course of the reaction forming the AlO(OH) attributed to the partial retention of the morphology of the
boehmite phase. These structural changes have been previously precursor as a consequence of the reported “memory effect” of
reported for impregnated metal alumina catalysts working under hydrotalcites.55 Likewise, some minor bands related to the
aqueous phase reforming conditions, leading to catalytic formation of hydrated boehmite were also observed at 1070 and
deactivation.52,53
730 cm-1. In accordance with the XRD pattern, TEM images of the
In order to study the reusability of the spent Cu-Al2O3 catalyst, used Co-Al2O3 catalyst reveal important differences with respect to
three consecutive reactions were conducted. Figure 8 represents the morphology of the fresh sample (Figure SI2c). Thus, the
HMF conversion and the yields of the main products after each run. previously observed relatively large particles of metallic cobalt (≈28
The results show a gradual decrease of HMF conversion from 100 % nm) are not seen in the micrograph represented in Figure 9b. In
to 76 % after three runs. Likewise, formation of HCPN decreases contrast, well-dispersed nanoparticles (2 - 4 nm) with spherical
from 81 to 21 % after 3 cycles and the yield of the key intermediate shape can be observed (Figure 9b). Therefore, it seems that during
FDM increases from 10 to 45 %. This progressive catalyst the reaction, the metallic cobalt particles undergo oxidation and
deactivation is associated with the observed hydration of the segregation, resulting in the formation of Co3O4 nanoparticles. This
alumina phase towards boehmite which involves reduction of both phenomenon could be related to the morphological changes
acidity and specific surface area, along with an agglomeration of observed in other Co-based catalysts, whose initial metallic particles
supported Cu particles.52 In this respect, TEM images of the spent are firstly transformed into hollow oxide spheres (based on
catalyst (Figure 7b inset) and EDX mapping (Figure SI6) show a Kirkendall effects), which finally break up into smaller particles
significant decrease in the dispersion of the Cu nanoparticles over under hydrogen atmospheres.56 The oxidation rate of the Co-Al2O3
the alumina support compared to the fresh material (Figure SI2b). catalyst could be reduced by adding promoters, such as Pd,57 or by
Regardless,
a significant restoration of catalytic activity was selecting an appropriate H2/H2O partial pressure ratio and
observed (Figure 8; run 4) after a regeneration treatment of the controlling cobalt crystallite sizes.58
used sample (air calcination 500 °C + H2 reduction 500 °C).
The reusability of Co-Al2O3 catalyst was also checked by
According to the X-ray diffraction of the regenerated catalyst performing a consecutive run after recovering the spent catalyst.
(Figure SI7a), the recovery of the catalytic activity can be associated Figure 10 shows the activity of the reused sample after drying at 20
to the removal of boehmite phase, maintaining a Cu particle size (≈ °C (Run 2a) and 120 °C (Run 2b). Thus, after the first run, a
45 nm) comparable to the initial catalyst.
pronounced decrease in HMF conversion (14 - 16 %) was observed
A similar characterization analysis was carried out on the used and HCPL was not formed at all. These results are in accordance
Co-Al2O3 catalyst (Figure 9). The obtained XRD pattern of the used with the catalytic activity showed by Zn-Al2O3 and Mg-Al2O3
Co-Al2O3 sample corresponds to an essentially amorphous material catalysts in which the absence of a reduced metallic phase limited
(Figure 9a). Compared to the X-ray diffraction pattern of the fresh the formation of further hydrogenated products.
material (Figure 3b), the diffraction peaks ascribed to the metallic
cobalt phase disappear completely, whereas traces of Co3O4 are
observed. The FT-IR spectrum (Figure SI8b) shows the characteristic
HMF
FDM
100
HCPN
80
60
40
20
0
Run 3
Run 1
Run 2
Run 4
Regeneration
Calcination: Air, 500 °C, 5 h.
Reduction: H2, 500 °C, 5 h.
Figure 8. Conversion of HMF and yields of FDM and HCPL after 6 h
Figure 9. (a) XRD pattern and (b) TEM images of the used Co-Al2O3
reaction over the Cu-Al2O3 catalysts (T = 140 °C, P = 20 bar).
catalyst.
10 | Green Chem., 2017, 00, 1-3
This journal is © The Royal Society of Chemistry 2017
Please do not adjust margins

Please do not adjust margins
Green Chemistry
Page 11 of 13
Journal Name
View Article Online
DOI: 10.1039/C7GC00315C
ARTICLE
reactions. Thus, the combination of these functionalities on both
Cu-Al2O3 and Co-Al2O3 favoured the formation of HCPN, which was
rapidly hydrogenated to HCPL in presence of Co0.
100
80
60
40
20
0
HMF
FDM
HCPN
HCPL
The reusability studies performed with the Cu-Al2O3 catalyst
suggested a progressive deactivation related to the formation of
hydrated boehmite phase (AlO(OH)). Likewise, the used Co-Al2O3
sample showed a poor activity after the first cycle due to the
formation of Co3O4 nanoparticles. However, both structural
changes were reversed by a simple regeneration treatment (air
calcination at 500 °C followed by a H2 reduction step), resulting in a
significant restoration of their catalytic activities.
Acknowledgements
This work was supported by the Engineering and Physical Sciences
Research Council (EPSRC), UK (EP/K014749). The MicroBioRefinery
(MBR, University of Liverpool) is kindly acknowledged for providing
the equipment for NH3 and CO2 chemisorption experiments.
Run 1
Run 2a Run 2b Run 3
Regeneration
Calcination: Air, 500 °C, 5 h.
Reduction: H2, 700 °C, 5 h.
Figure 10. Conversion of HMF and yields of FDM, HCPN and HCPL
after 6 h reaction over the Co-Al2O3 catalysts (T=140 °C, P=20 bar).
Run 2a after drying at 20 °C and Run 2b after drying at 120 °C.
References
1
2
R. Rinaldi and F. Schüth, ChemSusChem, 2009, 2, 1096.
R. W. Torget, J. S. Kim and Y. Y. Lee, Ind. Eng. Chem. Res.,
2000, 39, 2817.
Hence, the oxidation of metallic Co into Co3O4 nanoparticles during
the course of the reaction is most likely responsible for
deactivation. To reverse the catalyst deactivation, a regeneration
treatment (air calcination at 500 °C followed by H2 reduction at 700
°C) was applied to the used Co-Al2O3 catalyst. The XRD pattern of
the regenerated Co-Al2O3 sample (Figure SI7b) presents the peaks
associated to the Co metallic phase, with a mean particle size
around 20 nm. Thus, following these regeneration steps, the
catalyst completely recovered its activity (Figure 10; Run 3),
showing full conversion of HMF and a high selectivity to the
targeted HCPL (65 %).
3
4
5
K. Yan, G. Wu, T. Lafleur and C. Jarvis, Renew. Sust. Energ.
Rev., 2014, 38, 663.
Y. Román-Leshkov, J. N. Chheda and J. A. Dumesic, Science,
2006, 312, 1933.
R. J. van Putten, J. C. van der Waal, E. D. De Jong, C. B.
Rasrendra, H. J. Heeres and J. G. de Vries, Chem. Rev., 2013,
113, 1499.
6
7
8
J. P. Lange, E. van der Heide, J. van Buijtenen and R. Price,
ChemSusChem, 2012,
Y. Nakagawa, M. Tamura and K. Tomishige, ACS Catal., 2013,
, 2655.
A. Villa, M. Schiavoni, S. Campisi, G. M. Veith and L. Prati,
ChemSusChem, 2013, , 609; b) O. Casanova, S. Iborra and A.
Corma, ChemSusChem, 2009, , 1138.
5, 150.
3
Conclusions
6
The highly selective production of HCPN (86 %) and HCPL (94 %)
has been achieved over Cu-Al2O3 and Co-Al2O3, respectively,
emphasizing an efficient catalytic route to produce cyclopentanone
derivatives from HMF with readily available non-noble metals. The
followed synthesis method resulted in well dispersed metal phases
on alumina supports, exhibiting high specific surface areas and
acid/base properties. At T > 160 °C and PH2 > 35 bar, the Cu-Al2O3
catalyst showed a complete conversion of HMF into the targeted
HCPN and HCPL, whose yields can be tailored by varying the
reaction conditions. On the other hand, the Co-Al2O3 catalyst
showed a higher selectivity to the production of HCPL, a cyclic diol
whose formation from biomass derived resources is reported for
the first time.
2
9
R. Alamillo, M. Tucker, M. Chia, Y. Pagán-Torres and J. A.
Dumesic, Green Chem., 2012, 14, 1413; b) Y. Nakagawa, K.
Takada, M. Tamura and K. Tomishige, ACS Catal., 2014,
4
,
2718.
10 a) M. Chidambaram and A. T. Bell, Green Chem., 2010, 12
,
1253; b) Y. Román-Leshkov, C. J. Barrett, Z. Y. Liu and J. A.
Dumesic, Nature, 2007, 447, 982; c) D. P. Duarte, R. Martínez
and L. J. Hoyos, Ind. Eng. Chem. Res., 2015, 55, 54.
11 V. Choudhary, S. H. Mushrif, C. Ho, A. Anderko, V. Nikolakis,
N. S. Marinkovic, A. I. Frenkel, S. I. Sandler and D. G. Vlachos,
J. Am. Chem. Soc., 2013, 135, 3997; b) B. Girisuta, L. P. B. M.
Janssen and H. J. Heeres, Green Chem., 2006,
12 a) J. Tuteja, H. Choudhary, S. Nishimura and K. Ebitani,
ChemSusChem, 2014, , 96; b) T. Buntara, S. Noel, P. H. Phua,
8, 701.
The role of the metal and the acid/basic sites in the reaction
mechanism has been elucidated based on the correlation between
the main conversion pathways and the characterised catalytic
properties (Figure 6). Reduced metal phases were crucial to
promote the involved hydrogenation steps, whereas acid/basic sites
were proved to be necessary for the extent of ring-opening/closure
7
I. Melián-Cabrera, J. G. de Vries and H. J. Heeres, Top. Catal.
2012, 55, 612; c) S. Yao, X. Wang, Y. Jiang, F. Wu, X. Chen and
X. Mu, ACS Sust. Chem. Eng., 2013, 2, 173.
This journal is © The Royal Society of Chemistry 2017
Green Chem., 2017, 00, 1-3 | 11
Please do not adjust margins

Please do not adjust margins
Green Chemistry
Page 12 of 13
View Article Online
DOI: 10.1039/C7GC00315C
ARTICLE
Journal Name
13 J. Ohyama, R. Kanao, A. Esaki, A. Satsuma, Chem. Commun.,
2014, 50, 5633.
Jacobs, Y. Ji, B.H. Davis, D. Cronauer, A.J. Kropf, and C.L.
Marshall, Appl. Catal. A: Gen., 2007, 333, 177.
14 M. Hronec and K. Fulajtarová, Catal. Commun., 2012, 24, 38 H. A. Oliveira, D. F. Franceschini and F. B. Passos, J. Braz.
100;
Chem. Soc., 2014, 25, 2339.
15 M. Renz, Eu. J. Org. Chem., 2005,
6
, 979.
39 K. Lourvanij, PhD thesis, Partial dehydration of glucose to
oxygenated hydrocarbons in molecular-sieving catalysts,
Oregon State University, 1995.
16 K. A. Dubkov, G. I. Panov, E. V. Starokon, and V. N. Parmon,
React. Kinet. Catal. L., 2002, 77, 197.
17 T. Akashi, S. Sato, R. Takahashi, T. Sodesawa and K. Inui. 40 C. O. Veloso, C. N. Pérez, B. M. de Souza, E. C. Lima, A. G.
Catal. Commun., 2003,
4
, 411.
Dias, J. L. F. Monteiro and C. A. Henriques, Microp. Mesop.
Mater., 2008, 107, 23.
18 J. N. Chheda, G. W. Huber and J. A. Dumesic, Angew. Chem.
Int. Ed., 2007, 46, 7164.
41 J. I. Di Cosimo, V. K. Diez, M. Xu, E. Iglesia and C. R.
Apesteguia, J. Catal., 1998, 178, 499.
19 A. A. Rosatella, S. P. Simeonov, R. F. Frade, and C. A. Afonso,
Green Chem., 2011, 13, 754.
42 S. F. Wang, G. Z. Sun, L. M. Fang, L. Lei, X. Xiang and X. T. Zu,
Scientific Reports, 2015, 5.
20 a) M. Hronec, K. Fulajtárová, I. Vávra, T. Soták, E. Dobročka
and M. Mičušík, Appl. Catal. B: Environ., 2016, 181, 210; b) Y. 43 Y. Nakagawa and K. Tomishige, Catal. Commun., 2010, 12
Yang, Z. Du, Y. Huang, F. Lu, F. Wang, J. Gao and J. Xu, Green 154.
Chem., 2013, 15, 1932; c) X. L. Li, J. Deng, J. Shi, T. Pan, C. G. 44 S. Sitthisa and D. E. Resasco, Catal. Lett., 2011, 141, 784.
Yu, H. J. Xu and Y. Fu, Green Chem., 2015, 17, 1038; d) Y. 45 P. Gallezot, A. Giroir-Fendler and D. Richard, Catal. Lett.,
Wang, M. Zhou, T. Wang and G. Xiao, Catal. Lett., 2015, 145 1990, , 169.
1557; e) M. Zhou, H. Zhu, L. Niu, G. Xiao and R. Xiao, Catal. 46 P. Claus, Top. Catal., 1998,
,
,
5
5
, 51.
47 F. Liu, M. Audemar, K. D. O. Vigier, J. M. Clacens, F. De
Campo and F. Jérôme, ChemSusChem, 2014, , 2089.
Lett., 2014, 144, 235.
21 J. Ohyama, R. Kanao, Y. Ohira, A. Satsuma, Green Chem.,
2016, 18, 676.
7
48 F. Liu, M. Audemar, K. D. O. Vigier, J. M. Clacens, F. De
Campo and F. Jérôme, Green Chem., 2014, 16, 4110.
22 N. Perret, A. Grigoropoulos, M. Zanella, T. D. Manning, J. B.
Claridge and M. J. Rosseinsky, ChemSusChem, 2016,
9
, 521.
49 V. I. Baranenko and V. S. Kirov, At. Energ., 1989, 66, 30.
23 V. Rives and M. A. Ulibarri, Coord. Chem. Rev., 1999, 181, 61. 50 S. Sitthisa, T. Sooknoi, Y. Ma, P. B. Balbuena and D. E.
24 L. Balduzzi, F. Prinetto, G. Ghiotti, A. Bianchini, M. Livi and A.
Vaccari, Stud. Surf. Sci. Catal., 2001, 140, 67.
Resasco, J. Catal., 2011, 277, 1.
51 S. G. Wang, Y. Ma, Y. J. Shi and W. X. Gong, J. Chem. Tech.
Biotech., 2009, 84, 1043.
25 U. Constantino, F. Marmottini, M. Nocchetti and R. Vivani,
Eur. J. Inorg. Chem., 1998, 1439.
52 R. M. Ravenelle, J. R. Copeland, W. G. Kim, J. C. Crittenden
26 F. Cavani, F. Trifirò and A. Vaccari, Catal. Today, 1991, 11
,
and C. Sievers, ACS Catal., 2011, 1, 552.
173.
53 H. Li, Y. Zhao, C. Gao, Y. Wang, Z. Sun and X. Liang, Chem.
Eng. J., 2012, 181, 501.
27 J. J. Yu, Z. Jiang, L. Zhu, Z. P. Hao and Z. P. Xu, J. Phys. Chem.
B, 2006, 110, 4291.
54 J. T. Kloprogge and R. L. Frost, J. Solid State Chem., 1999,
146, 506.
28 A. Alejandre, F. Medina, P. Salagre, X. Correig and J. E.
Sueiras, Chem. Mater., 1999, 11, 939.
55 Q. Liu, B. Wang, C. Wang, Z. Tian, W. Qu, H. Ma and R. Xu,
Green Chem., 2014, 16, 2604.
29 Y. Lwin, M. A. Yarmo, Z. Yaakob, A. B. Mohamad and W. R.
W. Daud, Mater. Res. Bull., 2001, 36, 193.
30 V. Muñoz, F. M. Z. Zotin and L. A. Palacio, Catal. Today, 2015,
250, 173.
56 A. M. Saib, D. J. Moodley, I. M. Ciobîcă, M. M. Hauman, B. H.
Sigwebela, C. J. Weststrate, J. W. Niemantsverdriet and J.
Van de Loosdrecht, Catal. Today, 2010, 154, 271.
31 B. Jongsomjit, J. Panpranot and J. G. Goodwin, J. Catal., 2001, 57 T. Nowitzki, A. F. Carlsson, O. Martyanov, M. Naschitzki, V.
204, 98.
Zielasek, T. Risse, M. Schmal, H-J. Freund and M. Bäumer, J.
Phys. Chem. C, 2007, 111, 8566.
32 F. Kovanda, T. Rojka, P. Bezdička, K. Jirátová, L. Obalová, K.
Pacultová, Z. Bastl and T. Grygar, J. Solid State Chem., 2009, 58 E. van Steen, M. Claeys, M.E. Dry, J. van de Loosdrecht, E.L.
182, 27.
Viljoen and J.L. Visagie. J. Phys. Chem. B, 2005, 109, 3575.
33 F. E. López-Suárez, A. Bueno-López and M. J. Illán-Gómez,
Appl. Catal. B: Environ., 2008, 84, 651.
34 M. F. Luo, P. Fang, M. He and Y. L. Xie, J. Mol. Catal. A:
Chem., 2005, 239, 243.
35 W. Chu, P. A. Chernavskii, L. Gengembre, G. A. Pankina, P.
Fongarland and A. Y. Khodakov, J. Catal., 2007, 252, 215.
36 R. Bechara, D. Balloy, J. Y. Dauphin and J. Grimblot, Chem.
Mater., 1999, 11, 1703.
37 a) A.Y. Khodakov, J. Lynch, D. Bazin, B Rebours, N Zanier, B.
Moisson and P. Chaumette, J. Catal., 1997, 168, 16; b) G.
12 | Green Chem., 2017, 00, 1-3
This journal is © The Royal Society of Chemistry 2017
Please do not adjust margins

Page 13 of 13
Green Chemistry