One-step approach to 2,5-diformylfuran from fructose by proton- and vanadium-containing graphitic carbon nitride

Article abstract of DOI:10.1002/cctc.201402323

5-Hydroxymethylfurfural (HMF) was selectively synthesized from fructose by using protonated graphitic carbon nitride [g-C₃N₄ (H⁺)] as a solid acid. With g-C₃N₄ as the same precursor material, 2,5-diformylfuran (DFF) was efficiently obtained from aerobic oxidation of HMF by using vanadium-doped g-C₃N₄ (V-g-C₃N₄) as an environmentally benign heterogeneous catalyst with atmospheric pressure of molecular oxygen as the oxidant. In addition, a combination of g-C₃N₄(H⁺) and V-g-C₃N₄ successfully afforded the direct synthesis of DFF from fructose through g-C₃N₄(H⁺)-promoted fructose dehydration and V-g-C₃N₄-catalyzed successively aerobic oxidation of HMF to DFF in a one-pot and two-step process. Moreover, a bifunctional catalyst made of protonated V-g-C₃N₄ [V-g-C₃N₄(H⁺)] allowed onepot as well as one-step direct transformation of fructose into DFF. Under optimized conditions, 80% yield of HMF and 82% yield of DFF were obtained from g-C₃N₄(H⁺)-promoted fructose dehydration and V-g-C₃N₄-catalyzed HMF oxidation, respectively. Stepwise addition of g-C₃N₄(H⁺) and V-g-C₃N₄ in the one-pot process improved the DFF yield up to 63% from fructose. In contrast, V-g-C₃N₄(H⁺)-catalyzed direct one-step transformation of fructose into DFF led to a DFF yield of 45 %. The research highlights a one-step approach to DFF from fructose with functional g-C₃N₄ catalyst.

Full text of DOI:10.1002/cctc.201402323

CHEMCATCHEM
FULL PAPERS
DOI: 10.1002/cctc.201402323
One-Step Approach to 2,5-Diformylfuran from Fructose by
Proton- and Vanadium-Containing Graphitic Carbon
Nitride
Jinzhu Chen,*[a] Yuanyuan Guo,[a] Jiayi Chen,[a, c] Lu Song,[b] and Limin Chen[b]
5-Hydroxymethylfurfural (HMF) was selectively synthesized
from fructose by using protonated graphitic carbon nitride
[g-C3N4(H+)] as a solid acid. With g-C3N4 as the same precursor
material, 2,5-diformylfuran (DFF) was efficiently obtained from
aerobic oxidation of HMF by using vanadium-doped g-C3N4
(V–g-C3N4) as an environmentally benign heterogeneous cata-
lyst with atmospheric pressure of molecular oxygen as the oxi-
dant. In addition, a combination of g-C3N4(H+) and V–g-C3N4
successfully afforded the direct synthesis of DFF from fructose
through g-C3N4(H+)-promoted fructose dehydration and V–g-
C3N4-catalyzed successively aerobic oxidation of HMF to DFF in
a one-pot and two-step process. Moreover, a bifunctional cata-
lyst made of protonated V–g-C3N4 [V–g-C3N4(H+)] allowed one-
pot as well as one-step direct transformation of fructose into
DFF. Under optimized conditions, 80% yield of HMF and 82%
yield of DFF were obtained from g-C3N4(H+)-promoted fructose
dehydration and V–g-C3N4-catalyzed HMF oxidation, respec-
tively. Stepwise addition of g-C3N4(H+) and V–g-C3N4 in the
one-pot process improved the DFF yield up to 63% from fruc-
tose. In contrast, V–g-C3N4(H+)-catalyzed direct one-step trans-
formation of fructose into DFF led to a DFF yield of 45%. The
research highlights a one-step approach to DFF from fructose
with functional g-C3N4 catalyst.
Introduction
In view of diminishing fossil reserves and degradation of the
environment, a growing interest has recently been devoted to
fuels and chemicals from the sustainable biomass resource.[1–6]
Carbohydrates, obtained from renewable biomass, are current-
ly deemed as promising feedstocks for this purpose.[7–9]
Among the carbohydrate-derived platform chemical com-
pounds, 5-hydroxymethylfurfural (HMF) was identified as a ver-
satile molecule with particular interest.[10,11] Moreover, selective
oxidation of HMF to 2,5-diformylfuran (DFF) has received par-
ticular attention due to the various applications of DFF in fun-
gicides,[12,13] pharmaceuticals,[14] heterocyclic ligands,[15] and
functional polymers.[16,17]
(TEMPO) show good performance for this aerobic oxidation
process.[18–20] Moreover, various catalyst systems, such as
V2O5,[21] VOPO4·2H2O,[22]V-containing polymeric catalyst,[23]
Cu(NO3)2/VOSO4,[24] metal/bromide (Co/Mn/Br) catalyst,[25] poly-
mer-supported IBX amide,[26] Ru/HT,[27]K–OMS-2,[28] Ru/C,[29]Ag–
[33]
OMS-2,[30] TEMPO/Cu,[31]PVP–Pd NP,[32] and RuCo(OH)2CeO2
were recently developed to achieve this important process
under a more sustainable route with air or molecular O2 as an
oxidant. However, HMF, currently obtained from the acid-pro-
moted dehydration of fructose, is of high cost as raw material
for this purpose owing to its high energy cost for isolation and
purification. The progress in green chemistry could realize an
integrated chemical process in which multistep reactions are
consolidated in one pot. Accordingly, the direct transformation
of fructose to DFF in a one-pot process has recently attracted
much attention and a series of combined catalytic system, in-
cluding a H-form cation-exchange resin and VOHPO4,[21] Am-
berlyst-15 and Ru/HT,[27] Fe3O4–SBA–SO3H and K–OMS-2,[28]
have been investigated as a potential route for DFF production
from an economic and environmental point of view. In these
combined catalytic systems, both acid catalyst and redox cata-
lyst are essential for the one-pot but two-step approach to
DFF from fructose. Under optimal conditions, the overall DFF
yields from 45%[21,27] to 80%[28] calculated on fructose were ob-
tained by stepwise addition of these dual catalysts. Notably,
the coexistence of acid catalyst and redox catalyst in the initial
stage of reaction generally resulted in significantly decreased
DFF yields to 13%.[27] The low yield of DFF from fructose in the
one-step process was presumably related to the undesired de-
composition of fructose by redox catalyst in the presence of
In the literature for HMF-to-DFF transformation, a series of
classical oxidants such as pyridinium chlorochromate (PCC),
BaMnO4, NaOCl, and 2,2,6,6-tetramethylpiperidine-1-oxide
[a] Prof. Dr. J. Chen, Y. Guo, J. Chen
CAS Key Laboratory of Renewable Energy
Guangzhou Institute of Energy Conversion
Chinese Academy of Sciences
Guangzhou 510640 (P.R. China)
Tel./Fax: (+86)20-3722-3380
E-mail: chenjz@ms.giec.ac.cn
[b] L. Song, Dr. L. Chen
College of Environment and Energy
South China University of Technology
Guangzhou 510006 (P.R. China)
[c] J. Chen
University of Chinese Academy of Sciences
Beijing 100049 (P.R. China)
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/cctc.201402323.
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 3174 – 3181 3174

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
molecular oxygen.[27,28]In contrast to the one-pot but two-step
process for DFF production by using the above combined cat-
alytic system, a one-pot as well as one-step approach to DFF
with the yield of 60%from fructose was recently achieved by
us with Cs3HPMo11VO40 as a bifunctional and recyclable catalyst
and molecular oxygen as an oxidant under environmentally
benign conditions. Cs3HPMo11VO40 behaves as a heterogeneous
catalyst and can be easily reused at least four times with only
a slightly reduced DFF yield from 60 to 53%.[34]
Recently, Jiang and Han reported a selective synthesis of
phenol from benzene using a vanadium-doped graphitic
carbon nitride catalyst (V–g-C3N4) with hydrogen peroxide as
the oxidant.[35] Moreover, owning to the multifunctional cata-
lytic nature, the g-C3N4, upon deprotonation with tBuOK, ex-
hibits high efficiency for promoting Knoevenagel condensa-
tions and transesterification reactions as a heterogeneous solid
base catalyst, on the one hand.[36] On the other hand, g-C3N4
could be reversibly protonated by strong mineral acids, thus
modifying acidity, solubility/dispersibility, and surface area.[37]
Inspired by these works, herein, we report that a combination
of g-C3N4(H+) and V–g-C3N4 catalysts successfully catalyzed
direct synthesis of DFF from fructose by acid-catalyzed dehy-
dration and successive aerobic oxidation in a one-pot reaction
(Scheme 1). Stepwise addition of g-C3N4-derived binary cata-
Scheme 1. Conversion of fructose into HMF and DFF catalyzed by function-
alized g-C3N4.
lysts gave DFF in 63% yield. Moreover, a bifunctional catalyst
made of protonated V–g-C3N4 [V–g-C3N4(H+)] allowed one-step
direct transformation of fructose to DFF with a yield of 45%
(Scheme 1).
Figure 1. a) Effect of reaction temperature and time on fructose dehydration
to HMF using g-C3N4(H+) as a solid acid. Reaction conditions: fructose
(200 mg), g-C3N4(H+) (100 mg, acid density 0.32 mmolgÀ1), DMSO
(2 mL).b) Effect of acid density of g-C3N4(H+) on fructose conversion and
HMF yield. Reaction conditions: fructose (200 mg), g-C3N4(H+) (100 mg),
DMSO (2 mL), 1208C. The acid density of g-C3N4(H+) was determined by po-
tentiometric acid–base titration.
Results and Discussion
Dehydration of fructose to HMF over g-C3N4(H+)
both the conversion of fructose and the yield of HMF increased
with proton loading level in g-C3N4(H+) (Figure 1b).
According to the literature, g-C3N4(H+) was obtained by stirring
g-C3N4 with hydrochloric acid followed by centrifugal washing
with water until neutral condition.[37]In Figure 1a, the influence
of reaction temperature and time on the dehydration of fruc-
tose catalyzed by g-C3N4(H+) using DMSO as solvent is shown.
Both the conversion of fructose and the yield of HMF increased
with reaction time at all temperatures investigated. The yield
of HMF increased slowly to 34% after 2 h at 1008C, whereas it
increased quickly to 80% after 0.5 h at 1408C, confirming that
increasing the reaction temperature promotes the conversion
of fructose to HMF. The influences of acid density of g-
C3N4(H+)on fructose-to-HMF transformation suggested that
The developed g-C3N4(H+) was further compared with com-
mercial acid catalysts such as Amberlyst-15, H-ZSM-5 (Si/Al=
25), HY-zeolite (Si/Alꢀ5.2), and H-Beta (Si/Al=12.5) for fruc-
tose-to-HMF transformation. In Table 1it is shown that the cata-
lytic activity for fructose-to-HMF transformation at 1308C de-
creases in the order of g-C3N4(H+)>HCl>Amberlyst-15
(run 3)>H-ZSM-5 (Si/Al=25)>HY-zeolite (Si/Alꢀ5.2)>H-Beta
(Si/Al=12.5). The reusability of g-C3N4(H+) was performed and
in Figure 2the activity of g-C3N4(H+) catalyst is shown during
six consecutive uses. The recovered g-C3N4(H+) was separated
and cleaned, and then applied in the next run under the same
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 3174 – 3181 3175

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
Table 1. Fructose dehydration to HMF over various acid catalysts.[a]
Table 2. Catalytic activity of various metal-doped g-C3N4 for aerobic oxi-
dation of HMF to DFF.[a]
Run
Catalyst
Fructose
conv. [%]
HMF
yield [%]
[b]
Run
Catalyst M–g-C3N4
(M [mmolgÀ1])
HMF
DFF
conv. [%]
yield [%]
1
2
3
4
5
6
7
g-C3N4(H+)[a,b]
>99
>99
>99
>99
>99
>99
94
80
89
78
76
74
71
63
Amberlyst-15[a]
Amberlyst-15[a,c]
HCl[a,d]
1
2
3
4
5
6
7
8
g-C3N4
33
20
7
3
Cu–g-C3N4 (1.53)
Mn–g-C3N4 (1.25)
Co–g-C3N4 (1.13)
Ni–g-C3N4 (1.27)
Ru–g-C3N4 (1.66)
Fe–g-C3N4 (1.34)
Mo–g-C3N4 (1.44)
V–g-C3N4 (0.66)
V–g-C3N4 (1.01)
V–g-C3N4 (1.09)
V–g-C3N4 (1.39)
V–g-C3N4 (1.70)
24
4
H-ZSM-5(Si/Al=25)[a]
HY-zeolite (Si/Alꢀ5.2)[a]
H-Beta(Si/Al=12.5)[a]
24
3
23
6
20
8
35
45
76
83
92
>99
>99
16
23
68
69
75
82
54
[a] Reaction conditions: fructose (200 mg, 1.11 mmol), solid acid catalyst
(100 mg), DMSO (2 mL), 1308C, 1.5 h. [b] g-C3N4(H+) (acid density
0.32 mmolgÀ1). The acid density of g-C3N4(H+) was determined by poten-
tiometric acid–base titration. [c] An amount of 14 mg Amberlyst-15 (acid
density 2.25 mmolgÀ1) was used instead of 100 mg as an equimolar
amount of protons with g-C3N4(H+) (100 mg, acid density 0.32 mmolgÀ1).
[d] HCl (0.032 mmol) was used as the acid catalyst.
9
10
11
12
13
[a] Reaction conditions: HMF (126 mg, 1 mmol), catalyst (100 mg), DMSO
(2 mL), O2 (0.1 MPa), 1308C, 6 h. [b] Metal content in the catalyst was de-
termined by ICP–AES.
ferent metals (M–g-C3N4) under atmospheric pressure of mo-
lecular oxygen. Without doping, pristine g-C3N4 exhibited ex-
tremely inferior activity towards the aerobic oxidation of HMF
(Table 2, run 1). Moreover, the activity of Cu-, Mn-, Co-, and Ni-
doped g-C3N4 was negligible under the investigated conditions
(runs 2–5). Ru was doped into g-C3N4 as a catalyst for compari-
son purpose. However, as shown in Table 2 (run 6), Ru–g-C3N4
was not efficient for aerobic oxidation of HMF to DFF in this re-
action system. Fe- and Mo-doped g-C3N4 catalysts were found
to be moderate active in the synthesis of DFF from HMF (runs
7–8). Notably, V–g-C3N4 was proved to be considerably more
active and selective to afford DFF than other metals (runs 9–
13).
In addition to metallic sites, the influences of V contents on
the catalytic performances of V–g-C3N4 catalysts were further
explored for the HMF oxidation. The conversion of HMF in-
creased with the V content in V–g-C3N4, whereas DFF yields
smoothly increased with the V contents to a maximum DFF
yield of 82% at the vanadium content of 1.39 mmolgÀ1 in V-g-
C3N4. Further increased V content, however, led to a sharply re-
duced DFF yield (Table 2, runs 9–13). Such behavior may be ex-
plained as a result of the increase in the availability and
number of catalytically active V sites in V–g-C3N4, which pro-
motes the oxidation of HMF to DFF, on the one hand; howev-
er, it simultaneously promotes the subsequent oxidation of
DFF to 5-formyl-2-furancarboxylic acid (FFCA) on the other
hand as the V-loading increases. Notably, the desired DFF was
always the predominantproduct under the investigated condi-
tions.
Figure 2. Fructose dehydration to HMF during six consecutive catalytic uses
with g-C3N4(H+). Reaction conditions: fructose (200 mg), g-C3N4(H+) (100 mg,
acid density 0.32 mmolgÀ1), DMSO (2 mL), 1208C, 1.0 h. CFru =conversion of
fructose, YHMF =yield of HMF.
conditions. However, there was a dramatic decrease of conver-
sion in the second run. The ion chromatography analysis of
the reaction solution after the first run indicated that the con-
tent of chlorine in the solution was approximately 19% of the
chlorine in the catalyst.Notably, a subsequent treatment of the
catalyst with a 0.1m aqueous solution of HCl allowed recovery
of the catalytic activity of g-C3N4(H+). Apparently, the reduced
activity of g-C3N4(H+) can presumably be related to the acidic
proton lost from g-C3N4(H+) during its recovery.
In Figure 3, the influence of reaction temperature and time
on the HMF-to-DFF transformation is shown with V–g-C3N4 as
the catalyst. Generally, both the HMF conversion and DFF yield
increased with reaction time at the temperatures investigated.
In the case of the reaction temperature at 1308C, the DFF
yields increased with the reaction time with a maximum of
82% at 6 h. A further increased reaction time to 8 h, however,
Aerobic oxidation of HMF to DFF over V–g-C3N4
In addition to fructose dehydration to HMF catalyzed by solid-
acid g-C3N4(H+), we further explored aerobic oxidation of HMF
to DFF promoted by V–g-C3N4. An initial experiment was per-
formed to investigate the effect of the metallic sites on the
aerobic oxidation of HMF to DFF using g-C3N4 doped with dif-
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 3174 – 3181 3176

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
crease in catalytic activity could be presumably attributed to
minor leaching of V from V–g-C3N4.
Recently, Fu and co-workers proposed a redox mechanism
of K–OMS-2-promoted transformation of HMF into DFF using
molecular oxygen under base-free conditions.[28] The proposed
À
mechanism involves the oxidation of HMF by Mn4+O2
through two-electron transfer, and a simultaneous reduction of
À
Mn4+O2 to Mn3+ ion, which is reoxidized by O2. Moreover,
Han and co-workers suggested a mechanism of hydroxylation
of benzene to phenol using V–g-C3N4 as the catalyst with hy-
drogen peroxide as the oxidant.[35] In principle, our reported re-
sults seemed to agree with the above mechanisms. HMF is oxi-
dized by VV species, simultaneously the VV species is reduced
to VIV species (see the following X-ray photoelectron spectros-
copy analysis for details). Furthermore, VIV species could be
converted to VV species by O2 (Scheme S1 in the Supporting
Information).
Figure 3. Effect of reaction temperature and time on the aerobic oxidation
of HMF to DFF over V–g-C3N4. Reaction condition: HMF (126 mg, 1 mmol),
V–g-C3N4 (100 mg, V 1.39 mmolgÀ1), DMSO (2 mL), O2 (0.1 MPa).
One-pot conversion of fructose to DFF
We have examined the dehydration of fructose to HMF cata-
lyzed by g-C3N4(H+) and aerobic oxidation of HMF to DFF pro-
moted by V–g-C3N4. We thus further investigated the one-pot
synthesis of DFF from fructose by using a combination of g-
C3N4(H+) and V–g-C3N4 (Table 3). Addition of V–g-C3N4 as an ox-
led to a steeply reduced DFF yield to 65%, owing to subse-
quent oxidation of DFF to FFCA. The effect of reaction temper-
ature demonstrated that increasing the reaction temperature
promotes the aerobic oxidation of HMF to DFF.
Table 3. One-pot synthesis of DFF from fructose using g-C3N4(H+) and
V–g-C3N4 under atmospheric pressure of molecular oxygen.[a]
To probe the reusability of V–g-C3N4, a five-cycle experiment
was performed. The V–g-C3N4 catalyst was recovered by filtra-
tion, washed with DMSO, and then recycled for further reac-
tion. Notably, the measured yields of the DFF reduced from 82
to 70% (Figure 4). Inductively coupled plasma atomic emission
spectroscopy (ICP–AES) analysis of the reaction solution after
the first run indicated that the content of V in the solution was
approximately 2.7% of the V in the catalyst. Therefore, the de-
[c]
Run
V–g-C3N4
Fructose
conv. [%]
HMF
yield [%]
DFF
yield [%]
(V [mmolgÀ1])
1[a]
2[a]
3[a]
4[a]
5[a]
6[a,b]
V–g-C3N4 (0.66)
V–g-C3N4 (1.01)
V–g-C3N4 (1.09)
V–g-C3N4 (1.39)
V–g-C3N4 (1.70)
V–g-C3N4 (1.39)
>99
>99
>99
>99
>99
>99
14
3
8
39
47
52
63
25
26
1
trace
18
[a] Reaction conditions: fructose (200 mg, 1.11 mmol), g-C3N4(H+)
(100 mg, acid density 0.32 mmolgÀ1), DMSO (2 mL). After 2 h reaction
under N2 (0.1 MPa) at 1208C, V–g-C3N4 (100 mg) was added under O2
(0.1 MPa) at 1308C for 6 h. [b] HCl (0.032 mmol) was used as the acid in-
stead of g-C3N4(H+) (100 mg, acid density 0.32 mmolgÀ1). [c] The vanadi-
um content of catalyst was determined by ICP–AES.
idation catalyst afforded DFF directly from fructose in the pres-
ence of solid acid g-C3N4(H+). However, the existence of molec-
ular oxygen from the initial stage of the fructose dehydration
gave a trace amount (2% yield) of DFF, presumably owning to
the undesired oxidation of fructose by V–g-C3N4 in the pres-
ence of molecular oxygen.[27,28] Notably, the two-step synthesis
in the one-pot reaction without catalyst separation improved
the DFF yield of up to 63% (Table 3, run 4). In addition, with g-
C3N4(H+) as a solid acid for the first step, DFF yields, again, in-
creased with the vanadium contents in V–g-C3N4 to a maximum
of 63% at the vanadium content of 1.39 mmolgÀ1. Further in-
creased vanadium content to 1.70 mmolgÀ1, however, led to
a sharply reduced DFF yield to 25% (run 5). Notably, a combina-
Figure 4. Catalyst V–g-C3N4 recycling. Reaction condition: HMF (126 mg,
1 mmol), V–g-C3N4 (100 mg, V 1.39 mmolgÀ1), DMSO (2 mL), O2 (0.1 MPa),
6 h, 1308C.
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 3174 – 3181 3177

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
tion of HCl and V–g-C3N4 resulted in a decreased DFF yield to
26% (run 6), further indicating the high efficiency of g-C3N4(H+)
and V–g-C3N4 system (run 4).
Table 4. One-step synthesis of DFF from fructose using bifunctional
V–g-C3N4(H+) under atmospheric pressure of molecular oxygen.
Run
Catalyst[d]
V [mmolgÀ1 H+ [mmolgÀ1
Fructose
conv. [%] yield [%] yield [%]
HMF
DFF
In Figure 5, the reaction profile of these two-step sequential
reactions in a one-pot process is shown, the addition of g-
C3N4(H+) under atmospheric pressure of N2 resulted in the dis-
]
]
1[a]
2[a]
3[a]
4[a]
5[a]
6[a]
7[b]
8[c]
0.64
0.78
0.82
0.94
1.02
1.11
0.78
0.78
1.53
1.58
1.83
0.68
0.60
0.59
1.58
1.58
>99
>99
>99
>99
>99
>99
>99
>99
37
8
18
42
39
38
2
trace
21
36
34
9
11
12
45
39
[a] Reaction conditions: fructose (200 mg, 1.11 mmol), V–g-C3N4 (H+)
(100 mg), DMSO (2 mL), 1308C. After 2 h reaction under N2 (0.1 MPa), the
reaction was performed under O2 (0.1 MPa) for 6 h. [b] Reaction condition
as in [a], but O2 pressure was 1.0 MPa. [c] Reaction condition as in [a],
but O2 pressure was 2.0 MPa. [d] The vanadium content and acid density
of catalyst were determined by ICP–AES and potentiometric acid–base ti-
tration, respectively.
of HCl after each reuse. However, the DFF yield dramatically
decreased from 45% to 20% in the third run (Table S2).
Recently, Riisager and co-workers reported the aerobic oxi-
dation of HMF to DFF with zeolite-supported vanadia cata-
lysts.[38] Satsuma et al.,[39] Nijhuis et al.,[40] and our group[41] ob-
tained HMF by fructose dehydration using zeolite as solid acid.
Therefore, for comparison purpose, zeolite-supported vanadia
catalysts (V2O5/zeolite) were further investigated for one-pot
approach to fructose-to-DFF transformation. Zeolites such as
H-Beta (Si/Al=12.5), HY-zeolite (Si/Alꢀ5.2), and H-ZSM-5 (Si/
Al=25, 50 and 150) were chosen as supports owing to their
acidity, high thermal stability, and high surface areas. However,
in sharp contrast to V–g-C3N4(H+), the V2O5/zeolite exhibited
extremely inferior activity towards the direct synthesis of DFF
from fructose under the investigated conditions (Table 5). Ac-
cording to the literature, the Lewis acidity of the zeolite de-
Figure 5. Fructose conversion into HMF and DFF as a function of reaction
time. Reaction condition: fructose (200 mg), g-C3N4(H+) (100 mg, acid density
0.32 mmolgÀ1), DMSO (2 mL), 1308C. After 2 h reaction under N2 (0.1 MPa),
V–g-C3N4 (100 mg, V 1.39 mmolgÀ1) was added under O2 (0.1 MPa).
appearance of fructose and simultaneous formation of HMF
with a maximum yield of 80% at 2 h. Further addition of V–g-
C3N4 without catalyst separation under atmospheric pressure
of molecular oxygen afforded DFF from aerobic oxidation of
HMF. The overall yield of DFF is up to 63% at 6 h based on
fructose used under the investigated condition.
In addition to a dual catalyst system of g-C3N4(H+) and V–g-
C3N4, a bifunctional catalyst made of protonated V–g-C3N4
[V–g-C3N4(H+)] was further investigated for one-step approach
for fructose-to-DFF transformation. In Table 4, the results of
DFF formation from fructose by successive dehydration and
oxidation are listed using V–g-C3N4(H+) with variable acid den-
sity and V content. Under the investigated reaction conditions,
optimal results of 36% yield in DFF were obtained by using
the V–g-C3N4(H+) with acid density of 1.58 mmolgÀ1 and V con-
tent of 0.78 mmolgÀ1 under atmospheric pressure of molecular
oxygen (Table 4, run 2). In addition, the DFF yield, obtained
from the direct fructose transformation, increased with the O2
pressure to a maximum of 45% at 1.0 MPa (run 7). A further in-
creased O2 pressure to 2.0 MPa, however, led to a slightly re-
duced DFF yield to 39% (run 8). Moreover, the direct fructose-
to-HMF transformation was further performed under scale-up
conditions. A DFF yield of 19% was obtained with two grams
of fructose and one gram of V–g-C3N4(H+) in the presence of
molecular oxygen (Table S1 in the Supporting Information).
The reusability of V–g-C3N4(H+) was investigated, the catalyst
was regenerated by treatment with a 0.1m aqueous solution
Table 5. Catalytic activity of V2O5/zeolite for direct transformation of fruc-
tose to DFF.
Run
Catalyst(V2O5/zeolite)
Fructose
conv. [%]
HMF
yield [%]
DFF
yield [%]
1
2
3
4
5
V2O5/H-Beta (Si/Al=12.5)
V2O5/HY-zeolite (Si/Alꢀ5.2)
V2O5/H-ZSM-5 (Si/Al=25)
V2O5/H-ZSM-5 (Si/Al=50)
V2O5/H-ZSM-5 (Si/Al=150)
>99
>99
>99
>99
>99
12
24
9
7
6
5
7
8
5
4
Reaction conditions: fructose (200 mg, 1.11 mmol), V2O5/zeolite (100 mg,
V2O5 3 wt%), DMSO (2 mL), 1308C. After 2 h reaction under N2 (0.1 MPa),
the reaction was performed under O2 (0.1 MPa) for 6 h.
creases the selectivity of HMF formation by a fast initial con-
densation of fructose over the Lewis acid sites of the zeolite.[40]
The polymeric humins were thus presumably the major prod-
ucts.
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 3174 – 3181 3178

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
Characterization ofthe catalyst
The catalysts g-C3N4, V–g-C3N4, and g-C3N4(H+) were compared
by XRD and X-ray photoelectron spectroscopy (XPS) analy-
ses.[42–44] XRD patterns of g-C3N4 and V–g-C3N4 with different
vanadium contents are shown in Figure S1 (Supporting Infor-
mation). In the case of g-C3N4, the peaks at 2q=27.5 and 13.08
correspond to the stacking of the conjugated interlayers and
in-plane ordering of tri-s-triazine units, respectively.[45,46] V–g-
C3N4 and g-C3N4 samples displayed similar diffraction peaks;
however, the diffraction peaks of vanadium species were unob-
served (Figure S1). Notably, the intensity of characteristic dif-
fraction peaks observed in pristine g-C3N4 significantly de-
creased with increasing V contents if compared with V–g-C3N4
samples, which is in good agreement with the reported litera-
ture.[35] For g-C3N4(H+), XRD patterns of g-C3N4(H+) exhibited
a significant decrease of peak intensity at 2q=13.08, suggest-
ing a slightly distorted structure of nitride pores after the pro-
tonation of g-C3N4 (Figure S2).[37]
The XPS spectra of g-C3N4, V–g-C3N4, and g–C3N4(H+) are
shown in Figure S3. In the case of V–g-C3N4, the photoelectron
peaks of main elements on the surface appear at binding ener-
gies of 288.2 eV (C1s), 399.2 eV (N1s), 530.8 eV (O1s), and
517.1 eV (V2p), respectively. For the high-resolution V2p XPS
spectra of V–g-C3N4, two main peaks appeared at 523.5 and
517.3 eV corresponding to V2p1/2 and V2p3/2, respectively,[47]
which suggests that the V species is V2O5 (Figure 6a). In con-
trast, the other two peaks located at 522.1 and 515.2 eV corre-
spond to V2p1/2 and V 2p3/2, respectively, confirming VO2
species is formed.[35] In Figure 6b–d, the N1s XPS spectra of g-
C3N4, V-g-C3N4, and g-C3N4(H+) are shown, respectively. The two
peaks at 401.2 and 399.8 eV can be assigned to tertiary nitro-
gen and amino functional groups, respectively. The peak at
398.6 eV corresponds to the nitrogen bonded to two carbon
atoms. In the case of g-C3N4(H+), the additional signal at
402.3 eV can be attributed to the protonation of the nitrogen
(C=NH+), indicating the formation of protonated g-C3N4.[48]
Conclusions
We demonstrated the dehydration of fructose to HMF cata-
lyzed by protonated graphitic carbon nitride (g-C3N4(H+)) as
a solid acid and aerobic oxidation of HMF to DFF promoted by
vanadium-doped g-C3N4(V–g-C3N4). In addition, a combination
of g-C3N4(H+) and V–g-C3N4 successfully afforded direct synthe-
sis of DFF from fructose in a one-pot reaction. Moreover, a bi-
functional catalyst of V–g-C3N4(H+) achieved a one-step ap-
proach to fructose-to-DFF transformation by acid-catalyzed de-
hydration and successive aerobic oxidation. The research high-
lights a good prospect for catalytic application of g-C3N4-de-
rived catalysts for biomass carbohydrate conversion.
Figure 6. a) V2p XPS spectra of V–g-C3N4 (V 1.39 mmolgÀ1); N1s XPS spectra
of b) g-C3N4, c) V–g-C3N4 (V 1.39 mmolgÀ1), and d) g-C3N4(H+) (acid density
0.32 mmolgÀ1).
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 3174 – 3181 3179

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
Fructose conversion into DFF
Experimental Section
One-pot two-step: In a typical run, the procedure for one-pot two-
step synthesis of DFF was as follows: the reaction was divided into
two steps: firstly, fructose (200 mg), g-C3N4(H+) (100 mg), and
DMSO (2 mL) were added into a 25 mL Schlenk tube, the dehydra-
tion reaction was performed at 1308C for 2 h under nitrogen at-
mosphere (0.1 MPa). Then V–g-C3N4 (100 mg, V 1.39 mmolgÀ1) was
added into the Schlenk tube, the oxidation reaction was performed
at 1308C for 6 h under oxygen atmosphere (0.1 MPa).
Materials and Characterization
Materials and their characterization are provided in the Supporting
Information.
Synthesis of g-C3N4, M–g-C3N4, g-C3N4(H+), and V–g-C3N4(H+)
[36]
g-C3N4 and metal-doped g-C3N4 (M–g-C3N4; M=V, Ni, Cu, Mo,
One-pot one-step: In a typical run, the procedure for one-pot one-
step synthesis of DFF was as follows: fructose (200 mg), V–g-
C3N4(H+) (100 mg) and DMSO (2 mL) were added into a 25 mL
Schlenk tube. The dehydration reaction was performed at 1308C
for 2 h under nitrogen atmosphere (0.1 MPa). Then, the oxidation
reaction was performed at 1308C for another 6 h under oxygen at-
mosphere (0.1 MPa).
Mn, Fe, and Co)[49,50] were prepared according to the literature
methods (see the Supporting Information for details). The V–g-
C3N4 with different vanadium contents were labeled as V–g-C3N4
(x), in which x represents the molar concentration (mmolgÀ1) of V
in V–g-C3N4 determined by ICP–AES. g-C3N4(H+) and V–g-C3N4(H+)
were prepared by protonation of g-C3N4 and V–g-C3N4, respectively,
according to the literature methods (see the Supporting Informa-
tion for details). The acid density of g-C3N4(H+) and V–g-C3N4(H+)
was determined by potentiometric acid–base titration.
Acknowledgements
This work was financially supported by National Basic Research
Program of China (973 Program, 2012CB215304), 100 Talents
Program of the Chinese Academy of Sciences, Guangdong Natu-
Fructose conversion into HMF
In a typical run, the procedure for fructose dehydration into HMF
was as follows: fructose (200 mg), g-C3N4(H+) (100 mg), and DMSO
(2 mL) were added into a 25 mL Schlenk tube. The tube was
heated at 100–1408C for 2 h under N2. The concentrations of fruc-
tose and HMF were determined by HPLC (Shimadu LC-20AT)
equipped with UV and refractive-index detectors. The analysis con-
ditions were conducted with a Shodex Sugar SH-1011 column
(300ꢁ8 mm) by using HPLC-grade 0.005m H2SO4 water solution as
the eluent at a column temperature of 508C and a flow rate of
0.5 mLminÀ1. Fructose and HMF were detected with a refractive-
index detector and a UV detector, respectively. HMF: 1H NMR
(400 MHz, D2O) d=9.33 (s, 1H), 7.40 (d, J=3.6 Hz, 1H), 6.55 (d, J=
3.1 Hz, 1H), 4.57 ppm (s, 2H). 13C {1H} NMR (101 MHz, D2O) d=
180.5, 161.3, 151.8, 126.8, 110.9, 56.0 ppm. IR (KBr): n˜ =3377 (OÀH),
1670 cmÀ1 (C=O). Mass spectrum (EI) m/z=126 (M+).
ral
Science
Foundation
(S2013010012986
and
S2011010002274),and EU-Chinese Eco-fuel project (FP7-PEOPLE-
2009-IRSES Grant 246772).
Keywords: biomass · carbon · nitrides · oxidation · vanadium
[1] P. Gallezot, Chem. Soc. Rev. 2012, 41, 1538–1558.
[2] A. Corma, S. Iborra, A. Velty, Chem. Rev. 2007, 107, 2411–2502.
[3] G. W. Huber, S. Iborra, A. Corma, Chem. Rev. 2006, 106, 4044–4098.
[4] J. N. Chheda, G. W. Huber, J. A. Dumesic, Angew. Chem. 2007, 119,
7298–7318; Angew. Chem. Int. Ed. 2007, 46, 7164–7183.
[5] G. W. Huber, A. Corma, Angew. Chem. 2007, 119, 7320–7338; Angew.
Chem. Int. Ed. 2007, 46, 7184–7201.
[6] Y. T. Cheng, Z. Wang, C. J. Gilbert, W. Fan, G. W. Huber, Angew. Chem.
2012, 124, 11259–11262; Angew. Chem. Int. Ed. 2012, 51, 11097–11100.
[7] Y. Romꢂn-Leshkov, J. N. Chheda, J. A. Dumesic, Science2006, 312, 1933–
1937.
[8] J. B. Binder, R. T. Raines, J. Am. Chem. Soc. 2009, 131, 1979–1985.
[9] X. Qi, M. Watanabe, T. M. Aida, R. L. Smith, ChemSusChem 2010, 3,
1071–1077.
HMF conversion into DFF
In a typical run, the procedure for the HMF oxidation was as fol-
[10] R.-J. Van Putten, J. C. van der Waal, E. De Jong, C. B. Rasrendra, H. J.
Heeres, J. G. De Vries, Chem. Rev. 2013, 113, 1499–1597.
[11] A. A. Rosatella, S. P. Simeonov, R. F. M. Frade, C. A. M. Afonso, Green
Chem. 2011, 13, 754–793.
[12] M. Del Poeta, W. A. Schell, C. C. Dykstra, S. Jones, R. R. Tidwell, A. Czarny,
M. Bajic, A. Kumar, D. Boykin, Antimicrob. Agents Chemother.1998, 42,
2495–2502.
[13] A. Gandini, N. M. Belgacem, Polym. Int. 1998, 47, 267–276.
[14] K. T. Hopkins, W. D. Wilson, B. C. Bender, D. R. McCurdy, J. E. Hall, R. R.
Tidwell, A. Kumar, M. Bajic, D. W. Boykin, J. Med. Chem. 1998, 41, 3872–
3878.
[15] D. T. Richter, T. D. Lash, Tetrahedron Lett. 1999, 40, 6735–6738.
[16] A. S. Amarasekara, D. Green, L. D. Williams, Eur. Polym. J. 2009, 45, 595–
598.
[17] M. Baumgarten, N. Tyutyulkov, Chem. Eur. J. 1998, 4, 987–989.
[18] T. El-Hajj, J. C. Martin, G. Descotes, J. Heterocycl. Chem. 1983, 20, 233–
235.
[19] A. S. Amarasekara, D. Green, E. McMillan, Catal. Commun. 2008, 9, 286–
288.
[20] L. Cottier, G. Descotes, J. Lewkowski, R. SkowroÇski, E. Viollet, J. Hetero-
cycl. Chem. 1995, 32, 927–930.
[21] G. A. Halliday, R. J. Young, V. V. Grushin, Org. Lett. 2003, 5, 2003–2005.
lows: HMF (126 mg, 1 mmol), V–g-C3N4 (100 mg, V 1.39 mmolgÀ1
)
and DMSO (2 mL) were added into a 25 mL Schlenk tube. The tube
was heated at 1308C for 6 h under oxygen atmosphere (0.1 MPa).
After the reaction mixture was cooled, distilled water (8 mL) was
added and the mixture was filtrated; then the products were ex-
tracted by dichloromethane for qualitative analysis by GC–MS
(Aglient, 5973 Network 6890N). Byproducts of dehydration such as
levulinic acid and formic acid or further oxidation byproducts such
as 2-carboxy-5-(formyl)furan and furan-2,5-dicarboxylic acid were
undetected in the dichloromethane-extracted samples. The quanti-
tative analysis of DFF was performed by GC (FILI, GC-9790II) with
a KB-5 capillary column (30.0 mꢁ0.32 mmꢁ0.25 mm) and hydrogen
flame-ionization detector (FID). The products were detected by
using the following conditions: injector temperature 2608C, detec-
tor temperature 2808C, oven temperature 1008C was held for
1 min, 258CminÀ1 to 2008C, and held at 2008C for 1.5 min. o-
Cresol was used as an internal standard according to the litera-
1
ture.[51] DFF: H NMR (400 MHz, CDCl3) d=9.86 (s, 1H), 7.33 ppm (s,
1H).13C NMR (101 MHz, CDCl3) d=179.3, 154.4, 119.3 ppm.
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 3174 – 3181 3180

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
[22] C. Carlini, P. Patrono, A. M. R. Galletti, G. Sbrana, V. Zima, Appl. Catal. A
2005, 289, 197–204.
[39] K. Shimizu, R. Uozumi, A. Satsuma, Catal. Commun. 2009, 10, 1849–
1853.
[40] V. V. Ordomsky, J. van der Schaaf, J. C. Schouten, T. A. Nijhuis, Chem-
SusChem 2012, 5, 1812–1819.
[41] J. Z. Chen, K. G. Li, L. M. Chen, R. L. Liu, X. Huang, D. Q. Ye, Green Chem.
2014, 16, 2490–2499.
[42] Y. Wang, X. C. Wang, M. Antonietti, Angew. Chem. 2012, 124, 70–92;
Angew. Chem. Int. Ed. 2012, 51, 68–89.
[43] Y. Zhang, T. Mori, J. Ye, Sci. Adv. Mater. 2012, 4, 282–291.
[44] F. Dong, L. W. Wu, Y. J. Sun, M. Fu, Z. B. Wu, S. C. Lee, J. Mater. Chem.
2011, 21, 15171–15174.
[23] O. C. Navarro, A. C. Canꢃs, S. I. Chornet, Top. Catal. 2009, 52, 304–314.
[24] J. Ma, Z. Du, J. Xu, Q. Chu, Y. Pang, ChemSusChem 2011, 4, 51–54.
[25] W. Partenheimer, V. V. Grushin, Adv. Synth. Catal . 2001, 343, 102–111.
[26] Y.-S. Lee, H.-J. Yoon, J.-W. Choi, H.-S. Cho, J. Jang, J.-W. Byun, W.-J.
Chung, S.-M. Lee, Synlett 2010, 165–168.
[27] A. Takagaki, M. Takahashi, S. Nishimura, K. Ebitani, ACS Catal. 2011, 1,
1562–1565.
[28] Z. Z. Yang, J. Deng, T. Pan, Q. X. Guo, Y. Fu, Green Chem. 2012, 14,
2986–2989.
[45] B. Yue, Q. Li, H. Iwai, T. Kako, J. Ye, Sci. Technol. Adv. Mater. 2011, 12,
034401.
[46] W. Wang, G. Ding, T. Jiang, P. Zhang, T. Wu, B. X. Han, Green Chem.
2013, 15, 1150–1154.
[29] J. Nie, J. Xie, H. Liu, J. Catal. 2013, 301, 83–91.
[30] G. D. Yadav, R. V. Sharma, Appl. Catal. B 2014, 147, 293–301.
[31] T. S. Hansen, I. Sꢂdaba, E. J. Garcꢄa-Suꢂrez, A. Riisager, Appl. Catal. A
2013, 456, 44–50.
[47] J. S. Stevens, S. J. Byard, C. A. Muryn, J. Phys. Chem. B 2010, 114, 13961–
13969.
[32] B. Siyo, M. Schncider, M.-M. Pohl, P. Langer, N. Stcinfcldt, Catal. Lett.
2014, 144, 498–506.
[48] Z. X. Ding, X. F. Chen, M. Antonietti, X. C. Wang, ChemSusChem 2011, 4,
274–281.
[33] Y. Wang, B. Liu, K. Huang, Z. Zhang, Ind. Eng. Chem. Res. 2014, 53,
1313–1319.
[49] X. C. Wang, X. Chen, A. Thomas, X. Fu, M. Antonietti, Adv. Mater. 2009,
21, 1609–1612.
[34] R. L. Liu, J. Z. Chen, L. M. Chen, Y. Y. Guo, J. W. Zhong. ChemPlusChem
2014,
DOI:
10.1002/cplu.201402147.
http://dx.doi.org/10.1002/
[50] X. Xiang, L. He, Y. Yang, B. Guo, D. Tong, C. Hu, Catal. Lett. 2011, 141,
735–741.
[51] Y. Zheng, J. Liu, J. Liang, M. Jaroniec, S. Z. Qiao, Energy Environ. Sci.
2012, 5, 6717–6731.
cplu.201402147.
[35] G. D. Ding, W. T. Wang, T. Jiang, B. X. Han, H. L. Fan, G. Y. Yang, Chem-
CatChem 2013, 5, 192–200.
[36] F. Su, M. Antonietti, X. Wang, Catal. Sci. Technol. 2012, 2, 1005–1009.
[37] Y. Zhang, A. Thomas, M. Antonietti, X. Wang, J. Am. Chem. Soc. 2009,
131, 50–51.
Received: May 11, 2014
[38] I. Sꢂdaba, Y. Y. Gorbanev, S. Kegnæs, S. S. R. Putluru, R. W. Berg, A. Riisag-
er, ChemCatChem 2013, 5, 284–293.
Revised: June 11, 2014
Published online on September 15, 2014
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 3174 – 3181 3181