Efficient synthesis of 2,5-dihydroxymethylfuran and 2,5-dimethylfuran from 5-hydroxymethylfurfural using mineral-derived Cu catalysts as versatile catalysts

Article abstract of DOI:10.1039/c5cy00700c

Selective conversion of 5-hydroxymethylfurfural (HMF) can produce sustainable fuels and chemicals. Herein, Cu-ZnO catalysts derived from minerals (malachite, rosasite and aurichalcite) were employed for selective hydrogenation of HMF for the first time. High yields of 2,5-dihydroxymethylfuran (~99.1%) and 2,5-dimethylfuran (~91.8%) were obtained tunably over the catalyst with a Cu/Zn molar ratio of 2, due to the well-dispersed metal sites tailored by mineral precursors, well-controlled surface sites and optimized reaction conditions. The relationship between catalytic performance and catalyst properties was elucidated by characterization based on the composition and the structural and surface properties, and catalytic tests. The catalyst can also be extended to selective hydrogenation of other bio-derived molecules (furfural and 5-methylfurfural) to target products. The construction of mineral-derived Cu-ZnO catalysts and the revelation of the structure-performance relationship can be applied to further rational design and functionalization of non-noble Cu catalysts for selective conversion of bio-derived compounds.

Full text of DOI:10.1039/c5cy00700c

Catalysis
Science &
Technology
View Article Online
View Journal
PAPER
Efficient synthesis of 2,5-dihydroxymethylfuran and
2,5-dimethylfuran from 5-hydroxymethylfurfural
using mineral-derived Cu catalysts as versatile
catalysts†
Cite this: DOI: 10.1039/c5cy00700c
c
c
ac
Yifeng Zhu,‡ab Xiao Kong,‡ab Hongyan Zheng, Guoqiang Ding, Yulei Zhu
and
*
Yong-Wang Liac
Selective conversion of 5-hydroxymethylfurfural (HMF) can produce sustainable fuels and chemicals.
Herein, Cu–ZnO catalysts derived from minerals (malachite, rosasite and aurichalcite) were employed for
selective hydrogenation of HMF for the first time. High yields of 2,5-dihydroxymethylfuran (~99.1%) and
2,5-dimethylfuran (~91.8%) were obtained tunably over the catalyst with a Cu/Zn molar ratio of 2, due to
the well-dispersed metal sites tailored by mineral precursors, well-controlled surface sites and optimized
reaction conditions. The relationship between catalytic performance and catalyst properties was elucidated
by characterization based on the composition and the structural and surface properties, and catalytic tests.
The catalyst can also be extended to selective hydrogenation of other bio-derived molecules (furfural and
5-methylfurfural) to target products. The construction of mineral-derived Cu–ZnO catalysts and the revela-
tion of the structure–performance relationship can be applied to further rational design and
functionalization of non-noble Cu catalysts for selective conversion of bio-derived compounds.
Received 13th May 2015,
Accepted 18th June 2015
DOI: 10.1039/c5cy00700c
www.rsc.org/catalysis
DHMF is a promising diol, which could be either directly
used for producing shape memory and self-healing polymers
Introduction
Growing interest has been paid to utilization of renewable
biomass to replace the diminishing fossil resources for energy
and chemical supply. 5-Hydroxymethylfurfural (HMF) is an
important platform chemical which could be facilely obtained
from renewable biomass resources (e.g., cellulose, glucose
and fructose) using acid catalysts.1,2 It contains many func-
tional groups including CO, C–O, CC and the furan ring,
which allow its hydrogenation to various products (e.g., 2,5-
dihydroxymethylfuran (DHMF), 2,5-dimethylfuran (DMF), 2,5-
dihydroxymethyltetrahydrofuran (DHMTHF) and 2,5-dimethyl-
tetrahydrofuran (DMTHF)).1 Among these derivatives, DHMF
and DMF attract great interest. In this paper, we aim at tun-
able synthesis of DHMF and DMF over the same non-noble
catalyst, which is challenging but valuable work for both aca-
demic and industrial purposes.
or used for synthesis of 1,6-hexanediol which is also an
important polymer precursor.3–5 Besides, it can also be
employed for production of intermediates of drugs and
crown ethers.6 Selective hydrogenation of HMF to DHMF is a
hydrogenation reaction of the aldehyde group. The side reac-
tions involving CC, C–O and furan ring hydrogenation
should be suppressed.
DMF is an alternative fuel with high energy density (30 kJ
cm−3) and octane number (RON = 119), holding great poten-
tial to substitute for petroleum-based gasoline.7 As a diene,
DMF can also be converted to valuable benzene-based
chemicals via Diels–Alder reactions.8,9 Production of DMF
from HMF involves CO hydrogenation and C–O hydro-
genolysis reactions, while CC hydrogenation and ring-
opening reactions need to be avoided.
Previous studies on selective hydrogenation of HMF to
DHMF and DMF were mainly performed on noble metal cata-
lysts, which is a great impediment to industrial applications.
For DHMF synthesis, Alamillo et al.10 studied supported Ru,
Pd and Pt catalysts for HMF hydrogenation, of which Ru was
the most active metal with a yield of 81–94%. An Au/Al2O3
catalyst was proposed for DHMF production with 96% yield
at 140 °C and 3.8 MPa.11 Other catalysts were also reported,
including Pt/MCM-41,6 Pd/C12 and Ir–ReOx/SiO2.13 For DMF
a State Key Laboratory of Coal Conversion, Institute of Coal Chemistry, Chinese
Academy of Sciences, Taiyuan 030001, PR China. E-mail: zhuyulei@sxicc.ac.cn;
Fax: +86 351 7560668; Tel: +86 351 7117097
b University of Chinese Academy of Sciences, Beijing 100049, PR China
c Synfuels China Co. Ltd, Beijing, 101407, PR China
† Electronic supplementary information (ESI) available: Table S1 and Fig. S1–S3.
See DOI: 10.1039/c5cy00700c
‡ These authors contributed equally.
This journal is © The Royal Society of Chemistry 2015
Catal. Sci. Technol.

View Article Online
Paper
Catalysis Science & Technology
production, Dumesic et al.7 reported for the first time a Ru–Cu
catalyst with a yield of 76–79%. Many subsequent reports were
also mainly conducted on noble metals (e.g., Pd,12 Ru,14,15 and
Pt16). However, few reports on the development of non-noble
substitutes existed, due to their low reactivity. Ni-based (e.g.,
RANEY® Ni,17 Ni/Al2O3 (ref. 18) and Ni–W2C/AC19) catalysts
emerged for DMF synthesis; on the other hand, they were diffi-
cult to be employed in DHMF synthesis due to the strong CC
hydrogenation ability of Ni. Recently, Cu/MgO/Al2O3 and Ru-
modified Cu/MgO/Al2O3 catalysts derived from hydrotalcite have
been employed in synthesis of DHMF and DMF.20,21 Neverthe-
less, the development of non-noble catalysts still requires con-
siderable effort prior to their commercial applications.
Scheme
hydrogenation.
1
Tunable synthesis of DHMF and DMF from HMF
catalysts can be extended to selective conversion of other bio-
derived molecules (furfural and 5-methylfurfural), demon-
strating the great potential of these catalysts for biomass
utilization.
Another challenge for HMF hydrogenation is the poor
selectivity to products due to the presence of many functional
groups of HMF, DHMF and DMF. To obtain high yields of
DHMF/DMF, the efficiency of CO hydrogenation and/or Experimental section
C–O hydrogenolysis needs to be promoted. Commonly used
noble metal (e.g., Ru, Pd) and Ni catalysts are good candi-
Preparation and evaluation
Copper nitrate trihydrate, zinc nitrate hexahydrate and
dates, whereas the saturation of the furan ring easily pro-
ceeds over these catalysts with strong CC hydrogenation
abilities.17,18,22 For example, DHMF yield would be reduced
by the formation of DHMTHF over RANEY® Ni and Ni–Al2O3
catalysts.17,18 Ru-modified Cu/MgO/Al2O3 catalysts were
reported for HMF hydrogenation with high reactivity.20 How-
ever, Ru sites would also promote CC hydrogenation reac-
tions and lower the DMF yield. Control of surface acidity is
also crucial. An appropriate amount of acid sites and high
reaction temperature would facilitate DMF formation via
hydrogenolysis of CH2OH groups, whereas CH2OH groups
need to be preserved for DHMF production.20,23 Besides,
excessive amounts of acid sites would catalyze the ring-
opening reactions and etherification, and thus lower the
selectivity to DHMF and DMF.10 Thus, the choice of metal,
suitable surface acidity and reaction conditions are important
for tunable synthesis of DHMF and DMF from HMF, in par-
ticular over the same catalyst. Cu catalysts are one of the best
materials for selective hydrogenation, due to their high reac-
tivity to CO and C–O bonds and relatively low reactivity to
CC and C–C bonds.24–26 ZnO, as a catalyst support, has
moderate surface acidity, facilitating the possibility of tun-
able synthesis of DHMF and DMF.
In view of the above problems, rational design of efficient
non-noble catalysts based on the composition, structure and
properties of materials is a prerequisite. Herein, we reported
that mineral-derived Cu–ZnO catalysts (malachite, rosasite
and aurichalcite) can efficiently catalyze the selective conver-
sion of HMF into either DHMF or DMF for the first time
(Scheme 1). This work addresses at least the following
desired features: (1) highly selective and tunable synthesis of
DHMF and DMF over the non-noble Cu catalysts was real-
ized; (2) highly dispersed Cu sites and high reactivity of the
catalysts can be obtained by using Cu–Zn binary minerals as
precursors; (3) the revelation of the structure–performance
relationship can facilitate further rational development of
active Cu catalysts. Besides, the mineral-derived Cu–ZnO
sodium carbonate were purchased from Sinopharm Co., Ltd.
and used as-received. The catalyst precursors were prepared
by a facile constant-pH co-precipitation method.27 A solution
containing copper nitrate and zinc nitrate with preset Cu/Zn
molar ratios was prepared. An aqueous solution of Na2CO3 (2
mol L−1) was employed as the precipitant. The two solutions
were then simultaneously introduced into a 2 L vessel at 65
°C and a pH of 6.5 0.2. After precipitation, the suspension
was aged for 150 min in the mother liquor and then washed
and dried at 65 °C for 12 h, followed by calcination at 350 °C
for 4 h in air. The resulting samples were denoted as CuZn-x,
of which x indicates the molar ratio of Cu/Zn. For compari-
son, the catalyst without Zn2+ added was denoted as CuZn-
inf, of which inf means infinite.
Prior to the tests, the catalysts were ex situ reduced in a
tube reactor under H2 atmosphere at 250 °C (heating rate of
0.5 °C min−1) for 2 h. The tests were performed in an auto-
clave reactor (100 mL) after the ex situ reduction of the cata-
lysts. Typically, 1.5 g of HMF (Shanghai Demo Pharmaceuti-
cal Science and Technology Co., Ltd., 98%), 35 mL of
1,4-dioxane (Sinopharm, AR) and 0.5 g of the catalyst were
placed into the reactor. The reactor was then sealed and
purged with H2 5 times. After that, the reactor was filled with
H2 at the desired pressure and heated to the objective tem-
perature. After the tests, the products were collected and ana-
lyzed by using a GC instrument with an FID.
Characterization
The elemental compositions were determined by inductively
coupled plasma-optical emission spectroscopy (ICP, Perkin
Elmer Optima 2100DV). XRD experiments were conducted on
a Rigaku D/max-2000 diffractometer with Cu Kα radiation
operating at 40 kV. The BET surface areas, pore diameters
and pore volumes of calcined catalysts were measured via N2
physical adsorption at −196 °C using a Micromeritics ASAP
2420 instrument after the samples were degassed at 90 °C for
Catal. Sci. Technol.
This journal is © The Royal Society of Chemistry 2015

View Article Online
Catalysis Science & Technology
Paper
1 h and at 350 °C for 8 h. H2 temperature-programmed
reduction (TPR) and N2O titration experiments were
performed on a Tianjin XQ TP-5080 instrument equipped
with a TCD. The surface acidity of the catalysts was deter-
mined by temperature-programmed desorption of ammonia
(NH3-TPD) using a Micromeritics ASAP 2920 instrument and
a mass spectrometric detector. The infrared (IR) spectra of
the catalysts were obtained by using a Nicolet NEXUS 470 FT-
IR spectrometer in the range of 400–4000 cm−1 with a resolu-
tion of 4 cm−1. The Raman spectra of the samples were
recorded using a LabRAM HR800 system equipped with a
CCD detector at room temperature. A He–Cd laser with a
wavelength of 325 nm was employed as the excitation source
with a power of 30 MW. Thermogravimetry (TG) experiments
of fresh reduced and used catalysts were performed on a
Mettler Toledo TGA/SDTA 851 instrument (air atmosphere
and heating rate of 10 °C min−1). Prior to the test, both the
fresh reduced and used catalysts were dried in air at 100 °C
for 24 h.
Scheme 2 Synthesis of Cu–ZnO catalysts with mineral precursors
using a constant-pH co-precipitation method (pH value was adjusted
by controlling the dropping speed).
Moreover, the contents of the active phases and the com-
positions are highly flexible in a wide range for the mineral-
derived catalysts (e.g., tunable Cu/Zn ratios and substitution
of Cu2+/Zn2+ with other divalent ions),33–35 facilitating further
functionalization of the catalysts. The present work demon-
strated the design and construction of Cu–ZnO catalysts from
some simple mineral precursors, which is a promising direc-
tion for catalyst design in selective hydrogenation of bio-
derived molecules.
Results and discussion
Brief introduction of the precursors
Malachite is a carbonate with a formula of ijCu2IJCO3)IJOH)2].
It contains two crystallographically different CuO6 octahedral
sites which are coordinated to six O atoms of CO3 and OH−
2−
ions in a Jahn–Teller distorted octahedron.28,29 When some
Zn2+ ions were added (Cu/Zn molar ratio >2.7), single-phase
rosasite with a general formula of ij(Cu,Zn)2IJCO3)IJOH)2] can
be obtained.30 Although it has similar composition to mala-
chite, rosasite is not isomorphous with malachite. The main
difference between them is the orientation of the Jahn–Teller
elongated axes of the CuO6 octahedra.31 Addition of higher
amounts of Zn2+ ions (Cu/Zn molar ratio <2.7) results in the
formation of aurichalcite which has a general formula of
ij(Cu,Zn)5IJOH)6IJCO3)2].29,30 Aurichalcite contains four crystal-
lographically different metal sites (Cu2+ and/or Zn2+) in the
lattice (denoted as M1, M2, M3 and M4), of which M1 and
M2 are coordinated to six O atoms, M3 is coordinated to four
O atoms and M4 is located at a five-fold trigonal bipyra-
mid.28,29 M3 is totally occupied by Zn2+ while the others can
be occupied by Cu2+ and Zn2+ with equal probability.28,32
Despite the different and complex formulas and struc-
tures, malachite, rosasite and aurichalcite minerals can be
tunably synthesized simply through a constant-pH co-precipi-
tation method by varying the Cu/Zn molar ratios (Scheme 2).27
Due to the homogeneous distributions of elements (Cu and
Zn) in the minerals and different morphologies, the above
minerals are important precursors for construction and
meso-/nano-structuring of Cu–ZnO catalysts.30 Herein, the
above minerals were obtained at different preset Cu/Zn molar
ratios (1, 2, 3, 4 and infinite). Upon calcination and reduc-
tion, highly dispersed Cu–ZnO catalysts with different Cu
loadings were obtained which exhibited high performance
for tunable production of DHMF (99.1%) and DMF (91.8%)
from HMF hydrogenation.
Characterization of catalyst precursors
The different catalyst precursors were prepared via modulat-
ing the preset Cu/Zn molar ratios (CuZn-x, x indicates Cu/Zn
ratios and inf means infinite). Fig. 1 shows the XRD spectra
of the catalyst precursors in the range of 10–40°, which serves
as the “fingerprint region” for identification of malachite,
rosasite and aurichalcite minerals.27,29,36 Evident peaks at
Fig. 1 XRD patterns of the precursors of CuZn-x catalysts.
This journal is © The Royal Society of Chemistry 2015
Catal. Sci. Technol.

View Article Online
Paper
Catalysis Science & Technology
14.8° (020), 17.6° (120), 24.1° (220), 31.2° (20−1), 32.2° (21−1)
and 35.6° (240) were observed for the CuZn-inf precursor,
which can be assigned to the characteristic peaks of pure
malachite (PDF# 41-1390). Addition of a small amount of
Zn2+ ions (CuZn-4 and CuZn-3) would lead to a structural
change from malachite to rosasite. According to Behrens
et al.,31 the (20−1) and (21−1) diffractions can be employed to
distinguish the structural evolution because the decrease in
cell volume would proceed when the Cu–O bond is
substituted with the Zn–O bond. The peaks at 31.2° (20−1)
and 32.2° (21−1) shifted toward higher angles (32.1° and 33.0°)
while the peak intensity decreased, indicating the substitution
of Cu2+ with Zn2+ in the lattice and the formation of crystalline
rosasite. Further addition of Zn2+ (CuZn-2 and CuZn-1) resulted
in the continuous decrease of peak intensity and emergence of
peaks at 13.0° and 34.3°. The two peaks which appeared belong
to the characteristic peaks of aurichalcite (PDF# 17-0743). Only
the aurichalcite phase was observed over CuZn-1. Together with
the increase of peak width, the intensity of XRD reflections
decreased, indicating that the crystallinity of minerals was
reduced gradually by the addition of Zn2+. XRD results revealed
that pure malachite was produced over CuZn-inf, pure rosasite
was formed over CuZn-4 and CuZn-3 and the mixed composite
of both rosasite and aurichalcite was observed over CuZn-2
while only aurichalcite could be detected over CuZn-1.
IR spectra allowed visualization of the structure evolutions
of the precursors (Fig. 2a and b). The bands in the region of
400–600 cm−1 are attributed to Cu–O and/or Zn–O bonds.28
With the increase of Zn2+ content, these bands broadened,
shifted to lower frequencies and decreased in intensity.
According to Stoilova et al.,28 this could be attributed to the
stronger Cu–O interactions and higher polarization ability of
Zn2+ ions, indicating the substitution of Cu2+ with Zn2+ ions.
The bands at 712 and 752 cm−1 are attributed to the ν4 asym-
metric OCO bending modes,28 which shifted to lower
wavenumbers when Cu2+ ions were replaced with Zn2+ ions.
The peak at 820 cm−1 is indicative of the ν2 mode of mala-
chite and rosasite, which diminished gradually with the
increase of Zn2+ content.29 Furthermore, it is interesting to
note that a slight remnant at 820 cm−1 was observed over
CuZn-1, indicating the existence of a slight amount of
rosasite over the sample which is lower than the detection
limit of XRD. A new peak at 840 cm−1 emerged over CuZn-2
and CuZn-1 which can be attributed to the ν2 mode of
aurichalcite. The bands at 970 and 1200 cm−1 can also serve
as the fingerprint bands of aurichalcite.28,29 For malachite
and rosasite, the bands in the range of 1300–1600 cm−1 are
Fig. 2 IR ((a) 400–1250 cm−1, (b) 1250–2050 cm−1) and Raman spectra
((c) intensity of CuZn-inf was divided
precursors.
3 times) of the CuZn-x
2−
attributed to ν4 modes of CO3 ions (Fig. 2b). When mala-
chite and rosasite minerals changed to aurichalcite, the
bands in this range would be complicated due to the exis-
tence of two crystallographically non-equivalent CO3 ions
Raman spectroscopy has been proven to be very useful for
studying the mineral precursors (Fig. 2c). The band at 1070
2−
2−
bonded to different ions (Cu2+ and Zn2+). The band at 1390
cm−1 disappeared while the bands at 1370 and 1560 cm−1
emerged with the addition of Zn2+ ions. The IR results con-
firmed the structural evolutions of the precursors, as
evidenced by XRD results.
cm−1 corresponds to the CO3 symmetric stretching mode of
aurichalcite while the peak at 1100 cm−1 belongs to the
rosasite phase.32,37 For CuZn-2, the overlapped peaks at 1070
and 1100 cm−1 were attributed to the mixed crystal phases of
aurichalcite and rosasite. A different motif of CuZn-inf was
Catal. Sci. Technol.
This journal is © The Royal Society of Chemistry 2015

View Article Online
Catalysis Science & Technology
Paper
observed in the range of 1050–1150 cm−1, which could be
ascribed to the overlapped peak of O–H out-of-plane bending
and ν1 CO3 symmetric stretching modes of malachite.33,34
2−
Raman results supported the observations of IR experiments.
Characterization of calcined and reduced catalysts
Table 1 gives an overview of the main physicochemical and
surface properties of CuZn-x catalysts. The elemental compo-
sitions and Cu/Zn molar ratios, determined by ICP experi-
ments, fitted well with the nominal values. CuZn-inf was
found to have a surface area (SBET) of ~18 m2 g−1 which was
approximately one-third of those for CuZn-x (x = 1, 2, 3, and
4) catalysts. The pore volume (Vp) of CuZn-inf was also lower
than those for the other samples. The results indicated that
Zn2+ tuned the structures of the precursors and the micro-
structures of the catalysts. The amounts of surface metallic
Cu sites (SCu) and surface acid concentrations (Sacid) were
determined by N2O and NH3-TPD experiments, respectively.
CuZn-inf showed the lowest SCu and Sacid. SCu increased with
the decrease of Cu/Zn molar ratios and reached the maxi-
mum at a Cu/Zn ratio of 2. When the Cu/Zn ratio was further
decreased, a decrease of SCu was observed. In contrast, Sacid
increased monotonously when Cu2+ was substituted with
Zn2+, indicating that the surface acidity mainly originated
from the ZnO component.38
Calcination of the precursors resulted in the formation of
oxide composites containing an intimate mixture of CuO and
ZnO species. Fig. 3 shows the XRD patterns of calcined
CuZn-x catalysts. The CuO reflections at 35.6° (002) and 38.7°
(111) were broad over all the calcined catalysts, indicating
that CuO particles over CuZn-x catalysts were small (PDF#
45-0937). Moreover, the intensity of CuO peaks decreased
and the full widths at half maximum of the peaks increased
with the addition of Zn2+, demonstrating the decrease of
CuO particle sizes and the role of Zn2+ in the dispersion of
the CuO species. No crystalline ZnO was detected over
CuZn-x (2, 3, 4 and inf) catalysts, indicating the high dispersion
of ZnO species. Only a weak peak of ZnO at 36.3° (PDF#
36-1451, crystalline size is 9.4 nm) was observed over
CuZn-1. Because the diffraction of CuO(002) at 35.6° greatly
overlapped with that of ZnO(101), we therefore used the
reflection of CuO(111) at 38.7° to quantitatively determine
Fig. 3 XRD patterns of the calcined CuZn-x catalysts (CuO, PDF# 45-
0937; ZnO, PDF# 36-1451).
the CuO particle sizes (dCuO) (Table 1). The mean CuO crystalline
sizes of CuZn-x (x = 1, 2, 3, and 4) were 6.1–7.4 nm, whereas that
of CuZn-inf was 11.4 nm. The above results demonstrated that
mineral precursors can facilitate the high dispersion of catalyst
components. To investigate the reducibility and structural evolu-
tions of the CuZn-x catalysts, TPR experiments were performed
(Fig. 4). CuZn-inf exhibited the highest reduction temperature as
a result of the CuO crystallites being the largest and poorly
porous structures.39 Concomitantly, the TPR profile of CuZn-inf
was different from that of CuZn-x (x = 1, 2, 3, and 4) catalysts,
which might be because of the absence of ZnO. The CuZn-x (x =
1, 2, 3, and 4) catalysts all exhibited a broad peak composed of
multi-shoulders. With the increase of ZnO content, the peaks of
the CuZn-x catalysts shifted slightly towards the lower tempera-
ture side, of which CuZn-1 showed the lowest reduction tempera-
ture. The results fitted well with the observations of XRD
results.40 For the CuZn-x catalysts, the shoulder at the higher
temperature side can be attributed to the reduction of Cu2+ spe-
cies which interacted strongly with ZnO species.41 The TPR
results revealed that the reduction behaviors can also be compli-
cated by the interactions between CuO and ZnO.
The surface acidity of the CuZn-x catalysts was determined
by NH3-TPD experiments (Fig. 5). A main peak at 180 °C and
Table 1 The main physicochemical and structural properties of CuZn-x catalysts
Cu contenta
(wt.%)
Zn contenta
(wt.%)
Cu/Zn molar
ratios
SBET
dp
Vp
dCuO
SCu
Sacid
(mmol gcat )
b
b
b
c
d
e
−1
−1
Catalysts
(m2 g−1
)
(nm)
(cm3 g−1
)
(nm)
(mmol gcat
)
CuZn-inf
CuZn-4
CuZn-3
CuZn-2
CuZn-1
79.6
63.8
60.0
53.7
40.3
—
—
18.0
52.4
56.9
55.7
55.8
20.7
14.6
12.9
11.5
13.3
0.17
0.27
0.25
0.24
0.26
11.4
7.4
7.1
6.2
6.1
0.39
1.55
2.03
2.12
1.78
3.70 × 10−4
6.20 × 10−4
1.06 × 10−3
1.16 × 10−3
1.27 × 10−3
15.9
19.5
26.3
39.4
4.1
3.2
2.1
1.1
a
b
c
Determined by ICP experiments. Determined by N2 adsorption experiments. Calculated by using the CuO(111) reflection at 38.7° based on
d
e
the Scherrer equation. The amounts of surface metallic Cu sites (SCu) were determined by N2O titration measurements. The surface acid
concentrations (Sacid) were determined by NH3-TPD measurements.
This journal is © The Royal Society of Chemistry 2015
Catal. Sci. Technol.

View Article Online
Paper
Catalysis Science & Technology
Fig. 4 TPR results of the calcined CuZn-x catalysts.
a small one in the range of 400–500 °C were observed for all
the catalysts, which can be ascribed to the weak and strong
acid sites, respectively. The amounts of surface acid sites
were enhanced when the Zn2+ content was increased. We cal-
culated the amounts of surface acid sites (Table 1) and found
that the CuZn-1 catalyst exhibited the highest surface acid
concentration (1.27 × 10−3 mmol gcat−1) while CuZn-inf only
possessed slight surface acidity (3.70 × 10−4 mmol gcat−3). The
results indicated that the surface acidity mainly originated
from the ZnO species.38 It should be noted that all the cata-
lysts exhibited relatively low densities of surface acidity
(about 3 orders lower than SCu), mainly due to the fact that
ZnO is an oxide with low surface acidity. The moderate sur-
face acid sites would facilitate the tunable production of
DHMF and DMF under different conditions, which would be
discussed in the following sections.
Fig. 6 Correlations between the main physicochemical properties (a,
normalized SBET and dCuO; b, normalized SCu, Sacid and Cu dispersion)
and Cu contents of CuZn-x catalysts (normalized: the highest is 100%).
malachite to rosasite, and then changed to a mixture of
rosasite and aurichalcite phases, which brought about the
different properties of the resulting catalysts. As shown in
Fig. 6a, addition of Zn2+ ions increased the SBET of the cal-
cined catalysts by ~3 times and decreased the mean CuO par-
ticle size by approximately 0.5 times. Cu dispersion increased
with the decrease of Cu content, which resulted in a compre-
hensive volcano trend of the surface metallic Cu concentra-
tion (Fig. 6b). CuZn-2 displayed the highest amount of sur-
face Cu sites. In contrast, the amount of surface acid sites
increased with the decrease of Cu content, indicating that
ZnO species may be responsible for the surface acidity.
Fig. 6 illustrated the correlations between Cu contents,
mineral structures and main structural and surface proper-
ties of the CuZn-x catalysts. When Cu2+ ions were replaced
with Zn2+ ions, the precursor structures evolved from
Performance of CuZn-x catalysts for DHMF synthesis
The selective hydrogenation of HMF to DHMF over CuZn-x
catalysts was conducted at 100 °C and 1.5 MPa for 1 h. The
performance of the CuZn-x catalysts was screened (Table 2).
The commercial RANEY® metal (Ni, Co and Cu) catalysts
were also employed as references. A low HMF conversion of
~16.3% was obtained over CuZn-inf. With the decrease of
Cu/Zn ratios, HMF conversion first increased, reached a maxi-
mum of 83.5% over the CuZn-2 catalyst and then decreased,
while DHMF selectivity was almost constant around 99%. The
Fig. 5 NH3-TPD results of the calcined CuZn-x catalysts.
Catal. Sci. Technol.
This journal is © The Royal Society of Chemistry 2015

View Article Online
Catalysis Science & Technology
Paper
Table 2 HMF hydrogenation over CuZn-x catalysts and commercial
RANEY® metals at 100 °Ca
The time course of the CuZn-2 catalyst at 100 °C and 1.5
MPa was studied to determine the highest reactivity and
DHMF yield. As plotted in Fig. 7, HMF conversion increased
from 83.5% to 100% upon increasing the reaction time from
1 h to 2 h, while the selectivity was almost unchanged. Fur-
ther increase of reaction time (up to 20 h) had little effect on
DHMF selectivity. The performance of the CuZn-x catalysts at
100 °C for 20 h was also compared (Fig. S2†), which all
exhibited high DHMF yields. Very slight amounts of hydro-
genolysis and ring-saturation by-products (MFA, DMF and
DHMTHF) were detected under the conditions. The results
revealed the high selectivity of Cu–ZnO catalysts towards syn-
thesis of DHMF at 100 °C.
Sel. (%)
Conv.
(%)
b
Others
Catalysts
DHMF
DHMTHF
CuZn-inf
CuZn-4
CuZn-3
CuZn-2
CuZn-1
RANEY® Cu
RANEY® Co
RANEY® Ni
16.3
34.1
75.4
83.5
75.3
60.0
63.9
73.1
98.1
97.9
98.9
99.1
99.2
98.9
97.2
69.5
0
0
0
0
0
0
2.7
20.1
1.9
2.1
1.1
0.9
0.8
1.1
1.2
10.4
a
b
Conditions: 1.5 MPa, 1 h, catalyst 0.5 g, HMF 1.5 g. Others mainly
include DMF and 5-methylfurfuryl alcohol (MFA).
Selective conversion of HMF to DMF
Selective hydrogenation of HMF to DMF involves the hydroge-
nation of the CO bond and hydrogenolysis of C–O bonds,
which is a more challenging reaction than DHMF synthesis.
DHMF is the main intermediate of DMF over Cu catalysts
while 2,5-dimethyltetrahydrofuran (DMTHF) is the over-
hydrogenated product.20
activity follows the similar trend of surface metallic Cu con-
centration (Fig. S1†). Interestingly, CuZn-1 and CuZn-3 pos-
sessed different surface metallic Cu concentration, but showed
similar activity towards HMF hydrogenation to DHMF.
According to Morrison et al.,42 acids can help in the activation
of carbonyl groups and thus facilitate the reactivity. The simi-
lar reactivity of CuZn-1 and CuZn-3 can be attributed to the
synergy between surface metallic Cu sites and acid sites. The
results revealed that ZnO not only dispersed Cu sites but also
provided some acid sites for promoting HMF hydrogenation.
It should be noted that bare ZnO without metallic Cu sites can
hardly catalyze HMF hydrogenation in our cases, as evidenced
by model tests (Table S1†).
The effectiveness of the CuZn-2 catalyst was also com-
pared with that of commercial RANEY® Cu, Co and Ni cata-
lysts under the same reaction conditions. Among the
RANEY® metals, RANEY® Ni exhibited the highest activity
with a HMF conversion of 73.1%, which was still lower than
that for the CuZn-2 catalyst. The superior performance of the
CuZn-2 catalyst indicated that mineral precursors could facil-
itate the formation of highly dispersed metal sites and high
catalytic activity. Moreover, the DHMF selectivity of RANEY®
Ni was much lower than that of the CuZn-2 catalyst, mainly
due to the formation of DHMTHF. DHMTHF is formed by
ring hydrogenation of DHMF. Thus, the formation of
DHMTHF over RANEY® Ni indicates the relatively strong
CC hydrogenation ability of Ni, which is detrimental to
DHMF synthesis.17 In contrast, CuZn-x and RANEY® Cu cata-
lysts exhibited high selectivity to DHMF, indicating that the
catalysts have high reactivity to CO bonds while having rel-
atively low reactivity toward CC hydrogenation. Extensive
studies regarding the configuration of adsorbed aldehydes
have been conducted over the Cu surface. The η1IJO) configu-
ration is known to be formed;43,44 thus, Cu-based catalysts
could exhibit high selectivity to CO hydrogenation. The
above results confirmed that mineral-derived Cu-based cata-
lysts are good candidates for selective hydrogenation of HMF
to DHMF.
DHMF was already obtained effectively over the CuZn-2
catalyst at 100 °C. The hydrogenolysis of C–O bonds would
be significantly promoted by elevation of reaction tempera-
tures, as reported in previous studies.17,20 The effect of reac-
tion temperature on the performance of the CuZn-2 catalyst
was therefore studied (Table 3). At 80 °C, 53.7% of HMF was
converted into DHMF with ~99.5% selectivity. The conversion
increased to 83.5% when the temperature was elevated to 100
°C and the DHMF selectivity was maintained around 99%.
With a further increase of reaction temperature to 140 °C,
HMF was almost totally converted while DHMF selectivity
dropped due to the formation of MFA and DMF. Further
increasing the reaction temperature tuned the product distri-
bution dramatically. DHMF selectivity decreased, whereas
MFA and DMF selectivity increased monotonously. The best
DMF selectivity of ~40.6% was obtained at 220 °C. Moreover,
Fig. 7 Time course of catalytic performance of the CuZn-2 catalyst at
100 °C (1.5 MPa H2, HMF 1.5 g, catalyst 0.5 g).
This journal is © The Royal Society of Chemistry 2015
Catal. Sci. Technol.

View Article Online
Paper
Catalysis Science & Technology
Table 3 Influence of reaction temperature on HMF hydrogenation over
reaction temperature, the generated DHMF was converted,
whereas MFA, DMF and a very slight amount of DMTHF were
formed (Table 3). When the reaction time was increased,
DHMF was continuously converted (Fig. 8). In contrast, MFA
was first formed and then converted, whereas DMF was
formed substantially. The yield of DMTHF increased slightly
when the reaction time was increased. The reaction pathway
is summarized in Scheme 3.
The above results demonstrate the tunable and highly
selective conversion of HMF to DHMF or DMF using a binary
mineral-derived Cu–ZnO catalyst under different conditions
(Scheme 1), confirming the potential of non-noble Cu cata-
lysts for selective hydrogenation of HMF. Key to our success
is the highly dispersed active sites, well-controlled surface
acidity and suitable reaction conditions.
CuZn-2 catalysta
Sel. (%)
T
Conv.
(°C) (%)
DHMF
MFA
DMF
DMTHF Others
80
100
140
180 100
220 100
53.7 99.5
83.5 99.1
97.5 94.4
0
0.2
3.6
20.9
24.1
0
0
1.8
27.4
40.6
0
0
0
1.6
1.9
0.5
0.7
0.2
0.1
0.3
50.0
33.1
a
Conditions: 1.5 MPa, 1 h, catalyst 0.5 g, HMF 1.5 g.
DHMF and MFA were obtained with a total yield of 57.2%,
which could be converted to DMF by hydrogenolysis of the
side –CH2OH group.17 In summary, the results showed that
high reaction temperatures would promote C–O hydro-
genolysis reactions, while the DMF yield could be further
improved. Therefore, 220 °C was chosen to further optimize
the DMF selectivity. The above results facilitated the poten-
tially tunable synthesis of DMF and DHMF over the Cu–ZnO
catalysts.
Fig. 8 showed the time course of catalytic performance of
the CuZn-2 catalyst at 220 °C and 1.5 MPa. Upon increasing
the reaction time from 1 h to 2 h, the DMF yield did not
change dramatically, whereas DHMF yield decreased with the
elevation of MFA yield. After 5 h reaction time, DHMF and
MFA were almost totally consumed, giving rise to a high
DMF yield of 91.8%, which indicates that C–O hydrogenolysis
was promoted. Upon further increase in reaction time, DMF
selectivity decreased slightly due to the formation of DMTHF.
Nevertheless, DMF selectivity was maintained at a high level
under our reaction conditions.
Catalyst stability
HMF hydrogenation to DMF was chosen to further investigate
the reusability of the CuZn-2 catalyst due to the relatively severe
conditions (220 °C). A sharp decrease of HMF conversion and
DMF selectivity was observed (Fig. 9a), indicating that the cata-
lyst deactivated during the tests. Thermogravimetry (TG) experi-
ments of fresh reduced and used CuZn-2 catalysts were
performed to examine the deposition of carbonaceous species
(Fig. 9b and c). The weight of fresh reduced CuZn-2 increased,
which could be ascribed to the oxidation of Cu. In contrast, the
dramatic weight loss of used CuZn-2 accompanied by a sharp
exothermic peak revealed the oxidization of carbonaceous spe-
cies.18 The carbonaceous species can be strongly adsorbed on
the catalyst and hinder the activity.19 The XRD patterns revealed
that Cu particles were not evidently grown after 5 runs (Fig. S3†).
Besides, the recovery processes (e.g., washing) would inevitably
cause weight loss of the catalyst. The results suggested that deac-
tivation of CuZn-2 mainly resulted from deposition of carbona-
ceous species, which was consistent with previous reports.14,18
The slight Cu particle growth and weight loss of the catalyst may
also cause the loss of reactivity to some extent. The deposited
carbonaceous species can be removed by re-calcination and re-
reduction procedures.14,45 Besides, design of an efficient catalyst
that can catalyze HMF hydrogenolysis at low temperatures can
promote the stability due to the minimization of humin
formation.
The above results revealed a reaction pathway for HMF
hydrogenation under the reaction conditions. HMF was ini-
tially hydrogenated to DHMF (Fig. 7). Upon increasing the
Fig. 8 Time course of catalytic performance of the CuZn-2 catalyst at
220 °C (1.5 MPa H2, HMF 1.5 g, catalyst 0.5 g).
Scheme
conditions.
3
Reaction pathway of HMF hydrogenation under the
Catal. Sci. Technol.
This journal is © The Royal Society of Chemistry 2015

View Article Online
Catalysis Science & Technology
Paper
furfuryl alcohol (98.9%) and 2-methylfuran (94.5%) were also
achieved through tuning of the reaction conditions. These results
further demonstrated the great flexibility of mineral-based Cu–
ZnO catalysts for efficiently and selectively converting bio-derived
molecules. Further improvement and rational design of mineral-
derived Cu catalysts are in progress.
Conclusions
The non-noble mineral-derived (malachite, rosasite and
aurichalcite) Cu–ZnO catalysts prepared by a simple con-
trolled co-precipitation method have been demonstrated to
have high reactivity and selectivity for the conversion of HMF
to either DHMF (~99.1% yield) or DMF (~91.8% yield). The
tunable synthesis of both products was achieved by simply
tuning the reaction temperature and reaction time. The struc-
tural evolutions of minerals and structure–performance rela-
tionships were elucidated by various characterization tech-
niques based on the composition and structural and surface
properties, and catalyst tests. The Cu–ZnO catalyst derived
from the mixture of rosasite and aurichalcite (CuZn-2) was
found to possess the highest reactivity, which is due to its
surface Cu concentration being the highest, suitable acidity
and good microstructures. The results here are expected to
be useful for further rational design of earth-abundant Cu
catalysts. The catalyst was also extended to selective hydroge-
nation of other bio-derived molecules (furfural and
5-methylfurfural) to valuable products, demonstrating the
flexibility of Cu catalyst for biomass conversion.
Fig. 9 (a) Reusability of the CuZn-2 catalyst for DMF synthesis from
HMF hydrogenation (conditions: 220 °C, 1.5 MPa, 5 h, HMF 1.5 g, cata-
lyst 0.5 g, 1,4-dioxane 35 ml). (b) Thermogravimetry (TG) spectra of
fresh reduced CuZn-2 catalyst. (c) TG spectra of used CuZn-2 catalyst.
Acknowledgements
This work was financially supported by the Major State Basic
Research Development Program of China (973 Program) (no.
2012CB215305).
Applications in conversion of bio-derived molecules
Notes and references
The CuZn-2 catalyst was further extended to selective hydrogena-
tion of other bio-derived furan molecules IJ5-methylfurfural and
furfural). As shown in Table 4, MFA and DMF were obtained with
high yields of 99.2% and 91.6%, respectively, under different
reaction conditions when 5-methylfurfural was employed as the
reactant. For selective hydrogenation of furfural, high yields of
1 R. J. van Putten, J. C. van der Waal, E. de Jong, C. B.
Rasrendra, H. J. Heeres and J. G. de Vries, Chem. Rev.,
2013, 113, 1499–1597.
2 J. N. Chheda, G. W. Huber and J. A. Dumesic, Angew. Chem.,
Int. Ed., 2007, 46, 7164–7183.
3 C. Zeng, H. Seino, J. Ren, K. Hatanaka and N. Yoshie,
Polymer, 2013, 54, 5351–5357.
Table 4 Applications of CuZn-2 catalyst in the conversion of furanic
compounds
4 C. Zeng, H. Seino, J. Ren, K. Hatanaka and N. Yoshie,
Macromolecules, 2013, 46, 1794–1802.
5 C. Zeng, H. Seino, J. Ren and N. Yoshie, ACS Appl. Mater.
Interfaces, 2014, 6, 2753–2758.
6 M. Chatterjee, T. Ishizaka and H. Kawanami, Green Chem.,
2014, 16, 4734–4739.
7 Y. Roman-Leshkov, C. J. Barrett, Z. Y. Liu and J. A. Dumesic,
Nature, 2007, 447, 982–985.
Feedstock
Conditionsa
Conv. (%)
100
Sel. (%)
100 °C, 2 h
99.2
220 °C, 10 h
100 °C, 2 h
100
100
91.6
98.9
8 P. T. Do, J. R. McAtee, D. A. Watson and R. F. Lobo, ACS
Catal., 2013, 3, 41–46.
220 °C, 10 h
100
94.5
9 Y.-P. Li, M. Head-Gordon and A. T. Bell, J. Phys. Chem. C,
2014, 118, 22090–22095.
a
1.5 MPa, catalyst 0.5 g, feedstock 1.5 g.
This journal is © The Royal Society of Chemistry 2015
Catal. Sci. Technol.

View Article Online
Paper
Catalysis Science & Technology
10 R. Alamillo, M. Tucker, M. Chia, Y. Pagan-Torres and J.
Dumesic, Green Chem., 2012, 14, 1413–1419.
28 D. Stoilova, V. Koleva and V. Vassileva, Spectrochim. Acta,
Part A, 2002, 58, 2051–2059.
11 J. Ohyama, A. Esaki, Y. Yamamoto, S. Arai and A. Satsuma,
RSC Adv., 2013, 3, 1033–1036.
29 M. Behrens, F. Girgsdies, A. Trunschke and R. Schlögl, Eur.
J. Inorg. Chem., 2009, 2009, 1347–1357.
12 J. Mitra, X. Zhou and T. Rauchfuss, Green Chem., 2015, 17,
307–313.
13 M. Tamura, K. Tokonami, Y. Nakagawa and K. Tomishige,
Chem. Commun., 2013, 49, 7034–7036.
14 J. Jae, W. Zheng, R. F. Lobo and D. G. Vlachos,
ChemSusChem, 2013, 6, 1158–1162.
15 Y. Zu, P. Yang, J. Wang, X. Liu, J. Ren, G. Lu and Y. Wang,
Appl. Catal., B, 2014, 146, 244–248.
16 G. H. Wang, J. Hilgert, F. H. Richter, F. Wang, H. J. Bongard,
B. Spliethoff, C. Weidenthaler and F. Schuth, Nat. Mater.,
2014, 13, 293–300.
17 X. Kong, Y. Zhu, H. Zheng, F. Dong, Y. Zhu and Y.-W. Li,
RSC Adv., 2014, 4, 60467–60472.
18 X. Kong, R. Zheng, Y. Zhu, G. Ding, Y. Zhu and Y.-W. Li,
Green Chem., 2015, 17, 2504–2514.
19 Y. B. Huang, M. Y. Chen, L. Yan, Q. X. Guo and Y. Fu,
ChemSusChem, 2014, 7, 1068–1072.
30 M. Behrens, J. Catal., 2009, 267, 24–29.
31 M. Behrens and F. Girgsdies, Z. Anorg. Allg. Chem.,
2010, 636, 919–927.
32 R. L. Frost, M. C. Hales and B. Jagannadha Reddy,
Polyhedron, 2007, 26, 3291–3300.
33 R. L. Frost, D. L. Wain, W. N. Martens and B. J. Reddy,
Spectrochim. Acta, Part A, 2007, 66, 1068–1074.
34 R. L. Frost, J. Raman Spectrosc., 2006, 37, 910–921.
35 R. L. Frost, B. J. Reddy, D. L. Wain and W. N. Martens,
Spectrochim. Acta, Part A, 2007, 66, 1075–1081.
36 S. Zander, E. L. Kunkes, M. E. Schuster, J. Schumann, G.
Weinberg, D. Teschner, N. Jacobsen, R. Schlögl and M.
Behrens, Angew. Chem., Int. Ed., 2013, 52, 6536–6540.
37 R. L. Frost, W. N. Martens, L. Rintoul, E. Mahmutagic and
J. T. Kloprogge, J. Raman Spectrosc., 2002, 33, 252–259.
38 Y. Zhu, X. Kong, X. Li, G. Ding, Y. Zhu and Y.-W. Li, ACS
Catal., 2014, 4, 3612–3620.
20 A. J. Kumalaputri, G. Bottari, P. M. Erne, H. J. Heeres and K.
Barta, ChemSusChem, 2014, 7, 2266–2275.
39 Y. Zhu, Y. Zhu, G. Ding, S. Zhu, H. Zheng and Y. Li, Appl.
Catal., A, 2013, 468, 296–304.
21 T. S. Hansen, K. Barta, P. T. Anastas, P. C. Ford and A.
Riisager, Green Chem., 2012, 14, 2457.
40 L. Shi, C. Zeng, Y. Jin, T. Wang and N. Tsubaki, Catal. Sci.
Technol., 2012, 2, 2569.
22 Y. Nakagawa, K. Takada, M. Tamura and K. Tomishige, ACS
Catal., 2014, 4, 2718–2726.
41 J. Agrell, K. Hasselbo, K. Jansson, S. G. Järås and M.
Boutonnet, Appl. Catal., A, 2001, 211, 239–250.
23 E. R. Sacia, M. Balakrishnan and A. T. Bell, J. Catal.,
2014, 313, 70–79.
42 R. Morrison and R. Boyd, Organic Chemistry, Prentice Hall,
Englewood Cliffs, NJ, 6th edn, 1992.
24 Y. Zhu, X. Kong, D.-B. Cao, J. Cui, Y. Zhu and Y.-W. Li, ACS
Catal., 2014, 4, 3675–3681.
43 S. Sitthisa, T. Sooknoi, Y. Ma, P. B. Balbuena and D. E.
Resasco, J. Catal., 2011, 277, 1–13.
25 E. S. Vasiliadou, T. M. Eggenhuisen, P. Munnik, P. E. de
Jongh, K. P. de Jong and A. A. Lemonidou, Appl. Catal., B,
2014, 145, 108–119.
26 H. Tan, M. N. Hedhill, Y. Wang, J. Zhang, K. Li, S. Sioud,
Z. A. Al-Talla, M. H. Amad, T. Zhan, O. E. Tall and Y. Han,
Catal. Sci. Technol., 2013, 3, 3360.
44 Y. Nakagawa, H. Nakazawa, H. Watanabe and K. Tomishige,
ChemCatChem, 2012, 4, 1791–1797.
45 B. J. O'Neill, D. H. Jackson, A. J. Crisci, C. A. Farberow, F.
Shi, A. C. Alba-Rubio, J. Lu, P. J. Dietrich, X. Gu, C. L.
Marshall, P. C. Stair, J. W. Elam, J. T. Miller, F. H. Ribeiro,
P. M. Voyles, J. Greeley, M. Mavrikakis, S. L. Scott, T. F.
Kuech and J. A. Dumesic, Angew. Chem., Int. Ed., 2013, 52,
13808–13812.
27 B. Bems, M. Schur, A. Dassenoy, H. Junkes, D. Herein and R.
Schlögl, Chem. – Eur. J., 2003, 9, 2039–2052.
Catal. Sci. Technol.
This journal is © The Royal Society of Chemistry 2015