Inexpensive but Highly Efficient Co–Mn Mixed-Oxide Catalysts for Selective Oxidation of 5-Hydroxymethylfurfural to 2,5-Furandicarboxylic Acid

Article abstract of DOI:10.1002/cssc.201800989

A highly active and inexpensive Co–Mn mixed-oxide catalyst was prepared and used for selective oxidation of 5-hydroxymethylfurfural (HMF) into 2, 5-furandicarboxylic acid (FDCA). Co–Mn mixed-oxide catalysts with different Co/Mn molar ratios were prepared through a simple solid-state grinding method—a low-cost and green catalyst preparation method. The activity of these catalysts was evaluated for selective aerobic oxidation of HMF into FDCA in water. Excellent HMF conversion (99 %) and FDCA yield (95 %) were obtained under the best reaction conditions (i.e., 120 °C, 5 h, Co–Mn mixed-oxide catalyst with a Co/Mn molar ratio of 0.25 calcined at 300 °C (Co-Mn-0.25) and 1 MPa O₂). The catalyst could be reused five times without a significant decrease in activity. The results demonstrated that the catalytic activity and selectivity of the Co–Mn mixed-oxide catalysts prepared through solid-state grinding were superior to the same Co–Mn catalyst prepared through a conventional coprecipitation method. The high catalytic activity of the Co-Mn-0.25 catalyst was attributed to its high lattice oxygen mobility and the presence of different valence states of manganese. The high activity and low cost of the Co–Mn mixed-oxide catalysts prepared by solid-state grinding make it promising for industrial application for the manufacturing of polyethylene furanoate, a bioreplacement for polyethylene terephthalate, from sustainable bioresources.

Full text of DOI:10.1002/cssc.201800989

Accepted Article
Title: Inexpensive but highly efficient Co-Mn mixed oxide catalysts
prepared by simple solid-state mixing for selective oxidation of 5-
hydroxymethylfurfural to 2,5-furandicarboxylic acid
Authors: Chunbao (Charles) Xu, Tirumala Venkateswara Rao
Kasanneni, Jennifer Lorraine Rogers, Sadra Souzanchi, Luna
Dessbesell, and Madhumita Ray
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: ChemSusChem 10.1002/cssc.201800989
Link to VoR: http://dx.doi.org/10.1002/cssc.201800989
A Journal of
www.chemsuschem.org

ChemSusChem
10.1002/cssc.201800989
FULL PAPER
Inexpensive but highly efficient Co-Mn mixed oxide catalysts prepared by
simple solid-state mixing for selective oxidation of 5-hydroxymethylfurfural
to 2,5-furandicarboxylic acid
Kasanneni Tirumala Venkateswara Rao, Jennifer Lorraine Rogers, Sadra Souzanchi,[a] Luana
[
a]
[a]
Dessbesell, Madhumita Bhowmick Ray,[a] Chunbao (Charles) Xu*
[a]
[a]
involve either lengthy and expensive multi-step operations or
liquid/gas waste streams in the preparation processes. Therefore,
development of low-cost and greener catalyst preparation
methods is highly desirable.
Abstract: This work demonstrated high activity of inexpensive
Co-Mn mixed oxide catalysts for selective oxidation of 5-
hydroxymethylfurfural (HMF) into 2, 5-furandicarboxylic acid (FDCA)
–
a sustainable alternative to terephthalate. The Co-Mn mixed oxide
catalysts with different Co/Mn mole ratios were prepared simply by
solid-state grinding; a low-cost and greener catalyst preparation
method. The activity of these catalysts was evaluated for selective
aerobic oxidation of HMF into FDCA in water solvent. Under the best
On the other hand, utilization of biorenewable feedstock for
sustainable production of chemicals and transportation fuels
represents a promising way to reduce our dependence on fossil
fuels that are not only polluting but also depleting. For instance,
o
reaction conditions (i.e., 120 C, 5 h, Co-Mn mixed oxide catalyst with
Co/Mn molar ratio of 0.25 calcined at 300C (Co-Mn-0.25) and 1 MPa
7
C -C15 liquid alkanes could be selectively produced from biomass
2
O ), 99% HMF conversion, and 95% FDCA yield were obtained. The
derived carbohydrates by acid-catalyzed dehydration to form 5-
catalyst could be reused five times with an insignificant decrease in
activity. The results demonstrated that the catalytic activity and
selectivity of the Co-Mn mixed oxide catalysts prepared by solid-state
grinding, depending on Co/Mn molar ratio, are superior to the same
Co-Mn oxide prepared by conventional co-precipitation method. It was
found that the high catalytic activity of Co-Mn-0.25 catalyst can be
attributed to its high lattice oxygen mobility and the presence of
different valence states of manganese. The high activity and low-cost
of the Co-Mn mixed oxide catalysts prepared by solid-state grinding
make it promising for industrial application in the manufacturing of
polyethylene furanoate (PEF) – a bio-replacement for PET from
sustainable bioresources.
hydroxymethylfurfural (HMF), followed by aldol condensation over
solid base catalysts to produce long-chain alkanes.1 Recently,
intensified interest and effort have been devoted toward the
production of HMF from lignocellulose biomass or biomass
2
derived C
6
-carbohydrates such as either glucose or fructose, as
furan derivatives are ranked by the U.S. Department of Energy
DOE) in the “top 10 list of bio-based chemicals” in terms of their
(
market potential. HMF is a versatile intermediate platform
precursor for the synthesis of a wide range of biomass-based
alternative fuels and chemicals. HMF has two reactive functional
groups, i.e., an alcoholic and an aldehyde moiety, which allow
different types of chemical reactions like reduction, hydrogenation,
oxidation, and condensation to produce different value added bio-
fuel and bio-based chemicals. For example, it was demonstrated
by the authors’ group that HMF could replace formaldehyde to
1
. Introduction
Heterogeneous catalysts, in particular supported metallic or metal
oxide catalysts, are indispensable to many industrial processes
produce formaldehyde-free phenolic resins for composite
materials. 3 Selective oxidation of HMF can produce several
(such as petroleum refining) and many types of environmental
remediation, from vehicle emissions control systems to industrial
effluent and municipal waste treatment. The petroleum refinery
industry is the largest single user of catalysts, especially in fluid
catalytic cracking, hydro-processing and reforming or
isomerization processes for production of chemicals and
transportation fuels. The global market of energy and
environmental catalysts was worth $13.0 billion in 2009, and
the market was projected to grow at an average annual growth
important furan chemicals, and among them 2,5-furandicarboxylic
acid (FDCA) has attracted significant attention as it could be used
as a monomer to produce bio-based polyethylene furanoate
(PEF), which is considered as potential bio-replacement for fossil
based polyethylene terephthalate (PET) for the manufacturing of
packaging materials (e.g., bottles) and other value-added
4
products .
rate
of
12.8%
(http://www.the-
Selective oxidation of HMF to FDCA has gained growing research
interest in recent years . Both homogeneous and heterogeneous
infoshop.com/study/bc21724_energy_applications.html).
4
Supported metallic or metal oxide catalysts are conventionally
prepared by deposition of active component(s) onto the catalyst
support using wet impregnation, co-precipitation and CVD
methods. In contrast, the conventional methods for
heterogeneous catalysts preparation have limitations as they
noble metal (Au, Pd, Pt and Ru) catalysts have been explored for
4-10
oxidation of HMF to FDCA.
However, homogeneous catalysts
suffer from poor reusability, whereas the cost of noble metal
catalysts restricts their practical application in synthesis of FDCA.
Alternatively, inexpensive heterogeneous cobalt, iron and
manganese based catalysts proved to be effective for this
reaction.1
1-17
Among them manganese (Mn) based catalysts were
found to be highly active, owing to their different oxidation states
[
a]
Dr. K. T. V. Rao, J. L. Rogers, Dr. S. Souzanchi, L. Dessbesell, Prof.
M. Ray, Prof. C. Xu
and high oxygen storage capacity as well as faster oxygen
18-20
absorption and oxide reduction.
Nevertheless, these catalysts
Department of Chemical and Biochemical Engineering
Western University
were mostly prepared through a conventional co-precipitation
technique, where lengthy, multi-step operations and a significant
amount of waste were involved, as previously stated. Thus, it is
highly desirable to develop a low-cost and greener catalyst
preparation method for synthesis of manganese oxide based
catalysts for selective oxidation of HMF to FDCA.
London, Ontario, Canada N6A 5B9
E-mail: cxu6@uwo.ca
Supporting information for this article is given via a link at the end of
the document.
This article is protected by copyright. All rights reserved.

ChemSusChem
10.1002/cssc.201800989
FULL PAPER
Very recently, a simple and low-cost solid-state grinding method
has been reported for preparation of supported-metal or metal
0.25 (Mn/Co molar ratio=0.25) catalyst whose catalytic activity
was compared with Co-Mn-0.25 catalyst, in addition to FDCA,
FFCA was also detected, however the FFCA selectivity is far
lower than that of FDCA for most of the catalysts. As indicated in
oxide catalysts for various reactions. 2
1-23
The novel preparation
method via solid-state grinding for heterogeneous catalysts is not
only simple and low-cost, but also green and sustainable with
zero-waste emission during the catalyst preparation. Hence, it is
greatly advantageous over the conventional co-precipitation
techniques for heterogeneous catalyst preparation. Moreover, the
solid-state grinding method has great potential for large-scale
industrial applications. To the best of the authors’ knowledge,
cobalt-manganese (Co-Mn) mixed oxide catalysts prepared by
solid-state grinding for selective oxidation of HMF to FDCA had
not been studied previously.
Table 1, mono oxide of CoO
this result proved that MnO
x
failed to convert HMF to FDCA and
is a principal catalyst and Co is a
x
promoter for Co-Mn mixed to improve FDCA yield substantially.
With Co-Mn-0.25 catalyst, the FDCA yield as high as 95% was
obtained, which is the highest ever reported over any non-noble
catalysts in literature and even comparable with that of noble
metal based catalysts.4
Table 1. Catalytic activity of Co-Mn mixed oxide catalysts with various Co/Mn
o
molar ratios for selective oxidation of HMF at 120 C and 5 h under 1 MPa O
2
andEntry
Catalyst
HMF
conversion
FDCA yield
(%)
FFCA yield (%)
In this work, Co-Mn mixed oxide catalysts were prepared by
simple solid-state grinding of a mixture of Mn and Co-precursors
with oxalic acid (OA) at a specified molar ratio followed by drying
and calcination. The as-synthesized Co-Mn mixed oxide catalysts
were employed in selective aerobic oxidation of HMF into FDCA
in water medium. The influences of different reaction parameters
(%)
1
2
3
4
5
6
7
8
9
CoO
MnO
MnO a
x
36.5±3.8
> 99
-
-
x
46.1±5.1
80.8±2.2
75.7±3.5
95.2±1.1
89.1±2.7
86.8±4.2
83.9±1.7
34.02±2.91
33.8±2.6
(such as Co/Mn molar ratio, calcination temperature, reaction
x
> 99
-
temperature and time, oxygen pressure, etc.) have been studied
to maximize the FDCA product selectivity and yield. The
structural and surface properties of Co-Mn mixed oxide catalysts
were extensively investigated using various catalyst
characterization techniques to understand the relationship
between the structure, redox properties and catalytic performance.
Furthermore, a two-step process for the catalytic conversion of
carbohydrates into FDCA was tested, and the potential of
industrial production of FDCA from bioresources (starch and
cellulosic biomass) was examined.
Co-Mn-0.16
Co-Mn-0.25
Co-Mn-0.33
Co-Mn-0.5
> 99
3.91±1.8
> 99
-
> 99
-
> 99
-
Co-Mn-0.75
Mn-Co-0.25
> 99
-
> 99
23.46±3.75
2. Results and Discussion
a18 h reaction time.
2
.1 The catalytic performance of Co-Mn mixed oxide
2.1.2. Effects of the catalyst calcination temperature
catalysts in selective oxidation of pure HMF into FDCA
Calcination temperature for the supported catalysts could
significantly affect their catalytic properties . In this study, the
2
.1.1.Effects of Co/Mn molar ratio.
19
The catalytic performance of Co-Mn mixed oxide catalysts with
various Co/Mn molar ratios prepared simply by solid-state
grinding was evaluated for selective oxidation of pure HMF under
highly active Co-Mn-0.25 catalyst was calcined in air at various
o
temperatures in the range of 300 - 600 C and the corresponding
catalytic activities were examined for selective oxidation of HMF
1
MPa O in water with 0.5 mmol NaHCO as mild base additive
2
3
o
o
at 120 C for 5 h. As shown in Fig. 1, all catalysts obtained
at 120 C for 5 h unless stated otherwise. 5-HMF contains two
reactive oxidizable functional groups (-OH and -CHO) that can
produce different oxidation products, but selectivity of the
oxidation of HMF to FDCA is greatly influenced by the Co/Mn
molar ratio of the Co-Mn mixed oxide catalysts, as shown in Table
complete conversion of HMF (> 99%), but the FDCA yield varied
depending on calcination temperature of the catalyst: the FDCA
yield decreased from 95% to 35% with increasing calcination
o
temperature from 300 to 600 C; as similarly reported in a recent
17
1. It should be mentioned that in the absence of catalyst no FDCA
x 2
study on HMF oxidation over MnO -CeO catalyst . Along with
was detected. All mixed oxide catalysts and the mono oxide MnO
catalyst led to a nearly complete conversion of HMF (> 99%), but
the selectivity towards FDCA varied and strongly depended on
x
FDCA, FFCA was also detected as a side-product with the Co-
Mn-0.25-600 catalyst. This variation in catalytic selectivity with
calcination temperature (as shown here in Fig. 1) could be
attributed to the difference in structural properties of the Co-Mn-
x
Co/Mn molar ratio. The mono oxide MnO catalyst for 5 h reaction
obtained approx. 46% and 34% yields of FDCA and 2,5-
formylfurancarboxylic acid (FFCA), respectively, whereas 81%
yield of FDCA could be achieved for 18 h reaction. In contrast,
addition of cobalt as a promoter to MnO
binary mixed oxide catalysts enhanced the FDCA yield
substantially relative to mono oxide of MnO . The role of Co in this
mixed oxide catalyst is to increase the reducibility of MnO
thereby improve the redox properties of Co-Mn mixed oxide
catalysts, which will be discussed in the H -TPR results. As
0.25 catalyst, e.g., its crystallinity, BET surface area and the
lattice oxygen mobility, etc., which will be discussed in the
following Section 2.2. The results above demonstrate that the
x
to formulate Co-Mn
o
catalyst (Co-Mn-0.25) calcined at 300 C exhibited best
performance for HMF oxidation to FDCA. As such, the Co-Mn-
0.25 (300C calcination) was chosen as the best catalyst for
further investigation on the effects of other reaction parameters
such as reaction temperature, time, addition of base
compounds, oxygen pressure, etc.
x
x
,
2
observed, the FDCA yield increased from 75% to 95% when the
Co/Mn molar ratio was increased from 0.16 (Co-Mn-0.16) to 0.25
(
0
Co-Mn-0.25). With further increase of the Co/Mn mole ratio from
.33 to 0.75, the differences in FDCA yield among those catalysts
2
.1.3. Effects of reaction temperature
Effects of reaction temperature were examined for HMF
oxidation over Co-Mn-0.25 catalyst between 90 - 130 C for 5 h.
As expected, reaction temperature has a great influence on
were not significant, and in fact a decrease in FDCA yield was
observed at a higher Co/Mn mole ratio (> 0.25). On the other hand,
manganese was added as a promoter to CoO to derive Mn-Co-
x
o
This article is protected by copyright. All rights reserved.

ChemSusChem
10.1002/cssc.201800989
FULL PAPER
selective oxidation of HMF to FDCA. The oxidation products Also In our study, poor carbon mass balance was obtained over
o
were mainly FDCA, FFCA and traces of 2,5-diformylfuran (DFF). CoMn-0.25 catalyst at lower reaction temperatures (90 & 100 C)
o
However, no HFCA was detected under these reaction after 5 h reaction (Table 2, Entry 1 & 2). Accordingly, 120 C for
conditions, as shown in Table 2. Thus, the reaction might 5 h appeared to be the best reaction condition for FDCA
proceed through a pathway involving DFF intermediate instead production by selective oxidation of HMF over Co-Mn-0.25
of
2,5-hydroxymethylfurancarboxylic
acid
(HFCA) catalyst.
intermediate. As shown in the Table 2, both HMF conversion
17
and FDCA yields increased with rising reaction temperatures
Table 2. Effect of reaction temperature on selective oxidation of HMF over
o
o
from 90 to 120 C. Initially at 90 C, the HMF conversion and
FDCA yields were 64% and 4.6%, respectively, whereas at 120
Co-Mn-0.25 catalyst for 5 h
o
C these values increased to > 99% and 95%, respectively. It
Reaction
temperature
(oC)
HMF
conversion
(%)
FDCA yield
(%)
FFCA yield
(%)
Entry
indicates that selectivity of the catalyst for FDCA increased
drastically with reaction temperature. While further increasing
o
the reaction temperature to 130 C, the FDCA yield declined to
9
1% possibly due to HMF degradation at high reaction
1
2
3
4
5
90
64.0±3.6
93.2±2.9
99.1±0.8
> 99
4.6±1.8
32.1±2.7
86.4±2.6
95.2±1.1
90.8±2.4
21.8±2.6
24
temperatures. In addition to FDCA, FFCA also was formed,
and in fact, it was the dominant oxidation product at lower
reaction temperatures (90 C). However, its selectivity
100
110
120
130
29.8±1.9
o
1.2±0.5
decreased with increasing reaction temperature, and deceased
o
at > 110 C. Similarly, traces of DFF (< 1%) were detected at 90
-
-
o
C but not observed at >100 oC. Consequently, 120 oC was
actually the best reaction temperature for production of FDCA
by selective oxidation of HMF.
> 99
Addition of a base compound to the reaction medium could
4
positively influence the production of FDCA . In the present
HMF conversion (%) FDCA yield (%) FFCA yield (%)
investigation, we examined the effects of different base
compounds with varying strengths on selective oxidation of
1
00
o
HMF into FDCA over the Co-Mn-0.25 catalyst at 120 C for 5
h. As it can be seen from Fig. 2b, generally the yield of FDCA
decreased gradually with increasing base strength, as reported
80
60
40
20
0
16
previously . Weak bases such as NaHCO
found to be more effective for producing a high FDCA yield (95-
6%). In contrary, strong and medium range bases such as
NaOH, KOH, Na CO and K CO are less effective for selective
3 3
and KHCO were
9
2
3
2
3
oxidation of HMF to FDCA, although almost 100% HMF
conversion was achieved in all runs irrespective of the type of
base compound used, suggesting that the presence of a strong
base in the reaction medium promoted formation of more by-
16
products that are undetectable by HPLC . Conversely, bi-

+
+
3
carbonate (HCO ) bases irrespective of the cation (Na or K )
are the most effective additive for HMF to FDCA
transformation. Furthermore, effects of the additional amount
of NaHCO or the NaHCO /HMF molar ratio in the initial
3 3
3
00
400
500
600
o
Catalyst calcination temperature ( C)
Figure 1. Effects of catalyst calcination temperature on the activity of Co-Mn-
reaction medium on selective oxidation of HMF were
investigated with Co-Mn-0.25 catalyst at 120 C for 5 h, and
o
o
0.25 catalyst for selective oxidation of HMF at 120 C for 5 h.
the results are presented in Fig. S2. In the absence of an
external base, low conversion of HMF (67.2%) together with a
low FDCA yield (5.6%) was obtained, signifying the importance
of base additive in the selective oxidation of HMF4 . As can
be also seen from Fig. S2, the HMF conversion and FDCA
yield increased with increasing amount of the base compound
and attained the maximum values (approx. 100% and 95%,
respectively) at the base/HMF molar ratio of 2.0. With further
increase in base/HMF ratio from 2 to 3, FDCA yield decreased
slightly likely due to excess alkali that could promote
degradation of 5-HMF during the reaction. Therefore, the
optimum base/HMF ratio for this reaction over Co-Mn-0.25
catalyst appears to be 2.
2
.1.4. Effects of reaction time, addition of base compounds,
oxygen pressure and catalyst loading
,16
The effects of reaction time (ranging from 1 h to 6 h) on HMF
oxidation were investigated in the presence of Co-Mn-0.25
catalyst and NaHCO
as two major oxidation products in the experiments, and no HFCA
was observed, though traces of DFF (< 0.5%) were detected at
the initial stage of the reaction (1 h). It is postulated that the -OH
group of HMF could be first oxidized, followed by the oxidation of
the -CHO group. As shown in Fig. 2a, FDCA yield increased with
o
3
at 120 C. FDCA and FFCA were detected
prolonging the reaction time and reached a maximum (95%) after
5
o
h, although at 120 C HMF was completely converted in all
experiments even for a short time of 1 h. For a short reaction time
or at the initial stage of the reaction, the carbon mass balance in
the liquid phase (as determined by HPLC analysis) was poor. This
was likely due to to the strong adsorption of HMF on the catalyst
surface. Poor carbon balance was also reported by Besson and
co-workers in a study on HMF oxidation over Pt/C catalysts.25
The performance of the Co-Mn-0.25 catalyst in HMF oxidation
was also examined at a different oxygen pressure level at 120
C for 5 h, and the results are presented in Fig. S3. An increase
of oxygen pressure (from 0.5 to 2 MPa) led to increase in FDCA
yield to a certain extent. However, the yield plateaued off with
o
This article is protected by copyright. All rights reserved.

ChemSusChem
10.1002/cssc.201800989
FULL PAPER
the maximum at 1-1.5 MPa, though the HMF conversion was
nearly at 100% under at all oxygen pressure.
HMF conversion (%) FDCA yield (%) FFCA yield (%)
(a)
1
00
Effects of catalyst dosage on selective oxidation of HMF to
FDCA over Co-Mn-0.25 catalyst were also examined, while
keeping the other reaction parameters constant. As shown in
Fig. S4, varying the catalyst amount by changing the
HMF/catalyst weight ratio (w/w) from 0.31 to 1.26, i.e.,
decreasing the catalyst loading, FDCA yield is first increased
and then decreased. Whereas, there is an optimal catalyst
loading for greater production of FDCA, corresponding to
HMF/catalyst weight ratio of 0.63.
80
6
0
40
2
0
2.1.5. Solid-state grinding vs. conventional liquid-state co-
precipitation
0
1
2
3
4
5
6
The effectiveness of the Co-Mn-0.25 mixed oxide catalyst
prepared by simple solid-state grinding – a low-cost and
greener catalyst preparation method - was compared with that
of the same catalyst prepared by conventional liquid-state co-
precipitation method (Co-Mn-0.25-P). As shown in Fig. 3, both
catalysts though prepared by different methods achieved
complete conversion of HMF, but exhibited a remarkable
difference in products selectivity. The Co-Mn-0.25 catalyst
prepared by solid-state grinding is superior to Co-Mn-0.25-P,
prepared by liquid-state co-precipitation. The lower FDCA
selectivity for the Co-Mn-0.25-P could be mainly ascribed to its
Reaction time (h)
(
b)
HMF conversion (%)
FDCA yield (%)
1
00
8
0
0
6
2
2
smaller BET surface area (81 m /g vs. 154 m /g for Co-Mn-
.25 (Table S1) and its lower oxygen mobility, as will be
discussed in Section 3.2.
40
0
2
0
2.1.6. The catalyst stability and reusability
0
NaOH
KOH NaHCO3 KHCO3 Na2CO3 K2CO3
In this work, recyclability of the Co-Mn-0.25 catalyst was
8
Catalyst
evaluated by the following procedure, after the initial HMF
o
oxidation experiment at 120 C for 5h, the catalyst was
Figure 2 a. Effects of reaction time on HMF oxidation in the presence of Co-
Mn-0.25 catalyst and NaHCO
base compounds on HMF oxidation with the Co-Mn-0.25 catalyst at 120 C for
5h.
recovered from the reaction mixture by centrifugation, washed
at 120 oC, and b. effects of addition of various
3
o
with distilled water and ethanol, vacuum dried at 60 C for 12
o
h, and was then reused directly for the same reaction at the
same conditions. Repeating the procedure above for 5 times,
the FDCA yields from all five cycles are illustrated in Fig. 4. It
can be observed from Fig. 4 that the FDCA yield decreased
1
00
nd
rd
from 95% to 88% (the 2 cycle) and 83% (the 3 cycle), but
80
rd
remained almost constant after the 3 cycle, indicating
HMF conversion (%)
FDCA yield (%)
FFCA yield (%)
excellent stability of the Co-Mn-0.25 catalyst in the HMF
oxidation experiments. The reaction solution was analyzed by
ICP, and negligible concentrations of Mn and Co were detected
in the solution, demonstrating good chemical stability of Co-
Mn-0.25 catalyst with negligible leaching of active metals from
the catalyst during the reaction process.
6
0
0
4
20
0
Grinding Co-precipitation
Figure 3 Effects of catalyst (Co-Mn-0.25) preparation method on HMF oxidation
o
at 120 C for 5 h
This article is protected by copyright. All rights reserved.

ChemSusChem
10.1002/cssc.201800989
FULL PAPER
100
80
60
40
20
0
a
(
8)
1
2
3
4
5
Catalytic cycles
o
(7)
Figure 4. Recyclability of Co-Mn-0.25 catalyst for HMF oxidation at 120 C for
h.
5
2.2. Characterization of the fresh and spent catalysts
In order to understand the difference in catalytic selectivity of the
Co-Mn binary mixed oxides with varying Co/Mn molar ratios (as
shown previously in Table 1), these catalysts were
comprehensively characterized. Generally, oxidation reactions
over a metal oxide catalyst follow the Mars-Van-Krevelen
(
6)
mechanism involving exchange between the external O
2
and the
oxide’s lattice oxygen. Therefore, the catalytic activity of MnO
and Co-Mn mixed oxide catalysts would strongly depend on the
2
6
x
(
5)
1
9,27,28
mobility and accessibility of the lattice oxygen
characterized by H -TPR analysis. Fig. 5a displays the H
profiles of mono oxides of MnO and CoOx, and Co-Mn mixed
oxide catalysts with different Co/Mn molar ratios. The TPR profile
of MnO shows two-stage reduction, with the first reduction peak
, which can be
2
2
-TPR
x
x
o
at 350 C that could be ascribed to the reduction of MnO
2
2 3
/Mn O
o
to Mn
of Mn
also exhibits two reduction peaks at 358 C and 518 C, ascribed
3
O
4
and the second peak at 497 C is related to the reduction
(4)
29
3
4 2 x
O to MnO. The H -TPR profile of mono oxide of CoO
o
o
3+
2+
2+
30
to the reduction of Co to Co and Co to Co, respectively.
Interestingly with the Co-Mn mixed oxide catalysts, the addition of
CoO to MnO markedly promotes the reducibility of the catalysts
x
x
by significantly shifting both reduction peaks to lower
temperatures by 50-100C, compared with those of the mono
o
oxide of MnO
x
or CoO
x
. However, the high temperature (>650 C)
(
3)
peak in Co-Mn mixed oxide catalysts could attributed to the
reduction of spinel structure of Co-Mn mixed oxide.31 It is
reasonable to infer that the low temperature reduction peak in Co-
Mn mixed oxide catalysts resulted from the reduction of surface
oxygen species generated by the presence of oxygen
32
vacencies. It is noteworthy to mention that high oxygen mobility
favours oxygen transfer on the surface of the catalysts, hence
facilitating the catalytic oxidation reactions. 19,33 As clearly shown
in the figure, the reduction pattern of Co-Mn mixed oxides also
depends on the Co/Mn molar ratio. With a higher Co/Mn molar
ratio (> 0.5), the reduction temperature is slightly shifted to higher
temperature which indicates low oxygen mobility. From the above
results, one may conclude that the highest FDCA yield and
selectivity obtained with Co-Mn-0.25 catalyst can be explained by
the lowest reduction temperature, or the highest oxygen mobility
of the catalyst. It should be noted that other three catalysts with
different Co/Mo molar ratios (Co-Mn-0.16, Co-Mn-0.33 and Co-
Mn-0.5) also exhibit very low lattice oxygen reduction temperature,
but their selectivity towards FDCA is lower than that of the Co-Mn-
(2)
1)
(
0.25.
This article is protected by copyright. All rights reserved.

ChemSusChem
10.1002/cssc.201800989
FULL PAPER
2
Mn-0.25 catalyst is the highest (154 m /g), compared with the
2
2
other catalysts: Co-Mn-0.16 (83 m /g), Co-Mn-0.33 (132 m /g),
and Co-Mn-0.5 (127 m /g).
b. Mn 2p
2p3/2
2
2p1/2
The superb catalytic activity of the Co-Mn-0.25 catalyst was
further investigated by XPS analysis. The survey scan spectra of
the Co-Mn-0.25 catalyst consisting of Co, Mn and O elements are
presented in Figure S1. As shown in Fig. 5b, the Mn 2p spectra
show two distinguish binding energy opeaks that can be
recognized as Mn 2p3/2 and Mn2p1/2. The deconvoluted spectra of
Mn2p3/2 consists of two different binding energy values one at
2
Mn
+
3
Mn
+
6
37.8 eV and the other at 640.7 eV. These two peaks may
2
+
3+
characterize Mn and Mn species, respectively. The energy
separation between two spin-orbit components Mn 2p3/2 and Mn
2p1/2 is 11.4 eV, which is in good agreement with the existing
34
literature. However, compared to literature reports, the binding
energies of both Mn2 and Mn were shifted to lower values due
to doping of cobalt into MnOx. This led to a decrease in the binding
energy of manganese species as a result of formation of cobalt-
manganese spinel like structure.34 It also further indicates that
manganese is not in the form of individual oxide but in the form of
mixed oxide phase, as will be discussed in XRD results. In the
XPS spectra of Co (Fig. 5c), the doublets of Co 2p3/2 and Co 2p1/2
peaks are observed at 776.5 and 791.8 eV, with a separation of
binding energy of 15.3 eV, representing the coexistance of
+
3+
c. Co 2p
2+
3+ 35
Co /Co . Interestingly, the deconvolution of Co 2p, the peaks
3+
located at 776.5 and 798.6 eV belong to Co , and other peaks at
2+
7
2
81.9 and 791.8 eV indicate the Co . The H -TPR results further
support the XPS results, where the reduction pattern indicates the
presence of different valence states of manganese.The
deconvoluted O 1s spectra (Fig. 5d) of Co-Mn-0.25 catalyst show
two peaks. The binding energy peak particularly at 527.5 eV can
3
6
be assigned mainly to the lattice oxygen , although the presence
of other oxygen species such as adsorbed oxygen and surface
oxygen species cannot be ruled out. It has been known that
oxygen mobility of manganese oxide based catalyst strongly
depends on the transforming capacity of manganese species
between different oxidation states, for example between
d. O 1s
3+
4+
2+
3+ 37
Mn /Mn or Mn /Mn . According to the TPR results of the
catalysts, the reduction of higher oxidation states preferred to take
place at a lower temperature. It means that high concentration of
surface Mn3 would introduce more active oxygen that would be
+
related to the catalytic performance. Since CoO
effective for selective oxidation of HMF to FDCA, but the addition
of cobalt to MnO catalyst greatly promoted the activity of the
mono oxide of MnO for the reaction. It may thus be concluded
x
alone was not
x
x
that superb catalytic activity of Co-Mn-0.25 catalyst could be
attributed to the coexistence of highly mobile lattice oxygen and
surface Mn3 and Mn species.
+
2+
XRD patterns of CoOx, MnOx and Co-Mn mixed oxide catalysts
displayed in Fig. 6. The diffraction peaks of CoO
at 2θ equal to 31.5, 36.9, 38.6, 44.9, 59.5, and 65.5 ,
corresponding to spinel structure of Co (JCPDS No. 74-1656).
Whereas, the XRD pattern of MnO catalyst mainly exhibits
amorphous nature, however, it can show some crystalline peaks
x
catalyst exhibit
o
3
O
4
Figure 5 a. H
2
-TPR patterns of mono oxides and Co-Mn mixed oxides: (1)
x
, (3)Co-Mn-0.16, (4)Co-Mn-0.25, (5)Co-Mn-0.33, (6) Co-Mn-0.5,
x
MnO ,(2) CoO
x
and (7) Co-Mn-0.75, (8) Mn-Co-0.25; b. Mn-2p XPS spectra of Co-Mn-0.25
catalyst; c. Co-2p XPS spectra of Co-Mn-0.25 catalyst; d.O 1s XPS spectra of
Co-Mn-0.25 catalyst.
o
o
o
at 33 , 36.7 and 66 which can repersent the presence of Mn
3 4
O
(JCPDS 24-0734) and Mn (JCPDS 41-1442). On the other
2
O
3
hand, Co-Mn mixed oxide catalysts prepared through either solid-
state grinding or precipitation method displaying the characteristic
small and broad peaks at 2θ = 36.5 and 65.14, can be assigned
2
On the other hand, a distinct H -TPR pattern was observed
among Co-Mn-0.25 and Mn-Co-0.25 catalysts, however, the later
catalyst exhibits high oxygen mobility compared to the former.
The superior performance of the Co-Mn-0.25 catalyst in selective
oxidation of HMF to FDCA is likely owing to the combined effects
of the high lattice oxygen mobility/reducibility and a high surface
area. As illustrated in Table S1, the BET surface area of the Co-
2 4
to the spinel structure of (Co,Mn)(Co,Mn) O , which is in good
agreement with earlier work.38 Intensity of these XRD peaks
increases with increasing calcination temperature, suggesting
increased crystallinity of the catalysts calcined at a higher
temperature, as expected (Fig. S5).
This article is protected by copyright. All rights reserved.

ChemSusChem
10.1002/cssc.201800989
FULL PAPER
The TEM image of fresh Co-Mn-0.25 catalyst is shown in Fig. 7a.
The particle sizes are in nano meter range (4 to 14 nm). The
average particle size for Co-Mn-0.25 catalyst is about 6.5 nm
This variation in catalytic selectivity with calcination temperature
(as shown previously in Fig. 1) could be attributed to variation in
structural properties of the catalysts during thermal treatment
(calcination). In fact, the increase in crystallinity (Fig. 6) and
decrease in BET surface area (Table S1) of Co-Mn-0.25 catalyst
by thermal treatment at a higher calcination temperature may
account for the decline of catalytic selectivity. However, the redox
properties of the catalysts could play a major role in selective
(measured through Image software) which is in a good agreement
with a recent literature report of manganese modified cobalt oxide
catalyst prepared through solid-state grinding.3
9
The acid-base properties of the catalyst are measured by NH
3
-
TPD and CO -TPD techniques, and the results are presented in
2
oxidation of HMF to FDCA.17 Fig. 8a shows the H
2
-TPR pattern
Fig. S6 & S7. Both techniques reveal that these catalysts contain
both acidic and basic sites, as well as the same acid and base
type. The total acidity and basicity of the catalysts were measured
from peak areas of the TPD profile and these values are
presented in Table S3. All catalysts contain both weak and strong
acid sites and Co-Mn-0.25 catalyst exhibits low acidity among the
of the highly active Co-Mn-0.25 catalyst calcined at various
temperatures in the range of 300 – 600 C. The TPR profile would
o
infer that the number of reduction peaks decreased from three to
two when calcination temperature was increased from 300 to 600
o
o
C. In addition, the lowest temperature reduction peak at ~300 C
was shifted to a higher temperature and the intensity of this peak
o
others. On the other hand, the CO
2
-TPD profile of MnOx catalyst
was decreased with increasing calcination temperature (> 400 C),
shows sharp CO desorption peak in high temperature region,
2
implying decreased oxygen mobility of the catalysts calcined at a
higher temperature, which would account for reduced catalyst
activity for FDCA production. As such, the above results further
demonstrate that the superb activity of the Co-Mn-0.25 catalyst
whereas Co-Mn catalysts exhibit distribution of basic sites at low
and high temeperature regions, indicating the presence of strong
basic sites on the catalyst surface. It has been known that in
manganese based mixed oxide catalysts the basic sites are
o
calcined at 300 C due to the presence of rich active lattice oxygen
2

4+
3+
mainly O (Lewis type basicity), which would favor improving the
species, redox Mn /Mn species, a high surface area, and the
amorphous nature of the catalyst.
40
catalyst oxidation activity. The surface base concentration of Co-
Mn-0.25 catalyst was estimated to be 0.97 mmol/g, which is the
highest among the other catalysts analyzed. Moreover, the
presence of more basic sites rather than acid sites on Co-Mn-0.25
catalyst, could account for its greater oxidation activity, as
supported by some latest literature reports.4
a
0,41
(d)

(Co,Mn)(Co,Mn)
O
2 4
△ Mn O
3 4
 Mn
O
2 3



(c)
Co-Mn-0.25-P
Mn-Co-0.25
(b)
(a)
Co-Mn-0.75
Co-Mn-0.5
#
100
200
300
400
500
C)
600
700
800
Co-Mn-0.33
Temperature (
o
Co-Mn-0.25
Co-Mn-0.16
MnOx
b
Co-Mn-0.25
Co-Mn-0.25-P

△
△
CoOx
1
0 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85
2 (deg.)
x x
Figure 6. XRD pattern of CoO , MnO and Co-Mn mixed oxide catalysts
(a
(b
100
200
300
400
500
600
700
800
o
Temperature ( C)
2
Figure 8 a. H -TPR pattern of Co-Mn-0.25 catalyst calcined at different
temperature: (a) 300 C, (b) 400 C, (c) 500 C, (d) 600 C. b. H -TPR pattern
2
of Co-Mn-0.25 catalyst prepared by different methods.
As shown previously in Fig. 3, the catalysts prepared by different
methods, solid-sate grinding (Co-Mn-0.25) or liquid-state co-
precipitation (Co-Mn-0.25-P), showed remarkable difference in
catalytic activity for production of FDCA from HMF oxidation. The
Figure 7. The TEM images of Co-Mn-0.25 catalyst; (a) fresh and (b) spent
This article is protected by copyright. All rights reserved.

ChemSusChem
10.1002/cssc.201800989
FULL PAPER
lower catalytic selectivity of the Co-Mn-0.25-P could be mainly
2
2
ascribed to its smaller BET surface area, 81 m /g vs. 154 m /g for
Co-Mn-0.25 (Table S1). It can also be seen from Fig. 8b, the H
O
2
-
Mn4+O2-
OH
O
TPR profile of Co-Mn-0.25-P catalyst, the low temperature
reduction peak shifted slightly to a higher temperature compared
to that of the Co-Mn-0.25 catalyst, and the corresponding intensity
of the peak was weaker, indicating lower lattice oxygen mobility
for the Co-Mn-0.25-P catalyst. The combined effects of smaller
surface area and lower oxygen mobility of Co-Mn-0.25-P catalyst
Co-Mn-O
0
.5 O2
Mn3+
could account for its lower catalytic selectivity.
The spent Co-Mn-0.25 catalyst (after 5th cycle of experiment at
2
H O
O
DFF
O
1
20C for 5 h) was characterized by TEM, XPS, and FT-IR
analyses, and the characterization results are presented in Fig.
b, Fig. S8 & S9. The morphology and average particle size (6.5
O
2
H O
Base
7
nm) measured by TEM did not change even after consecutive
runs. The characterization results divulge that nearly no change
in chemical structure of the catalyst took place even after the 5
Mn4+O2-
HO
th
O
catalytic cycle of the experiment, indicating good chemical
stability of the catalyst.
OH
O
Co-Mn-O
0
.5 O2
2.2.1 The reaction mechanism
3
+
Mn
Based on the above discussion and catalyst characterization
results and the pathway proposed previously for Mn-based
OH
O
FFCA
2
H O
16,28,42,43
O
H
catalysts for oxidation of HMF,
a plausible reaction
O
mechanism for oxidation of HMF to FDCA over Co-Mn-O catalyst
is proposed and shown in Scheme 1. According to this
mechanism, the oxidation of HMF to FDCA proceeds through the
2
O
Base
4
+
3+
3+
Mn /Mn redox cycle. In the first step, Mn is oxidized with O
2
Mn4+O2-
HO
4
+
4+ 2-
4+ 2-
to Mn forming Mn O species, and Mn O could oxidize the
OH
3
+
O
HMF to form DFF intermediate and reduced Mn . The formed
DFF in the first step will undergo base catalysed hydration to form
hemiacetal followed by oxidative dehydrogenation over the Co-
Mn binary oxide catalyst to from FFCA. In the final step, the
aldehyde group of FFCA undergoes the same oxidation process
to the third step and finally generates FDCA. Moreover, lattice
oxygen also can participate in HMF oxidation and it can be proved
OH
HO
O
Co-Mn-O
0.5 O2
Mn3+
OH
by the reaction performed under N
2
atmosphere, HMF is
O
2
H O
O
converted into traces of oxygenated products mainly DFF and
FFCA, indicating the reactivity of lattice oxygen atoms on the Co-
Mn-O catalyst.
O
FDCA
Scheme 1. Plausible reaction mechanism for selective oxidation of HMF to
FDCA over Co-Mn-O catalyst
2
.2.2. Two-step catalytic conversion of carbohydrates into
H
2
O medium with NaHCO
3 2
, 1 MPa O ), an overall FDCA yield of
FDCA and the potential of industrial production of FDCA
from bioresources (starch and cellulosic biomass)
33% with respect to glucose was obtained for 24 h reaction (45%
yield of HMF from glucose in Step 1, and 73% yield of FDCA from
crude HMF in Step 2). Admittedly, the catalysed oxidation of the
crude HMF derived from glucose wasslower than that of pure
HMF, which might be due to water soluble impurities in the crude
HMF solution partially inhibit the access of HMF to the catalytic
active sites.
Though pure HMF is commercially available, it is very expensive
by far. Thus, the synthesis of FDCA directly from biomass-
derived carbohydrates rather than pure HMF is highly desirable.
In this regard, we developed a two-step catalytic process in which
carbohydrate (glucose or fructose or their mixtures) is first
catalytically dehydrated to produce HMF, followed by selective
42
In the real application, the crude HMF can be purified by fractional
distillation to obtain purified HMF, with which the oxidation
reaction can be accelerated and the FDCA yield in the Step 2
process could be increased to a value close to that achieved with
pure HMF (i.e., 95%). Furthermore, for industrial application, the
Step-2 process can also operate in a continuous-flow fixed-bed
reactor packed with the pelletized Co-Mn-0.25 catalyst. Based on
the above experimental results, a multi-step industrial process for
production of PEF (a green substitute for PET) from bioresources
(starch and cellulosic biomass) was proposed and conceptually
designed, as schematically shown in Fig. 9. The process involves
three main steps: (1) enzymatic hydrolysis of starch/cellulosic
biomass to produce glucose in a bio-reactor with immobilized
amylase and gluco-amylase, (2) catalytic dehydration of glucose
oxidation to produce FDCA. Niobium phosphate (NbP) or tin
phosphate (SnP) catalyst was used in the first-step process,
dehydration of carbohydrates (glucose or fructose or their
2
mixtures) to HMF in a biphasic reaction medium, H O/methyl
44,45
isobutyl ketone (MIBK) or H
2
O/tetrahydrofuran (THF).
At the
best operating conditions, the HMF yield of 60%, 45% and 53%
was obtained from fructose (over Amberlyst 36 catalyst), glucose
NbP catalyst) and HFCS-90 (NbP catalyst), respectively.44 After
(
the conversion of carbohydrate to HMF, crude HMF was
recovered by evaporation, followed by water extraction.46 The
crude HMF dissolved in distilled water was then fed into mini-
reactor for the oxidation reaction with the Co-Mn-0.25 catalyst.
o
Under the previously established best reaction conditions (120 C,
This article is protected by copyright. All rights reserved.

ChemSusChem
10.1002/cssc.201800989
FULL PAPER
or fructose to HMF in bi-phasic medium (water-MIBK) in a
continuous-flow fixed-bed reactor packed with NbP catalyst, and
CeO catalyst achieved a good yield of FDCA (91-93.8%), but
2
required a longer reaction time as compared with FDCA yield
(3) selective oxidation of purified HMF to FDCA in another
(95%) with a much shorter reaction time (5 h) with the Co-Mn-0.25
continuous-flow fixed-bed reactor packed with Co-Mn-0.25
catalyst. Preliminary techno-economic analysis of the above
process for production of 42,638 tonnes/year of FDCA from corn
starch was conducted by the authors. The analysis indicates that
the economic feasibility of the process, with a total capital
investment of approx. $27 million (2016 US$), is strongly
dependent on the selling price of FDCA (although the current
selling price is not available due to the unavailability of large
production of FDCA), as shown in Fig. 10. The breakeven selling
price of FDCA is approx. $1,800/t for a profitable production of
FDCA from starch via the proposed process. For instance, at a
selling price of FDCA of $2,000/t, the NPV would be $90 million
with a payback period of about 5 years. All parameters used in
the cost analysis are summarized in Table S2 in the
supplementary info.
catalyst in the present work. The Ru/MnCo catalyst obtained
2
O
4
a similarly high FDCA yield (99.1%) but the expensive nature of
Ru limits their practical application. The spinel mixed oxide
Li
FDCA yield (60.6 and 80%) compared with the present Co-Mn-
.25 catalyst, even though the reactions were performed under
2 3 8
CoMn O and Fe-Zr-O based catalysts exhibited a much lower
0
high temperatures and pressure. Moreover, many catalytic
reactions were performed in organic solvents and expensive ionic
liquids. On the other hand, inexpensive Mn0.75/Fe0.25 catalyst was
not as effective as the Co-Mn-0.25 catalyst in direct oxidation of
HMF to FDCA, obtaining much lower FDCA yield. Although, some
of these catalysts mentioned above showed good catalytic
performance, however, preparation of these catalysts involved
precipitation or co-precipitation technique where
a lot of
environmental waste is generated and needs to be disposed. In
contrast, the Co-Mn-0.25 catalyst, prepared in this work via simple
solid-state grinding - a greener and sustainable preparation
method without the need of any solvent, exhibits high catalytic
activity comparable to the noble metal based catalysts.
Table 3. Comparison of the catalytic activity of Co-Mn-0.25 catalyst with that of
other heterogeneous catalysts reported in the latest literature
Catalyst
Catalyst
quantity
Reaction
temp. ( C)
O
2
or air
Reaction
time (h)
FDCA
yield
(%)
Ref.
o
pressure
(MPa)
1
(mg)
Co-Mn-0.25
50
120
5
95
This
work
16
47
17
48
49
39
38
MnO
Ni-MnO
MnOx-CeO
Ru/MnCo
Li CoMn
Fe-Zr-O
2
100
100
100
200
205
10
100
100
110
120
150
160
90
1
0.8
2
2.4
5.5
2
24
28
12
10
8
91
93.8
91
99.1
80
2
Figure 9 Proposed industrial process for production of PEF (a green substitute
for PET) from bioresources
2
2
O
4
2
3
O
8
24
24
60.6
32
150
125
100
Mn0.75/Fe0.25
50
0.8
3. Conclusions
7
5
2
5
0
5
0
This work demonstrated that inexpensive Co-Mn mixed oxide
catalysts (non-noble metal catalysts) prepared by simple solid-
state grinding - a low-cost and greener catalyst preparation
method, are superior to that prepared by conventional liquid-state
co-precipitation method with respect to their catalytic performance
in selective oxidation of 5-hydroxymethylfurfural (HMF) into 2,5-
furandicarboxylic acid (FDCA) in water solvent. Some key
conclusions drawn from this study are summarized below:
1
500
1600
1700
1800
1900
2000
2100
-
-
-
25
50
75
-
-
-
-
-
100
125
150
175
200
(1)
The catalytic activity and selectivity of the Co-Mn
mixed oxide catalysts prepared by solid-state grinding strongly
depended on the Co/Mn molar ratio and calcination temperature
of the catalysts, reaction temperature and time, and the type and
addition amount of a base compound.
Selling price of FDCA (US$/t)
Figure 10. Net present value of the plant versus FDCA selling price
(2)
Co-Mn mixed oxide catalyst with Co/Mn molar
ratio of 0.25 calcined at 300C (Co-Mn-0.25) was found to be the
most active catalyst for FDCA production. The high catalytic
activity of Co-Mn-0.25 catalyst can be attributed to its high lattice
2
.2.3 Performance of Co-Mn-0.25 catalyst compared with
some other catalyts reported in the literature
4+
3+
oxygen mobility and the presence of surface Mn /Mn species.
o
(3)
At the best reaction conditions: 120 C, 5 h and
The catalytic activity of Co-Mn-0.25 catalyst for selective oxidation
of 5-HMF to FDCA in a water solvent is compared with that of
other heterogeneous oxidation catalysts reported in the literature
1
MPa O , 99% HMF conversion, and 95% FDCA yield were
2
obtained, and the catalyst could be reused for five times with
insignificant decrease in activity.
2 2
as shown in Table 3. The activated MnO , Ni -MnO and MnOx-
This article is protected by copyright. All rights reserved.

ChemSusChem
10.1002/cssc.201800989
FULL PAPER
(
4)
Production of FDCA directly from biomass
as 285 eV. The Kratos charge neutralizer system was used on all
specimens. The hydrogen temperature programmed reduction (H –TPR)
2
derived carbohydrates (glucose) rather than pure HMF has been
demonstrated in this work via a 2-step conversion process.
was carried out on a ChemBET Pulsar TPR/TPD automated chemisorption
analyser. In a typical measurement, about 0.1 g of the sample was pre-
(5)
A multi-step industrial process for production of
o
treated at 300 C for 1 h under a flow of helium (99.9%, 120 mL/min) and
PEF (a green substitute for PET) from bio-resources (starch and
cellulosic biomass) was proposed and conceptually designed.
The preliminary techno-economic analysis of the industrial
process designed for production of 60,000 tonnes/year of FDCA
from corn starch demonstrated economic feasibility of the
process, provided that the selling price of FDCA would exceed
o
cooled down to 50 C in the same atmosphere. The treated sample was
o
o
o
reduced from 50 C to 800 C at a linear heating rate of 10 C/min in 5%
/N with a flow rate of 120 mL/min. TEM images were acquired using a
H
2
2
JEOL JEM-2100F field emission transmission electron microscope
equipped with an ultra-high resolution pole-piece operating at 200 kV.
Total acidity and basicity of the catalysts was measured by NH
3
-TPD and
1
800 US$/t.
CO -TPD. TPD measurement was carried out on a ChemBET Pulsar TPR/
2
TPD automated chemisorption analyzer. In a typical experiment, about 0.1
o
Experimental Section
g of the sample was pre-treated at 300 C for 1 h under a flow of helium
-1
(
99.9%, 120 mL min ). Afer pretreatment, the sample was saturated with
at 50 C for 10 min and subsequently flushed
with He at 100 C for 30 min. Then, TPD analysis was carried out from
ambient temperature to 800 C with a heating rate of 10 C min and the
desorbed ammonia or CO was measured by thermal conductivity detector.
o
anhydrous ammonia or CO
2
Chemicals
o
3 2 2 2 2
Reagent grade Mn(NO ) xH O (98%), Co(OAC) 4H O (>98%), D-
o
o
-1
glucose (99.9%), 5- hydroxymethyl furfural (99%), oxalic acid, 5-
hydroxymethyl-2-furancarboxylic acid, 2,5-formylfurancarboxylic acid, 2,5-
furandicarboxylic acid (99%), 2,5-diformylfuran, trimethylamine (>99%),
2
3 2 3 3 2 3
NaHCO , Na CO , KHCO , NaOH, KOH and K CO were purchased from
Selective oxidation of HMF and product analysis
Sigma-Aldrich. Water (HPLC grade) and acetonitrile (HPLC grade) were
obtained from Caledon Laboratories. All reagents were used as received
without further purification.
The oxidation of HMF was carried out in a custom-made 25 mL mini-
reactor assembled with the standard Swagelock fittings and an SS 316L
tube. In a typical catalytic experiment, HMF (0.25 mmol), water (5 mL), the
prepared catalyst (50 mg), NaHCO
reactor and the reactor was purged and pressurized with O
3
(0.5 mmol) were added into the
(1 MPa). The
Catalyst preparation
2
reactor was left in an oil bath on a hot plate stirrer for reaction under
vigorous stirring (500 rpm) at 120 C for 5 h, unless stated otherwise. After
The Co-Mn binary mixed oxide catalysts were prepared via a solid-state
grinding method. In a typical preparation operation, the mixture of required
quantity of Mn(NO ) xH O, Co(OAC) H O and oxalic acid (OA) with a
3 2 2 2 2
o
the reaction, the reactor was quenched to room temperature in a cold-
water bath, and the reaction mixture was diluted three times with
water/acetonitrile (50/50). The catalyst particles were separated out by
centrifugation and the recovered catalyst was washed with water and
specific molar ratio (e.g., Co/Mn = 0.25 and OA/(Co+Mn) = 0.75) was
ground manually in a mortar and pestle for about 10 to 15 minutes at room
temperature. It should be noted that, the purpose of OA addition was to
improve the textural properties such as BET surface area and total pore
volume of the catalyst. The presence of OA could reduce the crystalline
grain size and promote the development of pore structure of the mixed
oxides during the solid-state grinding and the subsequent
o
ethanol, and then dried at 90 C for reuse.
The product analysis was performed using high performance liquid
chromatography (HPLC) (Waters 2690 HPLC) equipped with a photodiode
array detector (waters 2996) using an XBridge Amide analytical column
(3.5 µm, 4.6 × 250 mm), 80/20 acetonitrile/water with 0.2% triethylamine
as mobile phase with a flow rate of 0.6 mL/min. The column temperature
24
drying/calcination processes . The resulting pink coloured solid paste was
o
subjected to drying in an oven at 90 C overnight, followed by calcination
o
o
in air at 300, 400, 500 and 600 C, respectively for 3 h at 2 C/min. The
obtained mixed oxide catalyst samples were denoted as Co-Mn-0.25-300,
Co-Mn-0.25-400, Co-Mn-0.25-500, and Co-Mn-0.25-600, respectively.
More Co-Mn mixed oxide catalysts with a varying Co/Mn molar ratio were
prepared by following the same procedure as described above, but were
o
was maintained at 35 C. The contents of HMF, FDCA, 2,5-
formylfurancarboxylic acid (FFCA), 2,5-diformylfuran (DFF) and 2,5-
hydroxymethylfurancarboxylic acid (HFCA) in the products were quantified
by the external calibration curve method, which were constructed using the
pure compounds. Maximum absorption was found for HMF (λ = 281 nm),
FDCA (λ = 260 nm), FFCA (λ = 287 nm), HFCA (λ = 250 nm) and DFF (λ
o
o
calcined at 300 C for 3h @ 2 C/min, which were denoted as Co-Mn-0.16
Co/Mn = 0.16), Co-Mn-0.33 (Co/Mn = 0.33), Co-Mn-0.5 (Co/Mn = 0.5) and
Co-Mn-0.75 (Co/Mn = 0.75). The same procedure was also employed to
prepare mono oxide catalysts, MnO and CoO . For comparison, the
(
=
285 nm), respectively. All experiments were conducted in duplicate or
triplicate, and the HMF conversion, FDCA and FFCA yields, as calculated
by the following equations, are presented as (average value ± relative
error).
x
x
catalyst with the same composition as the Co-Mn-0.25 catalyst was
prepared by the conventional co-precipitation method as reported
25
elsewhere, and the resulting catalyst was denoted as Co-Mn-0.25-P.
ꢀ
ꢁꢂꢃꢄ ꢁꢅ ꢆꢀꢇ ꢈꢃꢉꢊꢋꢃꢌ
HMF conversion (%) =
× 100
(1)
(2)
(3)
ꢍꢎꢏꢋꢏꢉꢂ ꢐꢁꢂꢃꢄ ꢁꢅ ꢆꢀꢇ
Catalyst characterization
ꢀ
ꢁꢂꢃꢄ ꢁꢅ ꢇꢑꢒꢓ ꢔꢈꢁꢌꢕꢊꢃꢌ
ꢍꢎꢏꢋꢏꢉꢂ ꢐꢁꢂꢃꢄ ꢁꢅ ꢆꢀꢇ
FDCA yield (%) =
FFCA yield (%) =
× 100
Powder X-ray diffraction (XRD) patterns of all fresh catalysts were
measured using an X-ray diffractometer (Rigaku RINT 2500, Tokyo, Japan)
with Ni-filtered Cu Kα radiation (λ = 1.5406 Å) with a scan speed of 2° min−
ꢀ
ꢁꢂꢃꢄ ꢁꢅ ꢇꢇꢒꢓ ꢔꢈꢁꢌꢕꢊꢃꢌ
× 100
1
ꢍꢎꢏꢋꢏꢉꢂ ꢐꢁꢂꢃꢄ ꢁꢅ ꢆꢀꢇ
and a scan range of 2–80° at 30 kV and 15 mA. N
performed using Quadrasorb SI automated surface area and pore size
analyzer. Prior to the N adsorption measurement, the sample was
2
physisorption was
2
Acknowledgements
o
-3
degassed at 200 C until a stable static vacuum of less than 510 Torr
was achieved. Subsequently, N
adsorption was performed at 196 oC.
2
The authors acknowledge the financial support from NSERC
Discovery Grants and the funding from BioFuelNet Canada, a
Network of Centres of Excellence. One of the authors (C.X.) also
gratefully acknowledges the funding from the NSERC/FPInnovations
Industrial Research Chair program and the ORF-RE Program in
Forest Biorefinery. One of the authors (L.D.) also gratefully
acknowledges Brazil Science Without Borders/CNPq Graduate
Scholarship.
The surface area of the catalysts was measured using multi-point BET in
the relative pressure range of 0.05 – 0.3. The total pore volume was
estimated at the relative pressure of 0.99. Fourier transform infrared (FT-
IR) spectra of the catalysts were recorded on
a PerkinElmer IR
spectrometer using the KBr disc method. X-ray photoelectron
spectroscopy (XPS) measurements were conducted on an Axis Nova
spectrometer using a monochromatic Al K (alpha) source (15 mA, 14 kV),
calibrated using Au as the standard. For energy calibration, the carbon 1s
photoelectron lines were used. The carbon 1s binding energy was taken
This article is protected by copyright. All rights reserved.

ChemSusChem
10.1002/cssc.201800989
FULL PAPER
Qi, A. Riisager, Chem Select. 2017, 2, 6632-6639.
29] D. Meng, Q. Xu, Y. Jiao, Y. Guo, Y. Guo, L. Wang, G. Lu
Keywords: Selective oxidation •5-hydroxymethylfurfural • 2,5-
Furandicarboxylic acid • Co-Mn mixed oxide catalysts • Solid-
state grinding
[
[
W. Zhan, Apply. Catal. B: Environ. 2018, 221, 652-
663.
30] Y. G. Ji, Z. Zhao, A. J. Duan, G. Y. Jiang, J. Liu, J.
Phys. Chem. C. 2009, 113, 7186-7199.
[
[
1] G. W. Huber, J. N. Chheda, C. J. Barrett, J. A. Dumesic,
Science 2005, 30, 1446-1450.
2] B. Sha, M. M. Abu-Omar, Green Chem. 2014, 16, 24-38.
[31] S. Todorova, H. Kolev, J.P. Holgado, G. Kadinov, Ch.
Bonev, R. Pere˜niguez, A. Caballero, Appl. Catal. B:
Environ. 2010, 94, 46-54.
[
3] Y. Zhang, Z. Yuan, C. Xu, AIChE Journal, 2015, 61,
[32] C. Shu, Y. Wang, A. Zhu, B. Chen, C. Au, Catal.
Commun. 2012, 28, 18-22.
1275-1283.
[
[
4] Z. Zhang, K. Deng, ACS Catal. 2015, 5, 6529-6544.
5] W. Partenheimer, V. V. Grushin, Adv. Syn. Catal. 2001,
[33] Y. S. Wu, Y. Lu, C. J. Song, Z. C. Ma, S. T. Xing and Y. Z.
Gao, Catal. Today 2013, 201, 32-39.
3
43, 102-111.
[34] (a) Z.-Y. Tian, P. H. T. Ngamou, V. Vannier, K. Khose-
Hoinghaus, N. Bahlawane, Appl. Catal. B: Environ. 2012,
117-118, 125-134;(b) Z. Shu, Y. Chen, W. Huang, X. Cui,
L. Zhang, H. Chen, G. Zhang, X. Fan, Y. Wang, G. Tao,
D. He, J. Shi, Appl. Catal. B: Environ. 2013, 140–141, 42-
50; (c) A. Jha, K. R. Patil, C. V. Rode, ChemPlusChem.
2013, 78, 1384-1392.
[35] K. Cheng, F. Yang, G. Wang, J. Yin, D. Cao, J. Mater.
Chem. A 2013, 1, 1669-1676.
[36] H.F.Li, G. Z. Lu, Q. G. Dai, Y.Wang, Y. Guo, Y. Guo,
Appl. Catal., B: Environ. 2011, 102, 475-483.
[
[
[
[
[
6] N. K.Gupta, S. Nishimura, A. Takagaki, K.Ebitani, Green
Chem. 2011, 13, 824-827.
7] A.A. Rosatella, S. P. Simeonov, R. F. M. Frade and C. A.
M. Afonso, Green Chem. 2011, 42, 754-793.
8] H. A. Rass, N. Essayem, M. Besson, ChemSusChem.
2015, 8, 1206-1217.
9] X. Wan, C. Zhou, J. Chen, W. Deng, Q. Zhang, Y. Yang and
Y. Wang, ACS Catal. 2014, 4(7), 2175-2185.
10] G.Yi, S. P. Teong, Y. Zhang, Green Chem. 2016, 18,
9
79-983.
[
11] J. Deng, H. J. Song, M. S. Cui, Y. P. Du and Y. Fu,
[37] (a) H. Perez, P. Navarro, J. J. Delgado, M. Montes, Appl
Catal A: Gen 2011, 400, 238-248; (b) Y. Zhang, Z. Qin, G.
Wang, H. Zhu, M. Dong, S. Li, Z. Wu, Z. Li, Z. Wu, J. Zhang,
T. Hu, W. Fan, J. Wang Appl Catal B 2013, 129, 172-181.
[38] (a) M. Wu, W. Zhan, Y. Guo, Y. Guo, Y. Wang, L. Wang, G.
Lu, Appl. Catal. A:Gen, 2016, 523, 97-106; (b) N. J.
Tharayil, R. Raveendran, A. V. Vaidyan, Indian J. Pure
Appl. Phys. 2008, 46, 47-53.
ChemsusChem. 2014, 7, 3334-3340.
12] S. G. Wang, Z. H. Zhang, B. Liu, ACS Sustainable
Chem. Eng. 2015, 3, 406-412.
13] L. C. Gao, K. J. Deng, J. D. Zheng, B. Liu and Z. H. Zhang,
Chem. Eng. J. 2015, 270, 444-449.
14] B. Saha, D. Gupta, M. M. Abu-Omar, A. Modak, A.
Bhaumik, J. Catal. 2013, 299, 316-320.
15] D. Yana, J. Xin, Q. Zhao, K. Gao, X. Lu, G.Wang, S.
Zhang, Catal. Sci. Technol. 2018, 8, 164-175.
16] E. Hayashi, T. Komanoya, K. Kamata, M. Hara,
ChemSusChem. 2017, 10, 654-658.
[
[
[
[
[
[39] C. Liu, L. Gong, M. Lu, T. Sun, R. Dai, Q. Liu, X. Huang,
G. Liu, Z. Huang, Catal Surv Asia 2017, DOI:
0.1007/s10563-017-9235-0
[
40] F. Neatu, R. S. Marin, M. Florea, N. Petrea, O. D. Pavel, V.
I. Parvulescu, Appl. Catal. B: Env. 2016, 180, 751-757.
41] D.Yana, J. Xin, Q. Zhaob, K. Gaob, X. Lub, G. Wang, S.
Zhang, Catal. Sci. Technol. 2018, 8, 164-175.
[17] X. Han, C. Q.Li, X. H. Liu, Q. N. Xia, Y. Q. Wang, Green
Chem. 2017, 19, 996-1004.
18] Q. Zhang, X. Liu, W. Fan, Y. Wang, Appl. Catal B:
Environ. 2011, 102, 207-214.
19] W. X. Tang, X. Wu, D. Li, Z.Wang, G. Liu, H. Liu, Y.
Chen, J. Mater. Chem. A 2014, 2, 2544-2554.
[
[
[
[
[
42] J. Nie, H. Liu, J. Catal. 2014, 316, 57-66.
43] S. Zhang, X. Sun, Z. Zheng, L. Zhang, Catal. Commun.
2018, 113, 19-22.
[20] V. P.Santos, M. F. R. Pereira, J. J. M. Orfao and J. L.
[
44] S. Souzanchi, "Catalytic Conversion of Fructose, Glucose
and Industrial Grade Sugar Syrups to 5-
Figueiredo, Appl. Catal. B Environ. 2010, 99, 353-363.
21] (a) C. Marchal, A. Piquet, M. Behr, T. Cottineau, V.
Papaefthimiou, V. Keller, V. Caps, J. Catal. 2017, 352, 22-
[
Hydroxymethylfurfural, A Platform for Fuels and
Chemicals" 2016. Electronic Thesis and Dissertation
Repository. 4070. https://ir.lib.uwo.ca/etd/4070.
45] K. T. V. Rao, S. Souzanchi, Z. Yuan, M. B. Ray and C. Xu,
RSC Adv. 2017, 7, 48501- 48511.
46] S. Yi, S. P. Teong, X. K. Li and Y. G. Zhang,
ChemSusChem 2014, 7, 2131-2135.
47] K. Yu, Y. Liu, D. Lei, Y. Jiang, Y. Wang,Y. Feng, L. Lou, S.
Liu, W. Zhou, Catal. Sci. Technol. 2018, DOI:
34; (b) C. Marchal,M. Behr, F. Vigneron, V. Caps, V. Keller,
New J. Chem. 2016, 40, 4428-4435.
[
[
[
[
22] S. A. Kondrat, G.Shaw, S. J. Freakley, Q. He, J. Hampton,
J. K. Edwards, P. J. Miedziak, T. E. Davies, A. F. Carley,
S. H. Taylor, C. J. Kiely, G. J. Hutchings, J. Chem. Sci.
2012, 3, 2965-2971.
[
23] (a) H. Zhu, D. C. Rosenfeld, D. H. Anjum, V. Caps, J.-M.
Basset, ChemSusChem 2015, 8, 254-1263; (b) H. Zhu, D.
C. Rosenfeld, M. Harb, D. H. Anjum, M. N. Hedhili, S.
Ould-Chikh, J.-M. Basset, ACS Catal. 2016, 6, 2852-2866.
24] S. Albonetti, A. Lolli, V. Morandi, A. Migliori, C. Lucarelli,
F. Cavani, Appl. Catal.B: Environ. 2015, 163, 520-530.
25] H. A. Rass, N. Essayem, M. Besson, Green Chem. 2013,
1
0.1039/C7CY02455J.
48] D. Kumar Mishra, H. J. Lee, J. Kim, H. Lee, J. K. Cho, Y. Suh,
Y. Yid, Y. J. Kim, Green. Chem. 2017, 19, 1619-1623.
[
[
[
[49] A. Jain, S. C. Jonnalagadda, K. V. Ramanujachary, A.
Mugweru, Catal. Commun. 2015, 58, 179-182.
15, 2240-2251.
[
26] S. Biswas, B. Dutta, A. Mannodi-Kanakkithodi, R. Clarke,
W. Song, R. Ramprasad, S. L. Suib, Chem. Commun.
2017, 53, 11751-11754.
[
[
27] J. H. Zhang, Y. B. Li, L. Wang, C. B. Zhang , H. He,
Catal. Sci. Technol. 2015, 5, 2305–2313.
28] Z. Gui, S. Saravanamurugan, W. Cao, L. Schill, L. Chen, Z.
This article is protected by copyright. All rights reserved.

ChemSusChem
10.1002/cssc.201800989
FULL PAPER
Entry for the Table of Contents (Please choose one layout)
Layout 1:
FULL PAPER
Text for Table of Contents
K. T. V.Rao, J. L. Rogers, S. Souzanchi,
L. Dessbesell, M. Ray, C. Xu *
The Co-Mn mixed oxide
catalysts are prepared through
simple and green solid-state
Page No. – Page No.
Inexpensive but highly efficient Co-Mn
mixed oxide catalysts prepared by simple
solid-state mixing for selective oxidation
of 5-hydroxymethylfurfural to 2,5-
furandicarboxylic acid
grinding
method
which
exhibited excellent catalytic
performance for selective
oxidation HMF to FDCA.
This article is protected by copyright. All rights reserved.