Aqueous-Phase Oxidation of Furfural to Maleic Acid Catalyzed by Copper Phosphate Catalysts

Article abstract of DOI:10.1007/s10562-017-2191-5

Abstract: This work describes catalytic aqueous-phase oxidation of furfural to maleic acid. The heterogeneous Cu-phosphate catalysts were prepared by coprecipitation method at different atomic ratios of precursors and characterized by various techniques.

Full text of DOI:10.1007/s10562-017-2191-5

Catal Lett
DOI 10.1007/s10562-017-2191-5
Aqueous-Phase Oxidation of Furfural to Maleic Acid Catalyzed
by Copper Phosphate Catalysts
1
1
2
3
Tomáš Soták · Milan Hronec · Miroslav Gál · Edmund Dobročka ·
4
Jaroslava Škriniarová
Received: 19 July 2017 / Accepted: 13 September 2017
©
Springer Science+Business Media, LLC 2017
Abstract This work describes catalytic aqueous-phase
oxidation of furfural to maleic acid. The heterogeneous Cu-
phosphate catalysts were prepared by coprecipitation method
at different atomic ratios of precursors and characterized by
various techniques. The CaCu-phosphate catalyst showed
the best catalytic performance. Using this catalyst under
Graphical Abstract
reaction conditions (115 °C and 0.8 MPa of O ) 37.3 mol%
2
yield and 54.8% selectivity of maleic acid were achieved.
This catalyst can be recycled more than four times producing
practically the same selectivity as fresh one by 13% decrease
in furfural conversion. The proposed mechanism involves
furoic acid as an intermediate in the formation of maleic
acid.
Keywords Furfural · Oxidation · Metal pyrophosphates ·
Maleic acid · Furoic acid
1
Introduction
The major challenge today is to convert biomass as feedstock
into useful chemicals for the chemical and pharmaceutical
industries in an economically viable fashion. Furfural is a
renewable resource coming from agricultural byproducts,
hemicelluloses, by dehydration in acidic media. Currently
it is used as a solvent and as a platform compound for syn-
thesis of variety of chemicals and engine fuels [1, 2]. Selec-
tive oxidation of biomass-based furfural to maleic acid or
anhydride which is currently manufactured from benzene
and/or butane is an alternative route to the on-going fossil
based processes. Maleic acid is applied in the synthesis of
plasticizers, copolymers, resins, surface coatings, lubricants
and agricultural chemicals.
Electronic Supplementary Material The online version
of this article (doi:10.1007/s10562-017-2191-5) contains
supplementary material, which is available to authorized users.
*
Tomáš Soták
tomas.sotak@stuba.sk
*
Milan Hronec
milan.hronec@stuba.sk
1
Department of Organic Technology, Catalysis and Petroleum
Chemistry, Slovak University of Technology, Radlinského 9,
In literature is described oxidation of furfural in both
vapor and liquid phases using various types of metal cata-
lysts [3, 4]. In spite of oxidation of furfural in vapor phase
where the main products are maleic anhydride and carbon
dioxide, the oxidation in the liquid phase, depending on the
catalyst and solvent, can produce 2-furancarboxylic acid
8
12 37 Bratislava, Slovakia
2
3
4
Department of Inorganic Technology, Slovak University
of Technology, Radlinského 9, 812 37 Bratislava, Slovakia
Institute of Electrical Engineering, Slovak Academy
of Sciences, Dúbravská cesta 9, 845 41 Bratislava, Slovakia
Institute of Electronics and Fotonics, Slovak University
of Technology, Ilkovičova 3, 812 19 Bratislava, Slovakia
(
furoic acid) eventually its esters or furan-ring opening
Vol.:(0
1
123456789)
3

T. Soták et al.
product maleic acid. Generally, the main drawback of cata-
lyzed aerobic oxidations of organic substrates is low selec-
tivity. Traditionally, oxidations conducted in the liquid-phase
have been catalyzed by dissolved salts of redox-active met-
als. Several metallic catalysts were studied for the oxida-
tion of furfural or its derivatives, but usually with limited
selectivity for desired oxidized products [4, 5]. The major
problem of low selectivity is mostly connected with polym-
erization of furfural to resins under oxidative conditions.
Yin et al. [6] using combination of copper nitrate with
phosphomolybdic acid obtained 49.2% yield of maleic acid
with 51.7% selectivity. Kinetics of oxidation of furfural to
Furfural was purified by vacuum distillation and stored at
−15 °C.
2.2 Catalyst Preparation and Characterization
The copper–phosphate catalysts were prepared by precipi-
tation of the copper nitrate solution with an ammonium
monohydrate orthophosphate solution at different molar
ratios of Cu/P precursors. The hydrogen phosphate solu-
tion was added dropwise to the metal nitrate solution while
stirring vigorously at ambient temperature. The precipitate
was stirred for 3 h, then filtered and excessively washed
with deionized water to extract the residues of the precur-
sors. The mixed copper–metal (Mg, Ca, Sr, Ba)–phosphate
catalysts were prepared accordingly, mixing corresponding
metal-copper nitrate solutions at different atomic ratios of
M/Cu/P precursors. All the catalysts prior to catalytic tests
were calcined at 600 °C in air flow for 5 h, with a heating
2
-furancarboxylic acid using alum-impregnated activated
alumina studied Kumar et al. [7]. The more challenging
results were achieved recently using heterogeneous Au/ZrO
2
catalyst; however, in the oxidative esterification of furfural
with CH OH [8]. This catalyst is very active and selective
3
to methylfuroate (after 90 min both conversion and selec-
tivity are complete) and no acetal formation was observed.
Metal-free oxidative synthesis of succinic acid from furfural
and its homologues in the presence of hydrogen peroxide
described Erbitani et al. [9, 10]. Maximum succinic acid
yield and H O utilization efficiency of 74 and 85%, respec-
−
1
rate of 10 °C min . The Cu/P molar ratios of prepared Cu-
phosphate catalysts are summarized in Table S1 (see Sup-
plementary Data).
2.3 Apparatus and Oxidation Tests
2
2
tively, were achieved using Amberlyst-15 as a reusable solid
acid catalyst. Du et al. [11] studied ring-opening oxidation
of 4-hydroxymethyl furfural with molecular oxygen in ace-
Oxidation reactions were carried out in 10 mL stainless steel
autoclaves lined with glass and equipped with valve. The
autoclave was charged with furfural, water and catalyst and
after closing pressurized with oxygen (usually 0.8 MPa). The
reactors were heated in oil bath and the reaction mixture was
magnetically stirred. After reaction, the solid part containing
the catalyst was separated from the reaction mixture by cen-
trifugation. In the catalyst recycle study the reaction mixture
was centrifuged at 4500 rpm for 15 min and after centrifug-
ing the catalyst was located on the bottom of the vessel as
a solid phase. Then very carefully was removed about 95%
of the supernatant liquid and to the settled down solid phase
was added ca. 5 mL of deionized water. After mixing (ca.
5 min) and subsequent centrifugation the supernatant liquid
was again removed. This procedure was repeated twice to
ensure that all water soluble reaction products were recov-
ered from the catalyst (Table 1).
tonitrile. In the presence of homogeneous VO(acac) catalyst
2
maleic acid in about 52% yield was obtained. Li et al. [12]
studied binary Mo–V metal oxides as catalysts for the selec-
tive oxidation of furfural to the mixture of maleic acid and
maleic anhydride in acetic acid as a solvent. Under optimal
conditions over the Mo VO catalyst up to 65% yield of
4
14
maleic anhydride and maleic acid (in the ratio 1.8:1) was
achieved. The recent development on maleic acid synthe-
sis from biomass-derived chemicals over homogeneous or
heterogeneous catalysts was presented lately in the review
paper [13].
In the present contribution, we have investigated the oxi-
dation of furfural to maleic acid in water as a solvent. The
main objective was to study the influence of various hetero-
geneous copper-phosphate catalysts and reaction conditions
on the activity and selectivity of the formation of maleic
acid. Recycling experiments were conducted to check the
stability of selected heterogeneous catalysts during oxida-
tion reaction.
2.4 Analysis
The liquid samples were analyzed by high performance liq-
uid chromatography equipped with UV and RID detectors,
−
1
using a HPX-87H column and 5 mmol L sulfuric acid
−
1
2
Experimental
aqueous solution as the mobile phase with 0.65 mL min
flow rate. In selected samples the contents of maleic acid
was checked also by gravimetric analysis of the washed and
dried solid formed by precipitation with aqueous solution
of barium hydroxide. In some samples was determined by
gas chromatography (Shimadzu GC 2014 with TCD) the
2
.1 Materials
Furfural (98%), maleic acid (99%), furoic acid (98%),
and other chemicals were purchased from Sigma-Aldrich.
1
3

Aqueous-Phase Oxidation of Furfural to Maleic Acid Catalyzed by Copper Phosphate Catalysts
Table 1 Acidity of catalysts prepared at different molar ratios of pre-
Table 2 Electrode potentials and Cu(I) peak areas of monometallic
Cu P O and bimetallic M-Cu-phosphates prepared at different molar
cursors
2
2
7
ratios
Catalyst
Cu P O
Cu/P ratio Ca/Cu ratio Ca:Cu:P ratio a (mmol
T
−1
−9
g
)
M:Cu:P
Catalyst
Cu P O
Electrode potential Peak area×10
CAT
molar ratio
versus RE*/V
(Coulomb)
0.5
–
–
–
–
0.090
0.080
0.075
0.065
0.090
0.050
0.030
0.025
2
2
7
1
1
1
1
1
0
1
1
:1
−0.042± 0.021
−0.027± 0.018
−0.038± 0.012
−0.055± 0.014
−0.068± 0.008
−0.032± 0.012
−0.038± 0.012
No oxidation peak
0.5± 0.1
8.0± 1.1
14.7± 2.3
5.2± 0.8
1.1± 0.3
6.4± 1.2
14.7± 2.3
1
1
2
.0
.5
.0
2
2
7
:1:2
:1:2
:1:2
:1:2
.5:1:2
:1:2
.5:1:2
MgCuP O
2 7
–
–
–
–
–
–
CaCuP O
2
7
SrCuP O
7
CaCuP O7 0.5
2
2
BaCuP O
0
0
0
.5
.5
.5
0.5
1.0
1.5
0.5:1:2
1:1:2
1.5:1:2
2
7
7
7
7
CaCuP O
2
CaCuP O
2
CaCuP O
2
a Total acidity (mmol g−1CAT)
T
*
RE Reference electrode (saturated calomel electrode)
concentration of carbon oxides in the gas phase. The identi-
fication of main products was done by NMR spectroscopy.
The product yield was calculated by
rate of 3 mV/s were performed. The obtained results are
summarized in Table 2.
Y = [(productdetected∕substrateinput) × 100]
3
Results
2
.5 Catalyst Characterization
3
.1 XRD Characterization
The BET surface area of the samples was determined by
nitrogen adsorption at 77 K after activation of the sample in
vacuum at 300 °C for 2 h. Powder X-ray diffraction (XRD)
patterns were acquired on a Bruker AX S D8 diffractometer
using CuKα radiation. Crystalline phases were identified
by a comparison with the PDF file. The morphology of the
metal phosphate catalysts was followed by SEM analysis
using a JEOL JSM-7500F microscope. Temperature pro-
grammed reduction was carried out using a Micromeritics
Pulse Chemisorb 2700 apparatus. The acidity of catalysts
was measured by amine titration method. The sample of
solid catalyst (0.2 g) was suspended in 25 mL of water using
a magnetic stirrer and agitated for 30 min. The suspension
was titrated with 0.1 M solution of ethylamine dissolved in
water. During the titration procedure, the pH value of the
solution was continuously measured. On the base of eth-
ylamine consumption was calculated the acidity in terms
Copper phosphate formed after precipitation of copper
nitrate solution with an ammonium monohydrate orthophos-
phate solution is known to transform to copper pyrophos-
phate, Cu P O at the temperature of calcination of 600 °C
2
2
7
[14]. The XRD patterns of calcined samples of copper
phosphates prepared by precipitation of copper nitrate solu-
tion at the atomic ratios of Cu:P=0.5 and 1.25 confirmed
(Fig. 1) that both samples were well crystallized. The sam-
ple prepared at the ratio Cu:P = 0.5 contains a very pure
Cu P O phase (PDF No. 00-044-0182) with a particle
2
2
7
size 23.2± 0.2 nm, but in the sample prepared at the ratio
Cu:P=1.25 a very small amount of Cu(HPO )(H O) (PDF
4
2
No. 01-083-1857) and CuHPO (H O) (PDF No. 01-072-
3
2
2
1367) phases was present. The presence of the peak patterns
corresponding to the precursor Cu (PO ) ·3H O (Fig. 1a,
3
4 2
2
PDF No. 00-022-0548) was not registered in these samples.
The parent Ca–Cu phosphate represents a mixed phase
system. As being a co-precipitate, one can expect various
separated phosphates of copper and calcium, and bime-
tallic phosphates. The calcination temperature of the pre-
cipitate should favor the preference of pyrophosphates over
orthophosphates or hydrogen phosphates in the resulting
material [15, 16]. In the binary Ca–Cu phosphate sample
prepared at the atomic ratios of Ca/Cu/P precursors 1:1:2
monometallic copper pyrophosphate Cu P O and cal-
−
1
of mmol g catalyst. Abrasive linear sweep voltammetry
AbLSV) using a carbon paste electrode (CPE) as a working
(
electrode and a platinum mesh as a counter electrode was
performed. The solid catalyst samples were mechanically
deposited over the working electrode surface. The voltam-
metric patterns curves were recorded using PGSTAT 302N
(
Metrohm, Switzerland) in an argon-purged phosphate buffer
pH 7.4). Saturated calomel electrode (SCE) was served as
(
a reference electrode (RE). The measurements were carried
out at 25± 1 °C for each catalyst and were repeated at least
three times. Freshly cut and polished electrode surface was
used; and the measurements in the potential range from
2
2
7
cium phosphate Ca (PO ) (PDF No. 00-029-0359) were
3
4 2
found as dominant phases in the diffraction region below
37° (Fig. 2). In this sample are present peak patterns cor-
responding to the CaCuP O (PDF No. 01-080-0706) and
−
0.4 V versus RE to 0.4 V versus RE with potential sweep
2
7
1
3

T. Soták et al.
sample was not confirmed. However, the small shift of the
crystallographic parameters indicates that a partial mutual
substitution of Cu and Ca occurred in these phases, leading
to a deformation of the crystalline structures. This cation
exchange takes place probably during the calcination. A very
complex XRD pattern of sample of Ca–Cu phosphate shows,
that in the mixed Ca–Cu phosphate various purely defined
structures can be formed [17].
3
.2 SEM Characterization
The morphology of the monometallic Cu P O catalysts
2
2
7
can be significantly altered by changing the method of their
preparation. A precipitation of copper nitrate by various
amounts of ammonium hydrogen phosphate followed by
calcination at 600 °C resulted in different structures with
relatively large crystals, or in poorer crystalline. Figure 3
shows typical SEM images of calcined samples synthesized
at various conditions. The samples prepared at the atomic
ratios of Cu:P=0.5 and 1.0 possessed the agglomeration of
micron sized irregular particles with rather crystalline struc-
ture. The size of crystals was above 1 µm (Fig. 3a, b, e, f).
However, the sample prepared at higher ratio (Cu:P=1.25)
displays SEM images of porous spheres structures with
many nanopores (Fig. 3c, g). This phenomenon can be also
seen in Fig. 3d, h for the sample of mixed Ca–Cu phosphate;
we also find that there are many nanopores and the mor-
phology of the sample has the porous spheres structure. The
specific surface areas of the phosphate catalysts correspond
with their morphology observed by SEM. All samples had
Fig. 1 XRD patterns of Cu P O7samples prepared at molar ratio
2
2
Cu/P a 0.5; b 1.0; c 1.25; d sample(c) after catalytic test; phases of
Cu (PO )·3H O (PDF No. 00-022-0548), Cu P O (PDF No. 00-044-
3
4
2
2
2
7
0
182) and Cu (OH)PO (PDF No. 01-077-0922)
2 4
2
−1
a specific surface area smaller than 10 m g .
3
.3 TPR Characterization
TPR measurements were performed to investigate the
reducibility of both copper-phosphate and mixed Ca–Cu-
phosphate samples prepared at different molar ratios of the
precursors and calcined at 600 °C (Fig. 4). The copper-phos-
phate samples displayed two reduction peaks correspond-
ing to reduction of Cu(II) and Cu(I) species at lower and
higher temperatures. As is seen from Fig. 4a–c with increas-
ing Cu/P ratio during samples preparation both reduction
peaks are shifted to higher temperatures. As the SEM images
shows the shift of reduction peaks can be ascribed to dif-
ferent morphology of these samples. Similar dependence
of reducibility temperature of copper species on the molar
ratios of the precursors was observed also for the mixed
Ca–Cu-phosphate samples (Fig. 4d–f). Moreover, in com-
parison to Cu-phosphate samples in mixed Ca–Cu-phos-
phates the reduction temperatures of both Cu(II) and Cu(I)
species were more than 60 °C higher.
Fig. 2 Typical XRD image of mixed Ca–Cu-phosphate prepared by
calcination of precipitate synthetized at molar ratios Ca:Cu:P=1:1:2:
phases of Cu P O (PDF No. 00-044-0182), CaCuP O (PDF No.
2
2
7
2
7
0
1-080-0706), Ca (PO ) (PDF No. 00-029-0359) and Ca Cu (PO )
3 4 2 3 3 4 4
(
PDF No. 01-072-0203)
Ca Cu (PO ) (PDF No. 01-072-0203) phases. The presence
3
3
4 4
of Cu (PO ) ·3H O phase (PDF No. 00-022-0548) in this
3
4 2
2
1
3

Aqueous-Phase Oxidation of Furfural to Maleic Acid Catalyzed by Copper Phosphate Catalysts
Fig. 4 TPR profiles of Cu P O7and CaCuP O catalysts with
2
2
2
7
different Cu/P and Ca/Cu ratios: (a) Cu P O7 (Cu/P=0.5); (b)
2
2
Cu P O7(Cu/P=1.25); (c) Cu P O7(Cu/P=2); (d) CaCuP O (Ca/
2
2
2
2
2
7
Cu=0.5); (e) CaCuP O (Ca/Cu=1); (f) CaCuP O (Ca/Cu=1.5)
2
7
2
7
−
1
to 0.050 mmol g
and in the sample at the ratio Ca/
CAT
−
1
Cu = 1.5 to 0.025 mmol g
. The drop of acidity with
CAT
increasing Cu/P and Ca/Cu atomic ratios in the samples
was expected since copper and calcium possess properties
of bases.
3
.5 Catalytic Tests
Catalytic oxidations of furfural using prepared copper-phos-
phate type catalysts were carried out at 115 °C and 0.8 MPa
of oxygen pressure in aqueous solution. The major product
formed in oxidation was maleic acid present in an aqueous
phase and carbon dioxide in the gas phase. In most samples
was identified by HPLC in liquid phase also presence of
small amounts of furoic acid.
3
.5.1 Preliminary Study
In our preliminary study of furfural oxidation to maleic acid,
different copper based catalysts, e.g. Cu oxide, nitrate, or
acetylacetonate, were tested. Using these catalysts maleic
acid in the yields lower than 18% was achieved. Using the
mixed Cu–Fe phosphate (CuFeP O ) [18, 19] or monometal-
Fig. 3 SEM images of Cu P O prepared at Cu:P ratios (a, e) 0.5; (b,
2
2
7
f) 1.0; (c, q) 1.25 and (d, h) mixed Ca–Cu-phosphate synthetized at
molar ratios Ca:Cu:P=1:1:2 and their magnified views
2
7
lic Fe phosphate (Fe P O ) catalysts the yield of maleic acid
2
2
7
3
.4 Acidity of Catalysts
17.1 and 15.2% were obtained. However, when monometal-
lic Cu P O catalyst (Cu/P=1.0) was used for the reaction,
2
2
7
Table 1 shows acidity of samples synthesized at various
molar ratios of copper and phosphate precursors and cal-
cined at 600 °C. As is evident from results the total acidity
of Cu P O samples decreases from 0.090 to 0.065 mmol
the yield of furfural to maleic acid was improved to 25.6%.
Using this Cu-phosphate catalyst the effect of reaction tem-
perature (Fig. S1) and the influence of oxygen pressure (Fig.
S2) on the oxidation of furfural to maleic acid was stud-
ied (Supplementary Data). Over this catalyst the highest
yield of maleic acid, 25.7% at 64.5% furfural conversion,
was achieved at the temperature of 115 °C and 0.8 MPa of
oxygen pressure. Based on these results, the research was
2
2
7
−
1
g
by increasing Cu/P atomic ratio from 0.5 to 2.0.
CAT
Stronger decrease in acidity is observed in mixed Ca–Cu-
phosphates. The presence of calcium at the atomic ratio
Ca/Cu = 0.5 decreases the acidity of sample from 0.090
1
3

T. Soták et al.
focused on the preparation and characterization of differ-
ent copper phosphate catalysts and their testing for furfural
oxidation to maleic acid under the above described reaction
conditions.
polymerization of the reaction intermediates to resins and
promoting their conversion to maleic acid.
3.5.3 Oxidation Catalyzed by Mixed M–Cu‑phosphate
catalysts
3
.5.2 Oxidation Catalyzed by Cu P O Catalysts
2 2 7
The improvement of catalytic activity and selectivity to
maleic acid is achieved using mixed metal phosphate
catalysts prepared by coprecipitation of the metal precur-
sors followed by calcination at 600 °C. In Fig. 6 are the
results of furfural oxidation in the presence of copper
catalysts prepared in the presence of Mg, Ca, Sr and Ba
salts in the atomic ratio metal/Cu= 1. Their influence on
both activity and selectivity to maleic acid is in the order:
Ca>Mg>Sr>Ba. Among these bimetallic catalysts, Ca–Cu
phosphate provides the highest catalytic performance. In
comparison to monometallic Cu P O catalyst prepared at
The efficiencies of monometallic Cu P O catalysts prepared
2
2
7
at different Cu/P atomic ratios of precursors for the oxidation
of furfural to maleic acid are listed in Fig. 5. As the XRD
data confirmed (Fig. 1) the samples of Cu P O catalysts
2
2
7
prepared at the Cu/P atomic ratios of precursors 0.5 and 1.25
after calcinations at 600 °C display the structure of copper
pyrophosphate; however, their catalytic activities are sig-
nificantly different (Fig. 5). As is seen with increasing Cu/P
atomic ratio increases both activity and selectivity, reaching
the maximum at the ratio of 1.25 and then decreases. Among
these catalysts, the Cu P O catalyst prepared at Cu/P atomic
2
2
7
the same Cu/P atomic ratio (1.0) and tested at comparable
conditions the conversion of furfural and mainly selectivity
to maleic acid is increasing from 64.5 and 39.8% (Fig. 5)
to 68.0 and 54.8% (Fig. 6), respectively. On the other hand,
the catalysts doped with Sr and Ba are less active (48.9 and
50.0% conversion, respectively), however, in comparison to
monometallic Cu P O catalyst they are more selective to
2
2
7
ratio 1.25 provides at 62.0% furfural conversion 51.3%
selectivity to maleic acid and 4.5% to furoic acid. When the
reaction was carried out under the same reaction conditions
in the absence of metal catalyst 50.4% of furfural was con-
verted, but maleic acid was formed only in trace amounts. It
confirms that during non-catalytic oxidation the main part
of furfural (about 80%) was converted to polymeric prod-
ucts. Since at elevated temperatures and pressures of oxygen
the rate of formation of oligomers by parallel non-catalyzed
reaction is enhanced, the main problem of furfural oxidation
in liquid phase is how to avoid the formation of resins. As is
seen from the results in Fig. 5 a significant improvement of
selectivity to maleic acid is achieved in conduction of oxida-
tion reaction in the presence of copper catalysts. The effect
of these catalysts is related with the reduction of the rate of
2
2
7
maleic acid (selectivity 43.4 and 34.7%, respectively). Com-
parison the activity and the selectivity of studied bimetallic
copper phosphate catalysts with monometallic Cu P O cat-
2
2
7
alyst indicates that in their presence is substantially reduced
the polymerization of reaction intermediates to resins.
As was described above the highest catalytic perfor-
mance for furfural oxidation to maleic acid was achieved
using bimetallic Ca–Cu phosphate catalyst prepared
at the atomic ratios of precursors corresponding to
Ca:Cu:P = 1:1:2. During this experiment the molar ratio
of furfural/Ca–Cu phosphate catalyst (calculated on Cu)
Fig. 5 The effect of the Cu/P atomic ratio during Cu P O catalyst
2
2
7
preparation on the catalyst performance. Conditions: 115 °C; 0.8 MPa
oxygen; 0.250 g furfural; 5.0 mL water; 0.050 g of catalyst; reaction
time 18 h
Fig. 6 Catalytic performance of bimetallic MCuP O pyroph-
2
7
osphates prepared at atomic ratios Cu/P=0.5 and M/Cu=1;
(M:Cu:P=1:1:2). Conditions: See Fig. 5
1
3

Aqueous-Phase Oxidation of Furfural to Maleic Acid Catalyzed by Copper Phosphate Catalysts
Fig. 7 The effect of the Ca/Cu atomic ratio during preparation of
Fig. 8 The effect of the Cu/P atomic ratio during preparation of
CaCuP O catalyst (Cu:P=1:2) on the catalyst performance. Condi-
CaCuP2O7 catalyst (Ca:Cu=1:1) on the catalyst performance. Condi-
tions: See Fig. 5
2
7
tions: see Fig. 5
was 15:1. Since this bimetallic catalyst is highly efficient
for furfural oxidation to maleic acid, another bimetallic
catalysts with various Ca:Cu:P ratios were prepared and
evaluated. In Fig. 7 is shown the influence of calcium con-
centration on the catalytic properties of bimetallic Ca–Cu
catalysts prepared at the atomic ratio of Cu:P = 0.5. From
these results is clearly evident that in the range of the
atomic ratios of Ca:Cu=0–1 the catalytic activity is almost
the same (the conversion of furfural is about 66–68%),
and only at the ratio 1.5 conversion slightly decreases (to
ca. 62%). However, the presence of calcium has a highly
positive effect on the selectivity of maleic acid formation.
The selectivity increases from initial 28.8%, obtained with
monometallic Cu P O catalyst, to 54.8% in the presence
For better understanding of the effect of alkali earth
metals in the M–Cu-phosphate structure on their catalytic
performance some bimetallic phosphate catalysts using the
abrasive linear sweep voltammetry (AbLSV) method were
studied. As is seen from results in Table 2, for monome-
tallic Cu P O catalyst a very small oxidation peak with
2
2
7
−
9
area only 0.5 × 10 C was observed. This potential at
−0.042 ± 0.021 V corresponds to the position of the oxi-
dation peak observed for pure Cu O. In comparison to
2
the monometallic Cu-phosphate, in the bimetallic M-Cu-
phosphate catalysts, especially doped with calcium and
magnesium, intensive oxidation peaks were registered. The
position of these oxidation peaks in the catalysts is also the
same as the position of the oxidation peak observed for pure
2
2
7
of bimetallic catalyst prepared at the Ca/Cu atomic ratio 1.
Cu O. It suggests that in these catalysts copper in oxida-
2
At the ratio 1.5 the selectivity sharply decreases to about
tion state Cu(I) is present. The most intense oxidation peak
2
0%. The same catalytic activity (furfural conversion) but
was observed for Ca–Cu-pyrophosphate catalyst (peak area
−
9
a very strong effect of calcium on the improvement of
selectivity evidently must be related with the reduction of
non-selective polymerization reactions.
14.7 ± 2.3 × 10 C) followed by bimetallic Mg–Cu-phos-
−
9
phate (peak area 8.0 ± 1.1 × 10 C). As mentioned above
(Fig. 6), using these catalysts the highest selectivity to
maleic acid (54.8 and 49.0%) was achieved. For other types
of bimetallic phosphates (Sr and Ba) the oxidation peak
area and also their activity and selectivity to maleic acid
was lower. The electrode potential of bimetallic catalysts
decrease in the order: Ca > Mg > Sr > Ba. This order cor-
relates well with changes of the yields and selectivities to
maleic acid obtained using these catalysts.
As was described above (Fig. 5) the selectivity of oxida-
tion using monometallic Cu P O catalysts increases with
2
2
7
increasing Cu/P atomic ratio of precursors used in the prepa-
ration step, reaching the maximum at the value of 1.25. The
previous study of the influence of calcium on the catalytic
performance of Ca–Cu-phosphate catalysts was followed
using samples prepared at the atomic ratios Ca:Cu:P=1:1:2.
To achieve better results we prepared and tested bimetallic
Ca–Cu-phosphate catalysts with different Cu/P atomic ratio
and having the ratio Ca:Cu=1:1. In contrary to the results
depicted in Fig. 5, with increasing (Ca–Cu)/P atomic ratio
decreases sharply both the catalytic activity and selectiv-
ity to maleic acid (Fig. 8). It means that in the presence of
these catalysts a significant part of furfural was converted
to resins.
3.5.4 Catalyst Recycling
In order to study the catalyst deactivation during the furfural
oxidation, we have carried out four reaction runs using mon-
ometallic Cu P O catalyst and bimetallic Ca–Cu-phosphate
2
2
7
catalyst. The catalysts were prepared at Cu/P atomic ratio 1
and bimetallic catalyst at the atomic ratios Ca:Cu:P=1:1:2.
1
3

T. Soták et al.
At the end of each reaction the catalyst was filtered out,
washed only with water and submitted directly to the next
run. In Fig. 9 the results for both catalysts obtained after four
cycles are presented. As shown in the figure despite different
initial selectivity both catalysts exhibit practically the same
selectivity in the first and in the fourth cycle. Catalytic activ-
ity slightly decreases, while for both catalysts this decrease
exhibits the same trend; during fourth cycles conversion of
furfural dropped by about 8% and about 13% for Cu P O
Excluded cannot be the deactivation of the catalyst by res-
inous by-products adsorbed on the catalyst surface since
between recycling cycles the catalyst was washed only with
water. Therefore, polymeric products which are insoluble in
water, may remain adsorbed on the catalyst and their amount
is increasing gradually with the recycling runs.
2
2
7
and bimetallic Ca–Cu-phosphate catalyst, respectively.
4 Discussion
The phase composition of the Cu phosphate catalyst
should be changing during the catalytic tests, using oxy-
gen as oxidizing agent. X-ray diffraction patterns of the
Cu P O catalyst recorded after a catalytic test revealed a
In the literature are proposed various mechanisms for fur-
fural oxidation using homogeneous and heterogeneous cata-
lysts and oxygen or hydrogen peroxide as oxidation agents
[3–5, 20–22]. During heterogeneously catalyzed reaction
furfural molecule lies essentially flat on the surface and the
π-electrons in the furan ring and carbonyl interacts more
strongly with the surface [23, 24]. This interaction may lead
to abstraction of the hydrogen atom at the 5-position of fur-
fural or at the aldehyde group [12] and generate radicals
1,6 (Scheme 1). These radicals can attack another furfural
molecule and promote the conversion of furfural to resins, or
can be trapped by organic radical scavengers or some metal
ions. It is well known, that copper (II) is a highly efficient
trap for organic radicals to generate organic cation inter-
mediates. Accordingly, if in the reaction system is present
suitable copper catalyst the furfural radical is converted to
cation intermediate 2 leading in the next steps to small oxy-
gen containing molecules. As demonstrated in Scheme 1, in
the process of furfural oxidation the electron transfer reac-
tion generating furfural cation is crucial for competing with
furfural polymerization, i.e. for achieving high selectivity of
oxidation to the desired product. The rate of electron transfer
reaction depends on the redox potential of copper cation
(Table 2) and probably also on the geometric configura-
tion of the sites hosting the active species (Fig. 3). As the
results of furfural oxidation using monometallic Cu P O
2
2
7
small modification of the catalyst in comparison with the
fresh one (Fig. 1d). In the sample of recycled catalyst, the
presence of peaks characteristic to the Cu (OH)PO phase
2
4
(
PDF No. 01-077-0922), especially at 2Θ=15° was regis-
tered. Despite the small amount of the formed phase these
data indicate that during the catalytic test some changes of
the catalyst took place. Transformation of Cu P O phase
2
2
7
to copper oxides during the catalytic test was not observed.
Moreover, these oxides alone are catalytically almost inac-
tive for furfural oxidation. In order to understand the pos-
sibility that homogeneous metallic species dissolved from
catalyst should be real catalyst, the filtrate of Cu P O
2
2
7
catalyst treated under reaction conditions was used for the
reaction. After addition of furfural and oxidation reaction
comparable results with non-catalyzed oxidation of furfural
were obtained (about 50% furfural conversion and less than
5
% selectivity to maleic acid). Leakage of metals into solu-
tion was confirmed of less than 3 ppm in the filtrate of the
reaction mixture. Therefore, the loss of activity is not only
due to catalyst losses in the separation step but also by a
partial dissolution of catalyst in aqueous reaction medium.
2
2
7
and bimetallic CaCu phosphate catalysts showed that the
selectivity of the reaction depends on the Cu/P and Ca/Cu
atomic ratio. This ratio influences the acidity (Table 1) and
the morphology (Fig. 3) of catalysts and as evidenced by the
H -TPR (Fig. 4) and voltammetry measurements (Table 2)
2
it resulted in modified redox properties of catalysts. As was
observed experimentally pure Ca (PO ) was found to be
3
4 2
inactive, while the activity of bimetallic Ca–Cu phosphate
increased as the sample is enriched in calcium up to the
atomic ratio Ca:Cu = 1 (Fig. 7). Since Ca (PO ) alone is
3
4 2
inactive for maleic acid formation, the role of added calcium
might be attributed to the acceleration of electron transfer
steps that are crucial for selective oxidation of furfural to
maleic acid. Thus, adding calcium can provide the highest
yield of expected product, while other alkaline earth met-
als may also improve the electron transfer ability of copper
Fig. 9 Recycling of catalysts. Conditions: See Fig. 5; catalyst
Cu P O prepared at the ratio Cu/P=1 and catalyst CaCu P O pre-
2
2
7
2
2
7
pared at the ratios Ca:Cu:P=1:1:2
1
3

Aqueous-Phase Oxidation of Furfural to Maleic Acid Catalyzed by Copper Phosphate Catalysts
Scheme 1 A plausible mecha-
nism for furfural oxidation to
maleic acid. RR E 1,4-rearran-
ciement reaction; cat CuE-
phosphate catalyst
catalyst (Table 2) but with less improvement than calcium
Fig. 6).
Attack of cation intermediate 2 with water, followed by
,4-rearrangement and subsequent decarbonylation and elec-
or more likely to react with oxygen to produce furoic acid
(
7. To test whether furoic acid is the final product of furfural
oxidation or can be further transformed to maleic acid, we
conducted a control experiment where furoic acid was sepa-
rately used as a starting substrate for oxidation under the
identical conditions of furfural oxidation. Using CaCu-phos-
phate catalyst (Ca:Cu:P=1:1:2) 36.8 mol% yield of maleic
acid from furoic acid was obtained. This yield is almost
the same as was reached in the experiment with furfural
(37.3 mol%, Fig. 7). Moreover, 33 mol% yield of maleic
acid and 50% conversion of furoic acid was achieved already
after 3 h of reaction. The rapid conversion of furoic acid
to maleic acid explains why in most reaction mixtures of
furfural oxidation furoic acid is present only in low concen-
trations. It confirms that for the current system also the path-
way through furoic acid might be considered in the reaction
mechanism. Probably furoic acid in the presence of copper
1
tron transfer reaction of formed intermediate 3 generates
cation intermediate which is attacked with water producing
intermediate 4. This is subsequently oxidized to maleic acid
as the major product of furfural oxidation. The decarbonyla-
tion of intermediate 3 was indicated by the detection of CO
and CO in the gas mixture after reaction.
2
In experiments using copper phosphate catalysts was
observed that beside the desired product, maleic acid, the
presence of small amount of furoic acid was also registered
in the reaction mixtures. It should suggest that in the initial
step of furfural oxidation also hydrogen from the aldehyde
group is abstracted, forming radical intermediate 6. The
intermediate 6 can decarbonylate, followed to form resins
1
3

T. Soták et al.
catalyst is decarboxylated forming cation intermediate 8
which is attacked with water, followed by 1,4-rearrangement
to generate the intermediate 9. This intermediate is further
catalytically oxidized to maleic acid. Therefore, it might be
assumed that in the case of studied copper phosphate cata-
lysts a route involving furoic acid as an intermediate of fur-
fural oxidation in water plays a role in maleic acid formation.
acid provide an alternative route to the homogeneously
catalyzed reaction.
Acknowledgements Support from the fund APVV-0297 is deeply
appreciated.
References
1
2
.
.
Gallezot P (2012) Chem Soc Rev 41:1538
5
Conclusions
Luterbacher JS, Alonso DM, Dumesic JA (2014) Green Chem
1
6:4816
3
.
Alonso-Fagúndez N, Granados ML, Mariscal R, Ojeda M (2012)
Chem Sus Chem 5:1984
In this work, new recyclable heterogeneous catalysts have
been explored to convert furfural in liquid phase to maleic
acid. Several copper phosphate type catalysts were prepared
by coprecipitation at different atomic ratios of precursors.
The samples were calcined at 600 °C and characterized by
physical–chemical methods and evaluated for the oxidation
of furfural in water. Experimentally it was observed that
the M/Cu/P atomic ratio of precursors in copper pyroph-
osphate catalysts and in mixtures with Mg, Ca, Sr, Ba met-
als influences the acidity and the morphology of catalysts
and strongly also their catalytic performance. It is assumed
that these effects are linkage to the modification of redox
potential of copper cation and probably also to geometric
configuration of the sites hosting the active species. Doping
of copper phosphate catalyst with calcium showed the best
catalytic performance for furfural oxidation to maleic acid.
Under reaction temperature of 115 °C and 0.8 MPa of oxy-
gen pressure and using CaCu-phosphate catalyst prepared at
the atomic ratio of precursors Ca:Cu:P=1:1:2, 54.8% selec-
tivity to maleic acid can be achieved. This catalyst recovered
by filtration and washed only with water could be recycled
for more than four times. Despite partial leaching of metals
from the catalyst to the solution the selectivity in all runs
remain practically the same, and the catalytic activity (fur-
fural conversion) after fourth cycles decreases only by about
4
5
.
.
Lan J, Chen Z, Lin J, Yin G (2014) Green Chem 16:4351
Guo H, Yin G (2011) J Phys Chem C 115:17516
6. Shi S, Guo H, Yin G (2011) Catal Commun 12:731
7
8
.
.
Kumar NS, Srivastava VC, Basu S (2013) Indian Chem Eng
5
5:153
Pinna F, Olivo A, Trevisan V, Menegazzo F, Signoretto M, Man-
zoli M, Boccuzzi F (2013) Catal Today 203:196
9. Choudhary H, Nishimura S, Erbitani K (2012) Chem Lett 41:409
1
0. Choudhary H, Nishimura S, Erbitani K (2013) Appl Catal A Gen
4
58:55
1
1
1. Du Z, Ma J, Wang F, Liu J, Xu J (2011) Green Chem 13:554
2. Li X, Ho B, Zhang Y (2016) Green Chem 18:2976
13. Wojcieszak R, Santarelli F, Paul S, Dumeignil F, Cavani F, Gon-
çalves RV (2015) Sustain Chem Process 3:9
1
1
1
4. Induja S, Raghavan PS (2013) Catal Commun 33:7
5. Blazy P, Samama JC (2001) C R Acad Sci Paris 333:271
6. Özer AK, Gülaboglu MS, Bayrakceken S, Weisweiler W (2006)
Adv Powder Technol 17:481
1
1
7. Khan N, Morozov VA, Khasanov SS, Lazoryak BI (1997) Mater
Res Bull 32:1211
8. Štolcová M, Litterscheid C, Hronec M, Glaum R (2007) Stud Surf
Sci Catal 167:37
19. Polnišer R, Štolcová M, Hronec M, Mikula M (2011) Appl Catal
A Gen 400:122
2
0. Alba-Rubioa AC, Fierro JLG, León-Reina L, Mariscal R, Dumesic
JA, López Granados M (2017) Appl Catal B 202:269
1. Alonso-Fagundez N, Agirrezabal-Telleria I, Arias PL, Fierro JLG,
Mariscal R, Granados ML (2014) RSC Adv 4:54960
2. Li X, Lan X, Wang T (2016) Catal Today 276:97
3. Pang SH, Medlin JW (2011) ACS Catal 1:1272
4. Sitthisa S, Pham T, Prasomsri T, Sooknoi T, Mallinson RG,
Resasco DE (2011) J Catal 280:17
2
2
2
2
1
3%. The mechanistic study of furfural oxidation leads to the
assumption that in the presence of studied copper catalysts
a route involving furoic acid as an intermediate plays a role
in maleic acid formation.
The advantages of simple isolation and reusability of
studied heterogeneous catalysts also the yields of maleic
1
3