Transformation of cellulose and its derived carbohydrates into formic and lactic acids catalyzed by vanadyl cations

Article abstract of DOI:10.1002/cssc.201400150

The transformation of cellulose or cellulose-derived carbohydrates into platform chemicals is the key to establish biomass-based sustainable chemical processes. The systems able to catalyze the conversion of cellulose into key chemicals in water without the consumption of hydrogen are limited. We report that simple vanadyl (VO²⁺) cations catalyze the conversions of cellulose and its monomer, glucose, into lactic acid and formic acid in water. We have discovered an interesting shift of the major product from formic acid to lactic acid on switching the reaction atmosphere from oxygen to nitrogen. Our studies suggest that VO²⁺ catalyzes the isomerization of glucose to fructose, the retro-aldol fragmentation of fructose to two trioses, and the isomerization of trioses, which leads to the formation of lactic acid under anaerobic conditions. The oxidative cleavage of C=C bonds in the intermediates caused by the redox conversion of VO₂⁺/VO²⁺ under aerobic conditions results in formic acid and CO₂. We demonstrate that the addition of an alcohol suppresses the formation of CO ₂ and enhances the formic acid yield significantly to 70-75 %. Two birds with one stone: A shift of the major product from formic acid to lactic acid is observed on switching the reaction atmosphere from O₂ to N₂ during cellulose or glucose conversions catalyzed by vanadyl cations. Glucose is first isomerized into fructose, which undergoes retro-aldol fragmentation to form two trioses, which produce lactic and formic acid.

Full text of DOI:10.1002/cssc.201400150

CHEMSUSCHEM
FULL PAPERS
DOI: 10.1002/cssc.201400150
Transformation of Cellulose and its Derived Carbohydrates
into Formic and Lactic Acids Catalyzed by Vanadyl Cations
Zhenchen Tang, Weiping Deng, Yanliang Wang, Enze Zhu, Xiaoyue Wan, Qinghong Zhang,*
and Ye Wang*[a]
The transformation of cellulose or cellulose-derived carbohy-
drates into platform chemicals is the key to establish biomass-
based sustainable chemical processes. The systems able to cat-
alyze the conversion of cellulose into key chemicals in water
without the consumption of hydrogen are limited. We report
that simple vanadyl (VO2+) cations catalyze the conversions of
cellulose and its monomer, glucose, into lactic acid and formic
acid in water. We have discovered an interesting shift of the
major product from formic acid to lactic acid on switching the
reaction atmosphere from oxygen to nitrogen. Our studies
suggest that VO2+ catalyzes the isomerization of glucose to
fructose, the retro-aldol fragmentation of fructose to two trios-
es, and the isomerization of trioses, which leads to the forma-
tion of lactic acid under anaerobic conditions. The oxidative
cleavage of CÀC bonds in the intermediates caused by the
+
redox conversion of VO2 /VO2+ under aerobic conditions re-
sults in formic acid and CO2. We demonstrate that the addition
of an alcohol suppresses the formation of CO2 and enhances
the formic acid yield significantly to 70–75%.
Introduction
Cellulose, which is the main constituent of the abundant and
inedible lignocellulosic biomass and is a polymer of glucose
linked by b-1,4-glycosidic bonds, is a promising candidate for
renewable carbon resources to partially replace petroleum for
the production of chemicals and fuels.[1] The development of
chemocatalytic processes for the efficient transformation of
cellulose under mild conditions into platform chemicals, from
which various chemicals and fuels can be derived easily, has at-
tracted much attention in recent years.[2,3] Many studies have
been devoted to the acid-catalyzed conversion of cellulose to
its monomer, glucose, or to 5-hydroxymethylfurfural (HMF),
a platform compound.[4,5] The direct production of high-value-
added chemicals such as polyols, particularly ethylene or pro-
pylene glycol of commercial interest, from cellulose can also
be realized through the design of bi- or multifunctional catalyt-
ic systems.[6,7] However, these processes require expensive H2.
Studies on the transformation of cellulose into valuable chemi-
cals without the consumption of H2 are scarce.
As a key building-block chemical, lactic acid is widely used
in the chemical, food, pharmaceutical, and cosmetic indus-
tries.[8,9] Polylactic acid, the polymer of lactic acid, has attracted
much attention as a biodegradable plastic.[8,9] Currently, lactic
acid is produced mainly by the fermentation of carbohydrates
such as glucose and sucrose, but this biological process is ex-
pensive and cannot be applied directly to cellulose.[8,9] The de-
velopment of new chemical processes for the production of
lactic acid from cellulose is highly desirable. In addition, formic
acid is a promising source and storage platform for H2.[10] Many
recent studies have demonstrated that H2 can be produced
easily from formic acid under mild conditions.[11] The transfor-
mation of cellulose directly into formic acid under mild tem-
peratures would provide an energy-efficient alternative for the
sustainable production of H2.
A few studies have been devoted to the oxidative conver-
sion of cellulose or cellulose-derived carbohydrates to formic
acid.[12–14] Wasserscheid and co-workers reported that a Keggin-
type V-substituted heteropolyacid H5PV2Mo10O40 could catalyze
the conversion of glucose and some oligosaccharides under an
O2 atmosphere to formic acid.[12] A formic acid yield of ~50%
was obtained after 26 h of reaction at 353 K and an oxygen
pressure of ~3 MPa. With the assistance of p-toluenesulfonic
acid, cellulose could also be transformed to formic acid with
a yield of 22% after 66 h reaction at 363 K and an oxygen pres-
sure of ~3 MPa.[13] Fu and co-workers further demonstrated
that H5PV2Mo10O40 could catalyze the conversion of cellulose to
formic acid at high reaction temperatures in the absence of or-
ganic sulfonic acid; a formic acid yield of 35% was obtained
after 9 h of reaction at 443 K.[14]
[a] Z. Tang,+ W. Deng,+ Y. Wang, E. Zhu, X. Wan, Q. Zhang, Prof. Dr. Y. Wang
State Key Laboratory of Physical Chemistry of Solid Surfaces
Collaborative Innovation Center of Chemistry for Energy Materials
National Engineering Laboratory for Green Chemical Productions of Alco-
hols, Ethers and Esters
College of Chemistry and Chemical Engineering
Xiamen University
Xiamen 361005 (PR China)
E-mail: wangye@xmu.edu.cn
zhangqh@xmu.edu.cn
[+] These authors contributed equally to this work.
Supporting Information for this article is available on the WWW under
http://dx.doi.org/10.1002/cssc.201400150.
Part of a Special Issue on “The 6th Asia–Pacific Congress on Catalysis
(APCAT-6)“. To view the complete issue, visit:
http://onlinelibrary.wiley.com/doi/10.1002/cssc.v7.6/issuetoc.
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 2014, 7, 1557 – 1567 1557

CHEMSUSCHEM
FULL PAPERS
www.chemsuschem.org
However, only scarce studies have succeeded in the chemo-
catalytic production of lactic acid from carbohydrates under
mild conditions (<473 K). Holm et al. found that a beta zeolite
with SnIV in the framework as Lewis acid sites was efficient for
the conversion of glucose, fructose, or sucrose in methanol
into methyl lactate at 433 K.[15] SnIV grafted onto MCM-41
modified with carbon also catalyzed the transformation of car-
bohydrates to methyl lactate in methanol under similar condi-
tions.[16] However, these SnIV-based catalysts are limited to the
transformation of mono- or disaccharides. Solid Lewis acids
such as tungstated zirconia and alumina as well as lanthanide
triflates were reported to catalyze the conversion of cellulose
in water to lactic acid.[17,18] Recently, we found that a simple
PbII ion was highly efficient for the conversion of cellulose into
lactic acid in water.[19] Lactic acid yields of 68 and 62% were at-
tained for the conversion of ball-milled and microcrystalline
cellulose at 463 K for 4 and 15 h, respectively. The toxicity of
Pb may hamper the practical application of this system.
Figure 1. Catalytic performances of metal salts for the conversions of glu-
cose under aerobic conditions to formic acid and under anaerobic condi-
tions to lactic acid. Aerobic reaction conditions: glucose, 1.0 mmol; water,
20 cm3; catalyst, 0.20 mmol; temperature, 413 K; reaction time, 1 h; O2 pres-
sure, 2 MPa. Anaerobic reaction conditions: glucose, 1.0 mmol; water,
20 cm3; catalyst, 0.050 mmol; temperature, 433 K, time, 1.5 h; N2 pressure,
2 MPa.
It is highly desirable to develop more efficient and environ-
mentally friendly catalytic systems for the production of lactic
or formic acid from cellulose or cellulose-derived carbohy-
drates. Moreover, the formations of lactic and formic acid both
involve the selective cleavage of CÀC bonds in carbohydrates.
We still lack insight into how the CÀC bond is cleaved and
how to control the CÀC bond cleavage.
to lactic acid. The yield of lactic acid with VOSO4 under the
conditions of the experiments shown in Figure 1 (56%) was
better than that with the PbII catalyst (<30%), which was re-
ported to be an efficient catalyst in our previous paper.[19]
It becomes clear that VOSO4 is a unique catalyst for the for-
mation of both formic and lactic acid from glucose with yields
>50%. We have also investigated the catalytic performances
of several V compounds (which include VOSO4, VOHPO4,
VOC2O4, NH4VO3, and V2O5) for the conversion of glucose
under aerobic and anaerobic conditions. Under an O2 atmo-
sphere, the catalytic performance did not depend on the
source of the V catalyst significantly, and all the catalysts pro-
vided formic acid in yields of 45–55% (Figure S1 in the Sup-
porting Information). If N2 was used instead of O2, lactic acid
was formed for all the V compounds examined. Thus, it is the
V cation and not the counteranion that is responsible for the
formation of both formic and lactic acid. The yield of lactic
acid is somewhat dependent on the source of the V com-
pounds, and VOSO4 exhibits the highest lactic acid yield.
We have further investigated the effect of the concentration
of VOSO4 on its catalytic behavior. In the absence of a catalyst,
we found that the conversion of glucose was low under either
O2 or N2 and the major products were fructose and HMF,
which were expected to be formed by the isomerization of
glucose and the dehydration of fructose (Scheme S1). The
presence of VOSO4 increased the conversion of glucose signifi-
cantly under both aerobic and anaerobic conditions (Fig-
ure S2). The formic acid and lactic acid yields increased with
the VOSO4 concentration to reach 64 and 56% at VOSO4 con-
centrations of 25 and 2.5 mm, respectively. The highest turn-
over frequencies (TOFs) for the formation of formic and lactic
acid, that is, the moles of formic and lactic acid produced per
mole of catalyst per unit time, were 63 and 21 hÀ1 at 413 and
433 K, respectively (Figure S2).
Very recently, we found that a cheaper and less toxic vanadi-
um salt, VOSO4, could catalyze the transformations of glucose
and cellulose into either formic acid or lactic acid by simply
tuning the reaction atmosphere from O2 to N2. This paper re-
ports a new “two birds, one stone” catalytic system. The reac-
tion pathways and mechanisms for the formation of lactic and
formic acid are investigated and discussed to gain an insight
into the chemistry that controls the cleavage of CÀC bonds in
carbohydrates and the derived intermediates.
Results and Discussion
Catalytic behavior under aerobic and anaerobic conditions
As mentioned above, H5PV2Mo10O40 has been studied exten-
sively as a homogeneous catalyst for the oxidative conversion
of glucose to formic acid.[12–14] To explore the possibility of
other catalysts for this transformation, we have investigated
the catalytic performances of various metal cations for the con-
version of glucose in water under O2. Without a metal catalyst
(blank reaction), no formic acid was formed. Among the metal
salts examined, vanadyl sulfate showed the highest yield of
formic acid, followed by Fe, Cu, and Cr salts (Figure 1). The
yield of formic acid reached 53%, which is better than the best
yield obtained for the conversion of glucose with the
H5PV2Mo10O40 catalyst.[12] Thus, simple VOSO4 is a highly effi-
cient catalyst for the oxidative conversion of glucose into
formic acid in water under an O2 atmosphere.
If the reaction was performed under N2 instead of O2, we
found that lactic acid was formed as a major product instead
of formic acid in the presence of some metal salt catalysts
(Figure 1). Interestingly, VOSO4 also exhibited the highest yield
VOSO4 was also efficient for the transformation of cellulose
into formic or lactic acid. For the conversions of ball-milled
and microcrystalline cellulose at 433 and 453 K under O2,
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 2014, 7, 1557 – 1567 1558

CHEMSUSCHEM
FULL PAPERS
www.chemsuschem.org
Table 1. Catalytic performance of VOSO4 for the conversion of cellulose
and other biomass under aerobic and anaerobic conditions.[a]
Entry
Substrate
Yield[b] [%]
FA
Glu.
Fru.
LA
HMF
1[c]
2[d]
3[e]
4[f]
5[g]
6[h]
7[g]
8[h]
cellulose (ball-milled)
cellulose (ball-milled)
cellulose (microcrystalline)
cellulose (microcrystalline)
inulin
inulin
starch
starch
1.1
2.1
0
1.7
0
1.9
0.4
4.6
0
1.4
0
0.7
0
1.9
0
39
3.8
39
4.9
39
2.2
46
0
54
0
24
0
54
0
46
0
11
0
2.6
0
9.5
0
11
5.0
3.0
[a] Reaction conditions: glucose or fructose units in biomass, 1.0 mmol;
water, 20 cm3. [b] Glu., Fru., FA, and LA denote glucose, fructose, formic
acid, and lactic acid, respectively. [c] O2, 2 MPa; VOSO4, 0.10 mmol; tem-
perature, 433 K; time, 2 h. [d] N2, 2 MPa; VOSO4, 0.050 mmol; tempera-
ture, 453 K; time, 2 h. [e] O2, 2 MPa; VOSO4, 0.20 mmol; temperature,
453 K; time, 2 h. [f] N2, 2 MPa; VOSO4, 0.050 mmol; temperature, 453 K;
time, 4 h. [g] O2, 2 MPa; VOSO4, 0.10 mmol; temperature, 413 K; time,
1.5 h. [h] N2, 2 MPa; VOSO4, 0.050 mmol; temperature, 433 K; time, 1.5 h.
Figure 2. Effect of the fraction of O2 in the reaction atmosphere on the prod-
uct distribution in the conversion of glucose catalyzed by VOSO4. Reaction
conditions: glucose, 1.0 mmol; catalyst, 0.10 mmol; temperature, 423 K; re-
action time, 1 h; total pressure, 2 MPa.
major byproducts became fructose and HMF, and no CO2 was
formed. The other byproducts in the absence of O2 were glyc-
eraldehyde, 1,3-dihydroxyacetone, and pyruvaldehyde (denot-
ed Others I). If O2 was added to the reaction system, the yield
to Others I decreased, and the byproducts that included glyc-
eric acid, glycolic acid, oxalic acid, and acetic acid (denoted
Others II) were detected.
a formic acid yield of 39% was obtained (Table 1, entries 1 and
3). Under a N2 atmosphere at 453 K, lactic acid in yields of 54
and 24% was obtained for the conversions of ball-milled and
microcrystalline celluloses, respectively (Table 1, entries 2 and
4). Our catalytic system was also applicable to the conversion
of other biomass such as starch and inulin into formic or lactic
acid. For example, inulin, which is also an inedible polysaccha-
ride composed of fructose units with terminal glucose units
and exits in many plants such as jerusalem artichoke,[20] can be
transformed into formic and lactic acid with yields of 39 and
54% after reaction for 1.5 h under O2 at 413 K and under N2 at
433 K, respectively (Table 1, entries 5 and 6).
The analysis of the time course for the conversion of glucose
catalyzed by VOSO4 can provide insights into the reaction
pathways that lead to the formation of different products
under aerobic and anaerobic conditions. Our results showed
that the conversion of glucose increased almost linearly in the
initial 40 min under both aerobic and anaerobic conditions
(Figures S3a and S4a). It is of interest that, in both cases, fruc-
tose is formed as a major product at the initial stage (Figur-
es S3b and S4b). If the reaction time was prolonged, the selec-
tivity to fructose decreased and that to formic acid under O2 or
to lactic acid under N2 increased. These observations suggest
that the isomerization of glucose to fructose occurs in the first
step, and formic acid or lactic acid is formed by the consecu-
tive conversion of fructose. The selectivity to HMF increased
with reaction time under N2 (Figure S4b), which confirms that
HMF was a secondary product via fructose (Scheme S1). Glycer-
aldehyde and 1,3-dihydroxyacetone were also formed under
N2, the selectivities of which first increased and then decreased
with increasing reaction time (Figure S4b), which implies that
these trioses were reaction intermediates during the formation
of lactic acid. Similarly, our result suggests that glyceric and
glycolic acid are intermediates during the formation of formic
acid under O2 (Figure S3b).
Acidity is typically required for the hydrolysis of cellulose
and other biomass to monosaccharides (glucose or fructose),
which are expected to undergo further conversions into formic
and lactic acid. The pH value of the aqueous solution that con-
tains 0.10 mmol VOSO4 in 20 cm3 H2O was approximately 3.7–
3.8 at ambient temperature. In addition, it is known that H3O+
can be generated reversibly in hot water because of the in-
creased ionization constant of water at high temperatures.[6b,21]
We speculate that both may contribute to the hydrolysis of
cellulose or other biomass to monosaccharides. Furthermore,
the product, that is, formic or lactic acid, may autocatalyze the
hydrolysis of the biomass.[19]
Reaction pathways under aerobic and anaerobic conditions
To further clarify the role of the reaction atmosphere to deter-
mine the reaction route, we investigated the effect of the frac-
tion of O2 in the atmosphere on the product distribution in
glucose conversion catalyzed by VOSO4 at 423 K. A decrease in
the fraction of O2 leads to the gradual decrease of the yield of
formic acid and increases that of lactic acid (Figure 2). The
major product gradually shifts from formic acid to lactic acid
with a decrease in O2 fraction. CO2 is a major byproduct at
a higher fraction of O2 (Figure 2). The yield of CO2 decreased
upon decreasing the fraction of O2. In the absence of O2, the
Based on these results, we propose a scheme for the conver-
sion of glucose into formic and lactic acid catalyzed by VOSO4
(Scheme 1). Fructose, formed by the isomerization of glucose
in the first step, undergoes retro-aldol fragmentation in the
second step to form two C3 intermediates, that is, glyceralde-
hyde and 1,3-dihydroxyacetone. Lactic acid is formed by the
isomerization of the trioses under anaerobic conditions. How-
ever, the oxidative conversions of the trioses provide formic
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 2014, 7, 1557 – 1567 1559

CHEMSUSCHEM
FULL PAPERS
www.chemsuschem.org
was a key intermediate. If glyceraldehyde and 1,3-dihydroxya-
cetone, two intermediates formed by the retro-aldol fragmen-
tation of fructose, were used as the reactants, lactic acid yields
were quite low in the absence of VOSO4, although the conver-
sions were high. Unknown polymeric products and humins
were formed. The presence of the VOSO4 catalyst increased
the lactic acid yield to >90% for both glyceraldehyde and 1,3-
dihydroxyacetone. In short, VOSO4 plays a significant catalytic
role in the isomerization of glucose, the retro-aldol fragmenta-
tion of fructose, and the conversion of the formed trioses into
lactic acid under anaerobic conditions.
Under an O2 atmosphere, the conversion of glucose in the
absence of VOSO4 was low (2.3%) at 413 K (Table 3). Although
Scheme 1. Reaction scheme for the conversion of glucose in water into
formic and lactic acid under O2 and N2 atmospheres catalyzed by VOSO4.
Table 3. Conversion of possible intermediates catalyzed by VOSO4 under
aerobic conditions.[a]
acid and CO2 as the major products via glyceric and glycolic
acid under aerobic conditions. As we have not observed the
formation of any C4 or C5 intermediates (Figure S5), we specu-
late that the selective cleavage of the C3ÀC4 bond in fructose
occurs as a major pathway in our system.
Substrate[b]
Catalyst[c]
Conversion
[%]
Yield[d] [%]
FA
0.1
5.2
45
44
11
14
48
35
GA1
GA2
CO2
glucose
–
2.3
55
>99
>99
88
88
>99
>99
0
0
0.3
0
0.9
14
43
53
14
24
27
37
fructose
glucose
fructose
glyceraldehyde
1,3-DHA
–
VO2+
VO2+
–
0.5
0.7
0
1.3
3.1
0
Conversions of possible intermediates under aerobic and
anaerobic conditions
–
0
0
glyceraldehyde
1,3-DHA
VO2+
VO2+
1.3
0
8.4
5.5
To understand the catalytic functions of VOSO4 in the forma-
tion of lactic acid under anaerobic conditions in depth, we per-
formed the conversions of several possible intermediates
under a N2 atmosphere. In the absence of VOSO4, neither lactic
acid nor C3 intermediates were formed from glucose under N2
(Table 2). The conversion of fructose without VOSO4 was higher
than that of glucose to produce HMF as a major product along
with minor amounts of C3 products. This indicates that the de-
hydration of fructose proceeds preferentially. The presence of
VOSO4 enhanced the conversions of both fructose and glucose
significantly to 96% with high yields of lactic acid (56 and
58%, respectively) at 433 K. The product distribution for the
conversion of fructose was almost the same as that for the
conversion of glucose, which further confirms that fructose
[a] Reaction conditions: substrate, 6.0 mmol based on carbon; H2O,
20 cm3; O2, 2 MPa; temperature, 413 K; reaction time, 1.0 h. [b] 1,3-DHA
denotes 1,3-dihydroxyacetone. [c] None: without VOSO4; VO2+: VOSO4,
0.10 mmol. [d] FA, GA1 and GA2 denote formic acid, glyceric acid, and
glycolic acid, respectively.
the noncatalytic conversion of fructose was relatively high
(55%), the reaction was unselective and a number of products,
which included glyceraldehyde, 1,3-dihydroxyacetone, glyceric
acid, pyruvaldehyde, pyruvic acid, acetic acid, formaldehyde,
formic acid, and CO2, were obtained. In the presence of VOSO4,
the conversions of glucose and fructose were both accelerated
to >99%, and formic acid and CO2 were produced as the
major products (Table 3). The minor products were glyceric
and glycolic acid. The presence of VOSO4 also enhanced the
oxidative conversion of the C3 intermediates, that is, glyceral-
dehyde and 1,3-dihydroxyacetone, and increased the selectivi-
ties to formic acid and CO2. These observations allow us to
speculate that the formations of formic acid and CO2 also pro-
ceed via glyceraldehyde and 1,3-dihydroxyacetone.
Table 2. Conversion of possible intermediates catalyzed by VOSO4 under
anaerobic conditions.[a]
Substrate[b]
Catalyst[c] Conversion
[%]
Yield[d] [%]
GA 1,3-DHA HMF
LA
0
glucose
fructose
glucose
fructose
glyceraldehyde
1,3-DHA
glyceraldehyde VO2+
–
–
26
56
0
0
11
25
13
13
0
0
0
0
6.9 3.1 1.6
VO2+
VO2+
–
96
96
99
56
58
11
3.3 2.5
2.9 3.1
Oxidative cleavage of CÀC bonds in model compounds
–
9.1
–
–
88
5.2 8.3
under aerobic conditions
>99
>99
91
92
–
0
1.3
–
VO2+
Our present work has demonstrated that the simple VOSO4
can catalyze the cleavage of the CÀC bonds in carbohydrate
molecules to produce formic acid and CO2 selectively under
aerobic conditions. It is of high significance to understand the
factors that control the cleavage of CÀC bonds in carbohydrate
molecules as such information is very limited. Thus, we have
1,3-DHA
[a] Reaction conditions: substrate, 6.0 mmol based on carbon; H2O,
20 cm3; N2, 2 MPa; temperature, 433 K; reaction time, 1.5 h. [b] 1,3-DHA
denotes 1,3-dihydroxyacetone. [c] None: without VOSO4; VO2+: VOSO4,
0.05 mmol. [d] LA and GA denote lactic acid and glyceraldehyde, respec-
tively.
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 2014, 7, 1557 – 1567 1560

CHEMSUSCHEM
FULL PAPERS
www.chemsuschem.org
studied the oxidative conversion behaviors of 12 model com-
pounds comprehensively using the present catalytic system.
First, we checked the stability of formic acid under our reac-
tion conditions in the presence of VOSO4. No conversion of
formic acid was observed after 1 h reaction at 413 K (Table 4).
This clarifies that formic acid is stable and does not undergo
acid, 2-bromo-2-phenylacetic acid, or a-amino acid, in which
the a-carbon atom of the carboxylic group is linked by a OH,
Br, or amino group, respectively, can occur by CÀC cleavage
catalyzed by V.[22]
Besides these C2 carboxylic acids, we further investigated
the oxidative conversions of some other model compounds.
The conversion of glyoxal provided formic acid as the major
product with 80% selectivity (Table 4, entry 6). Thus, the CÀC
bond that links two aldehydic groups can be cleaved easily in
the presence of VOSO4 under O2, and the aldehydic groups are
oxidized into carboxylic groups [Scheme 2, Eq. (6)]. CO2 was
also formed with a selectivity of ~12%. It has already been
demonstrated that formic acid cannot be oxidized to CO2 in
our system [Scheme 2, Eq. (1)]. We propose that a small part of
glyoxal may be first converted to glyoxylic acid [Scheme 2,
Eq. (6), structure in the square brackets], which then under-
goes decarboxylation to CO2 and formic acid. The oxidative
conversion of glycolal dimer was performed as glycolal itself is
too reactive to be stable (Table 4, entry 7). As expected, the
main products were formic acid and HCHO, which suggests
that the CÀC bond that links an aldehyde group and a carbon
atom joined to a OH group could also be cleaved [Scheme 2,
Eq. (7)]. However, although acetaldehyde could be converted
in our system, acetic acid was the major product, and the
cleavage of the CÀC bond in acetaldehyde did not proceed.
Thus, only a CÀC bond with an O-containing group joined to
each end can be cleaved. Notably, the conversion of ethylene
glycol under our reaction conditions was low (10%) (Table 4,
entry 9), which indicates that the direct cleavage of CÀC bond
only joined to hydroxyl groups in polyols was difficult. Further
analysis of the oxidative conversions of lactic acid, pyruvalde-
hyde, and 3-hydroxy-2-butanone [Scheme 2, Eqs. (10)–(12)]
suggests again that the CÀC bond that links a carboxylic (or al-
dehyde or ketone) group and a carbon atom joined to an O-
containing group can be cleaved easily, and the group at each
side is oxidized.
Table 4. Conversion of model compounds catalyzed by VOSO4 under
aerobic conditions.[a]
Entry Substrate
Conversion
[%]
Yield[b] [%]
GA AA FAld CO2
FA
–
2.5
2.1
24
0.2
80
1
2
3
formic acid
formaldehyde
oxalic acid
0
4.6
70
91
2.0
>99
>99
83
10
98
–
0
0
0
–
0
0
–
0
0
–
2.1
65
45
1.1
0.5 12
34 3.6
4
5
glycolic acid
acetic acid
0.4 17
0
0
0
0
6
glyoxal
1.6
1.1
0
7
8
9
10
11
12
glycolal dimer
acetaldehyde
ethylene glycol
lactic acid
pyruvaldehyde
3-hydroxy-2-butanone >99
53
0
0
74
0
0
1.0
5.6
1.6
4.9 0.9
7.5
30
3.0
0
0
0
46
63
70
7.4 24
2.3
3.9
>99
2.0
2.1
[a] Reaction conditions: substrate, 6.0 mmol based on carbon; VOSO4,
0.010 mmol; water, 20 cm3; temperature, 413 K; O2 pressure, 2 MPa; time,
1.0 h. [b] FA, GA, AA, and FAld denote formic acid, glycolic acid, acetic
acid, and formaldehyde, respectively.
consecutive oxidation to CO2 under our reaction conditions.
The oxidation of HCHO could take place, which produced
formic acid and CO2 with selectivities of 54 and 46%, respec-
tively. However, the conversion of HCHO under our reaction
conditions was low (4.6%), which suggests that the oxidation
of HCHO is not the main pathway for the formation of formic
acid and CO2.
Then, we studied the oxidative conversions of some C2
model compounds. The conversion of oxalic acid proceeded
easily under our reaction conditions to produce CO2 with a se-
lectivity of 93% (Table 4, entry 3). This suggests that the CÀC
bond, which links two carboxylic groups, can be broken to
form CO2 [Scheme 2, Eq. (3)]. The conversion of glycolic acid, in
which the a-carbon atom of the carboxylic group is joined to
an OH group, provided formic acid, HCHO, and CO2 (Table 4,
entry 4). The selectivity to CO2 was close to the sum of the se-
lectivities to formic acid and HCHO. This reveals that the CÀC
bond in glycolic acid can be cleaved, and the carboxylic group
is transformed to CO2 [Scheme 2, Eq. (4)]. The remaining
moiety after decarboxylation is expected to be oxidized into
HCHO. As formic acid and HCHO have actually been formed
and the consecutive oxidation of HCHO to formic acid is
a slow reaction (Table 4, entry 2), we speculate that some of
the glycolic acid may be first oxidized into glyoxylic acid
[Scheme 2, Eq. (4), structure in the square brackets], which fur-
ther undergoes oxidative decarboxylation to CO2 and formic
acid. The conversion of acetic acid was very low (2.0%). Thus,
the decarboxylation of acetic acid almost did not proceed in
our system. Actually, the oxidative decarboxylation of mandelic
Based on these studies on model molecules, we have ob-
tained the following insights into the oxidative cleavage of the
CÀC bond and the formation of products catalyzed by VOSO4
under aerobic conditions.
(1) Formic acid is stable under our reaction conditions and
does not undergo consecutive oxidation to CO2. The oxidation
of HCHO under our reaction conditions proceeds slowly.
(2) The decarboxylation of carboxylic acid intermediates
causes the formation of CO2. The presence of an O-containing
group (ÀOH or =O) at the a-carbon atom of the carboxylic
acid is required for the decarboxylation. The remaining moiety
is oxidized accompanied by the formation of CO2.
(3) The CÀC bond that links an aldehyde (or ketone) group
and a C atom joined to an O-containing group (ÀOH or =O)
can be cleaved easily. The aldehydic (or ketonic) group is trans-
formed into formic acid (or another carboxylic group), and the
remaining moiety is oxidized.
(4) The direct cleavage of a CÀC bond only joined to hydrox-
yl groups in polyols is difficult.
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 2014, 7, 1557 – 1567 1561

CHEMSUSCHEM
FULL PAPERS
www.chemsuschem.org
HCHO or oxidized to glyoxylic acid, which is further
converted to formic acid and CO2. There are also two
possible pathways for the conversion of 1,3-dihydrox-
yacetone (Scheme 3b). One is the direct oxidative
CÀC cleavage to form HCHO and glycolic acid, and
the other is the oxidation to hydroxypyruvaldehyde,
which can further undergo oxidative CÀC cleavage to
form glycolic acid and formic acid. Finally, formic
acid, CO2, and HCHO are produced from glycolic acid
as described above. During the conversion of glucose
or fructose, formic acid and CO2 were formed as the
major products. This suggests that the pathway via
glyoxylic acid is probably a major route.
Strategy to increase formic acid yield under aero-
bic conditions
The yield of formic acid is limited to ~50% because
of the formation of CO2 not only for VOSO4 but also
for the H5PV2MoO10O40 catalyst reported previous-
ly.[12–14] It would be highly attractive to find a strategy
to suppress the formation of CO2 and to increase the
selectivity of formic acid. From the above mechanistic
studies, we can expect that the suppression of the
formation of carboxylic acid intermediates by pro-
tecting the aldehydic group may increase the selec-
tivity to formic acid. It is known that the presence of
an alcohol can transform reactive aldehydic groups
into hemiacetal intermediates.[23] If we added metha-
nol or ethanol into the reaction system, we succeed-
ed in the suppression of the formation of CO2 during
the conversions of glucose and cellulose under aero-
bic conditions (Table 5). The yield of formic acid in-
creased to 70–75%. HCHO was also formed with
a yield of ~10%. We performed control reactions in
the absence of glucose or cellulose. The conversion
of methanol or ethanol catalyzed by VOSO4 under an
O2 atmosphere was negligible under our reaction
conditions. In other words, formic acid could not be
formed from the direct oxidation of methanol or eth-
anol. To the best of our knowledge, this is the first
report to describe the highly selective production of
formic acid from cellulose or cellulose-derived carbo-
hydrates without the significant formation of CO2.
Scheme 2. Key reactions involved in the conversion of model C2 and C3 compounds cata-
lyzed by VOSO4 in the presence of O2. The number in parenthesis represents product se-
lectivity, and some of the reactant may be transformed into the structure in brackets
before the CÀC bond was cleaved. The numbering of the reaction corresponds with the
entry number in Table 4.
Mechanism for the formation of formic acid and CO2
Characterization of V species
Based on the insights gained from the studies using model
molecules, we propose a possible mechanism for the conver-
sions of glyceraldehyde and 1,3-dihydroxyacetone, two major
C3 intermediates, into formic acid and CO2 (Scheme 3). Glycer-
aldehyde may undergo oxidative CÀC cleavage to formic acid
and glycolal or may be oxidized to glyceric acid, which has
been observed as an intermediate under an O2 atmosphere
(Scheme 3a). The decarboxylation of glyceric acid produces
CO2 and glycolal. Glycolal is a highly reactive molecule, which
probably undergoes fast oxidation to glycolic acid, a detected
intermediate. Glycolic acid may be decarboxylated to CO2 and
The electron spin resonance (ESR) spectrum of VOSO4 in an
aqueous solution that contained glucose showed eight distinct
peaks with a g value of 1.97 (Figure 3A). This is consistent with
the hyperfine coupling of the unpaired electron (3d1) in VO2+
with the 51V nucleus (I=7/2).[24] After reaction under a N2 at-
mosphere at 433 K for 1.5 h, the spectrum of the mixture did
not change significantly. In other words, VO2+ remained during
the reaction, and no other V species were detectable. We spec-
ulate that VO2+ is the active species for the conversion of glu-
cose into lactic acid. However, we cannot completely rule out
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 2014, 7, 1557 – 1567 1562

CHEMSUSCHEM
FULL PAPERS
www.chemsuschem.org
Figure 3. ESR spectra of V species in an aqueous solution that contains glu-
cose (V/glucose=1:10). VOSO4 system (left): (a) before reaction, (b) after re-
action under N2 at 433 K for 1.5 h, (c) after reaction under O2 at 413 K for
1 h, (d) after reaction under O2 at 413 K for 12 h. V2O5 and NH4VO3 systems
(right): (a) V2O5 before reaction, (b) V2O5 after reaction under N2 at 433 K for
1.5 h, (c) V2O5 after reaction under O2 at 413 K for 1 h, (d) V2O5 after reaction
under O2 at 413 K for 12 h, (e) NH4VO3 before reaction, (f) NH4VO3 after reac-
tion under N2 at 433 K for 1.5 h, (g) NH4VO3 after reaction under O2 at 413 K
for 1 h, (h) NH4VO3 after reaction under O2 at 413 K for 12 h. See Figure 1 for
other reaction conditions.
Although the yield of lactic acid was lower, V2O5 and NH3VO3
were also effective for the conversion of glucose into lactic
acid under N2 (Figure S1, Supporting Information). If these two
compounds were used as the catalyst precursors, almost no
ESR signals were observed before reaction (Figure 3B, spectra a
and e). This is in agreement with the fact that VV in these com-
pounds is ESR silent. After the reaction at 433 K for 1.5 h, the
ESR signal attributed to VO2+ could be observed if either V2O5
or NH4VO3 was used as the catalyst precursor. This supports
the idea that VO2+ is a key species for the conversion of glu-
cose into lactic acid under a N2 atmosphere. After the conver-
sion of glucose in the presence of V2O5 or NH4VO3 under an O2
atmosphere at 413 K for 1 h, we also observed the ESR signal
ascribed to VO2+ species. The VO2+ signal became significantly
weakened or disappeared after a longer reaction time under
O2. This change is similar to that observed for VOSO4.
Scheme 3. Possible mechanism for the formation of formic acid and CO2
from glyceraldehyde (a) and 1,3-dihydroxyacetone (b) catalyzed by VOSO4
under aerobic conditions.
Table 5. Effect of the addition of methanol or ethanol on the catalytic
performance of VOSO4 for the conversion of glucose and cellulose under
aerobic conditions.[a]
Additive
Reactant
Conversion
[%]
Yield[b] [%]
FA
GA
AA
FAld
CO2
none
glucose
glucose
–
glucose
cellulose[d]
–
100
95
–
91
92
–
41
75
0
73
70
0
0.2
2.8
0
0.5
0.2
0
0.3
0
0
0.5
0
0
0.6
11
tr.[c]
9.7
9.5
0
51
2.5
0
2.1
2.9
0
CH3OH
CH3OH
C2H5OH
C2H5OH
C2H5OH
It is known that 51V NMR spectroscopy can be used to study
VV (3d0) species,[25] whereas VIV (3d1) species are NMR silent as
a result of considerable line broadening caused by the pres-
ence of a paramagnetic center in the neighborhood of the
nuclei.[26] Our 51V NMR measurement for the VOSO4 aqueous
solution after oxidation under an O2 atmosphere at 413 K for
0.5 h showed one distinct signal with a chemical shift at d=
À542.9 ppm (Figure 4A, spectrum a), which was attributable to
[a] Reaction conditions: glucose, 1.0 or 0 mmol; VOSO4, 0.10 mmol; H2O,
15 cm3; alcohol, 5 cm3; O2, 2 MPa; temperature, 413 K; reaction time,
3.0 h. [b] FA, GA, AA, and FAld denote formic acid, glycolic acid, acetic
acid, and HCHO, respectively. [c] A trace amount of HCHO was formed
from CH3OH. [d] Ball-milled cellulose based on glucose units, 1.0 mmol;
catalyst, 0.1 mmol; H2O, 15 cm3; alcohol, 5 cm3; O2, 2 MPa; temperature,
433 K; reaction time, 5.0 h.
+
the possibility that another highly active species with a low
concentration may work as the true catalyst. After the reaction
under an O2 atmosphere at 413 K for 1 h, VO2+ was again ob-
served in the ESR spectrum of the mixture (Figure 3A, spec-
trum c). However, if the reaction was performed under O2 for
a longer time (12 h) and no reductive reactant and intermedi-
ates remained, the ESR signal for the VO2+ species disap-
peared, which indicates the oxidation of VIV to VV under oxida-
tive circumstance as VV (d0) is ESR silent.
VO2 species in aqueous solution.[25] This indicates that the oxi-
+
dation of VO2+ to VO2 occurs under an O2 atmosphere at
+
413 K. The addition of glucose into the VO2 aqueous solution
caused a decrease in the 51V NMR signal even at 298 K for 24 h
+
(Figure 4A, spectrum b). The 51V NMR signal of VO2 cations
disappeared after the addition of glucose followed by heating
the mixture at 393 K for 10 min (Figure 4A, spectrum c). Mean-
while, an ESR signal attributable to VO2+ appeared (Figure 4B).
+
These observations suggest that the VO2 can be reduced
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 2014, 7, 1557 – 1567 1563

CHEMSUSCHEM
FULL PAPERS
www.chemsuschem.org
lyst is efficient for CÀC bond cleavage but not good at isomeri-
zation. Secondly, the higher reactivity of fructose than glucose
in our catalytic system may also be responsible for the prefer-
ential occurrence of the cleavage of CÀC bond in fructose but
not in glucose (Tables 2 and 3). Furthermore, for the W-based
catalytic system,[7] a significantly higher reaction temperature
(518 K) was employed than that in our system (433 K). Possibly,
the energy barrier for the C2ÀC3 bond cleavage in glucose is
higher than that for the C3ÀC4 bond cleavage in fructose.
Our previous study uncovered that the CÀC bond cleavage
of fructose is triggered by proton transfer from C4ÀOH to
C=O.[19] The coordination of a Lewis acid such as PbII to the O
atoms of fructose increases the positive charge on the H atom
of C4ÀOH.[19] This would enhance the acidity of the C4ÀOH frag-
ment and facilitate the proton transfer and the CÀC bond
cleavage of fructose to trioses. We speculate that VO2+ is also
a better Lewis acid to be coordinated by the O atoms in fruc-
tose, which contributes to its outstanding ability in the retro-
aldol fragmentation.
Figure 4. 51V NMR (left) and ESR (right) spectra of V species in aqueous solu-
tion after (a) oxidation of VOSO4 under O2 at 413 K for 0.5 h,( b) addition of
glucose to (a) (V/glucose=1:10) and keeping the mixture at 298 K for 24 h,
(c) addition of glucose to (a) (V/glucose=1:10) and keeping the mixture
under N2 at 393 K for 10 min, and d) addition of glucose to (a) (V/glu-
cose=1:10) and keeping the mixture at 413 K for 1 h under O2.
easily into VO2+ by glucose at our reaction temperatures. The
Our present work has further demonstrated that VO2+ can
catalyze the formation of formic acid under an O2 atmosphere.
Without VO2+, the conversion of glucose was quite low (2.3%)
under our aerobic reaction conditions (Table 3). Although the
conversion of fructose reached 55% under noncatalytic condi-
tions (Table 3), the reaction did not result in the selective for-
mation of specific products. The presence of VO2+ not only in-
creased the conversion of either glucose or fructose to >99%
but also led to the selective formation of formic acid together
with CO2. We have clarified that formic acid is very stable and
does not undergo decomposition or consecutive oxidation
under our reaction conditions (Table 4). The oxidative CÀC
cleavage of C3 intermediates and subsequent C2 intermediates
results in the formation of formic acid along with CO2. Our de-
tailed studies using various model molecules have uncovered
that the decarboxylation of carboxylic acid intermediates with
an O-containing group at the a-carbon atom occurs easily in
our catalytic system to produce CO2 and an oxidized moiety
(Scheme 2). The CÀC bond that links an aldehyde group and
a carbon atom joined to an O-containing group can also be
cleaved easily to produce formic acid and an oxidized moiety.
An electron-transfer and oxygen-transfer (ET-OT) mechanism
has been proposed for polyoxometalate H5PV2Mo10O40-cata-
lyzed selective oxidations in the liquid phase.[30] In this mecha-
nism, the two VV centers in H5PV2Mo10O40 accept two electrons
from the substrate to be reduced to VIV and donate one O
atom at the same time. For some heterogeneous V-catalyzed
oxidation reactions, it is proposed that the redox conversion of
VV/VIII may participate in the catalytic cycle, in which one VV
center could accept two electrons from the substrate.[31] Our
ESR and 51V NMR spectra suggest that the redox conversion
+
redox conversion of VO2 /VO2+ has also been observed in
other systems.[27–29] Therefore, we propose that the redox reac-
+
tion between VO2 and VO2+ is responsible for the formation
of formic acid under an O2 atmosphere.
Possible mechanism of the V catalyst
Our characterization results reveal that VO2+ is a key species
during the conversion of glucose to lactic acid under a N2 at-
mosphere. The comparative studies for the conversions of pos-
sible intermediates in the absence and presence of the VOSO4
catalyst under a N2 atmosphere suggest that the VO2+ species
catalyzes a series of tandem steps, which include the isomeri-
zation of glucose to fructose, the retro-aldol fragmentation of
fructose into trioses, and the conversion of trioses into lactic
acid (Scheme 1). In particular, we found that in the absence of
VO2+ HMF was formed as a major product instead of lactic
acid during the conversion of either fructose or glucose
(Table 2). This indicates that the dehydration of fructose to
HMF is a major route without a catalyst. The presence of VO2+
not only increases the conversion of fructose or glucose but
also shifts the major product to lactic acid. This suggests clear-
ly that VO2+ plays a crucial role in the retro-aldol fragmentation
of fructose.
The cleavage of the CÀC bond (C2ÀC3) in glucose, which
leads to the simultaneous formation of C2 and C4 compounds,
did not proceed as a major pathway in our system, which
might be because of several possibilities. The first reason may
be the higher catalytic ability of our VO2+ catalyst in the iso-
merization reaction. In our system, VO2+ may function as
a Lewis acid, which can catalyze the isomerization of glucose
to fructose rapidly. Taarning et al. and Sels et al. found a similar
cleavage of the C3ÀC4 bond in fructose during the conversion
of glucose in the presence of Lewis acidic SnIV-based cata-
lysts.[15,16] However, Zhang and co-workers proposed that a W-
based catalyst mainly catalyzed the cleavage of the C2ÀC3
bond in glucose.[7] This is possibly because the W-based cata-
+
between VO2 and VO2+ occurs during the reactions under
aerobic conditions. This VV/VIV redox couple may participate in
the oxidative cleavage of CÀC bonds to form formic acid along
with CO2 as the major product. We speculate that such oxida-
tive CÀC cleavage may also follow the ET-OT mechanism sug-
gested for the H5PV2Mo10O40 catalyst. If we consider that two
electrons should be transferred at the same time, we propose
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 2014, 7, 1557 – 1567 1564

CHEMSUSCHEM
FULL PAPERS
www.chemsuschem.org
Scheme 4. Proposed ET-OT mechanism for the oxidative CÀC cleavage for three possible reaction intermediates.
tentatively that two VV species are required for the oxidative
CÀC cleavage in our case. Moreover, our reaction path studies
suggest that the reaction proceeds via C3 and C2 intermediates
such as glyceraldehyde, 1,3-dihydroxyacetone, glyceric acid,
glyoxylic acid, and glycolic acid (Scheme 3). The possible reac-
tion mechanism that we have in mind for the oxidative CÀC
cleavage of three typical intermediates is displayed in
Scheme 4. If we take glyceraldehyde as an example, we specu-
late that one glyceraldehyde molecule may interact with two
plays a key catalytic role in the formation of lactic acid by ac-
celerating the isomerization of glucose, the retro-aldol frag-
mentation, and the further isomerization of the trioses. On the
other hand, the oxidative CÀC cleavage of the C3 and subse-
quent C2 intermediates under an O2 atmosphere produces
formic acid, a stable product under our reaction conditions,
along with CO2. The CÀC bond that links an aldehyde group
and a carbon atom joined to an O-containing group (ÀOH, =O)
can be cleaved easily to form formic acid, whereas CO2 is
mainly produced by the decarboxylation of the carboxylic acid
intermediates in which the a-carbon atom is joined to an O-
containing group. We suppressed the formation of CO2 and
enhanced the yield of formic acid to 70–75% by the addition
of an alcohol such as methanol or ethanol. We propose that
the redox conversion of VO2+/VO2+ plays a crucial role in the
oxidative CÀC cleavage reactions.
+
VO2 species or its hydrolyzed form (H2VO3+) in H2O[27] to form
an intermediate of binuclear VV coordinated by glyceraldehyde.
Each VV center accepts one electron from the VÀO bond con-
nected to the CÀO bond to be reduced to VIV. This drives the
formation of two C=O bonds and the simultaneous cleavage
of the CÀC bond to form glycolal and formic acid. The process
involves the transfer of two electrons and one O atom. The re-
+
duced VO2+ can be reoxidized into VO2 in the presence of O2
at the temperature used in our reactions as indicated by ESR
and 51V NMR studies. Glyoxylic or glycolic acid would undergo
similar oxidative CÀC cleavage to form formic acid and CO2 or
HCHO and CO2.
Experimental Section
Materials
The chemicals used primarily in this work were as follows: d-glu-
cose (99.5%, Alfa-Aesar), d-fructose (99.0%, Alfa-Aesar), 1,3-dihy-
droxyacetone (98.0%, J&K Chemical), glyceraldehyde (90.0%,
Sigma Aldrich), dl-lactic acid (85.0%, TCI), formic acid (99.0%, Sino-
pharm), and VOSO4·xH2O (99%, H2O 24.4 wt%, Strem Chemicals).
The other chemicals used as substrates (which include glyceric
acid, pyruvaldehyde, pyruvic acid, glycolic acid, acetic acid, acetal-
dehyde, oxalic acid, glyoxal, glycolal dimer, glyoxylic acid, ethylene
glycol, and glycerol) were purchased from Sinopharm, Alfa-Aesar,
or TCI and their purity was higher than 98%. The other metal salts
used as catalyst candidates (which include VOC2O4, NH4VO3, V2O5,
KCl, CaCl2, CrCl3, MnSO4, FeCl3, CoCl2, NiCl2, CuCl2, ZnCl2, Al2(SO4)3,
ZrOCl2, In(NO3)3, SnCl2, Pb(NO3)2, and Bi(NO3)3) were purchased
from Sinopharm and their purity was higher than 99%. VOH-
PO4·0.5H2O was prepared according to the reported procedure.[32]
Microcrystalline cellulose was purchased from Alfa-Aesar. Ball-
milled cellulose was obtained by ball-milling the microcrystalline
Conclusions
We have demonstrated that simple VOSO4 is an efficient ho-
mogeneous catalyst for the transformations of cellulose and its
monomer (glucose) into lactic or formic acid depending on the
reaction atmosphere. We have discovered that the major prod-
uct can be tuned by changing the atmosphere; a decrease in
the fraction of O2 in the atmosphere decreases the yield of
formic acid and increases that of lactic acid. We have clarified
that the reaction proceeds through the isomerization of glu-
cose to fructose, which subsequently undergoes retro-aldol
fragmentation to two trioses, that is, glyceraldehyde and 1,3-
dihydroxyacetone. The isomerization of these trioses under N2
leads to the formation of lactic acid. The VO2+ cation in water
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 2014, 7, 1557 – 1567 1565

CHEMSUSCHEM
FULL PAPERS
www.chemsuschem.org
cellulose at 60 rpm by using a SFM-1 Desk-top planetary ball miller
(MTI Co.) for 48 h.
the
Doctorial
Program
of
Higher
Education
(No.
20130121130001).
Keywords: biomass · carbohydrates · homogeneous catalysis ·
reaction mechanisms · vanadium
Catalytic reactions
The conversion of glucose, cellulose, possible reaction intermedi-
ates, model compounds, and other biomass was performed by
using a batch-type Teflon-lined stainless-steel autoclave reactor. If
we take the conversion of glucose as an example, the catalyst and
glucose were added to the reactor, which was precharged with
water (20 cm3). After the purge with O2 or N2 and the introduction
of O2 or N2 with a fixed pressure (typically, 2 MPa), the reactor was
placed in an oil bath. When the system reached the reaction tem-
perature, the reaction was started by stirring vigorously. The reac-
tion was terminated quickly by cooling the reactor to RT in cold
water after a fixed time. The liquid products were analyzed by
HPLC (Shimazu LC-20A) with refractive index (RI) detection and
a Shodex SUGARSH-1011 column (8 mmꢁ300 mm) using a dilute
H2SO4 aqueous solution as the mobile phase. The gas-phase prod-
ucts were analyzed by GC equipped with a thermal conductivity
detector (TCD) and a Porapak Q column. We analyzed the amount
of formaldehyde formed in both the liquid and gas phases by
HPLC and GC, respectively. Formaldehyde was detected only in the
liquid phase because of its high solubility in water. The total
amount of organic carbon (TOC) in the reaction solution was deter-
mined by using a Multi N/C3100 TOC-TN analyzer (Analytik Jena,
Germany). The conversion of the substrate and the selectivity to
a specific product were defined as the molar percentage of the
substrate converted and the molar percentage of the specific prod-
uct in the substrate converted, respectively. Both the conversion
and selectivity were calculated on a carbon basis. For example, the
conversion of cellulose was calculated as the carbon molar per-
centage of the organic carbon products formed in the reaction so-
lution detected as TOC and the carbon-containing products
formed in the gas phase quantified by GC in the cellulose added
to the system. The carbon balance was typically better than 90%,
and the lost carbon appeared mainly as humins.
[1] a) A. J. Ragauskas, C. K. Williams, B. H. Davison, G. Britovsek, J. Cairney,
C. A. Eckert, W. J. Ferderick Jr., J. P. Hallett, D. J. Leak, C. L. Liotta, J. R.
Mielenz, R. Murphy, R. Templer, T. Tschaplinski, Science 2006, 311, 484–
489; b) M. Stçcker, Angew. Chem. Int. Ed. 2008, 47, 9200–9211; Angew.
Chem. 2008, 120, 9340–9351.
[2] a) B. Kamm, Angew. Chem. Int. Ed. 2007, 46, 5056–5058; Angew. Chem.
2007, 119, 5146–5149; b) J. C. Serrano-Ruiz, R. Luque, A. Sepulveda-Es-
cribano, Chem. Soc. Rev. 2011, 40, 5266–5281.
[3] a) P. L. Dhepe, A. Fukuoka, ChemSusChem 2008, 1, 969–975; b) S. Van
de Vyver, J. A. Geboers, P. A. Jacobs, B. F. Sels, ChemCatChem 2011, 3,
82–94; c) H. Kobayashi, H. Ohta, A. Fukuoka, Catal. Sci. Technol. 2012, 2,
869–883; d) W. Deng, Q. Zhang, Y. Wang, Dalton Trans. 2012, 41, 9817–
9831; e) P. Gallezot, Chem. Soc. Rev. 2012, 41, 1538–1558; f) H. Kobaya-
shi, K. Fukuoka, Green Chem. 2013, 15, 1740–1763.
[4] a) R. Rinaldi, F. Schꢂth, Energy Environ. Sci. 2009, 2, 610–626; b) K. Shi-
mizu, A. Satsuma, Energy Environ. Sci. 2011, 4, 3140–3153.
[5] J. B. Binder, R. T. Raines, J. Am. Chem. Soc. 2009, 131, 1979–1985.
[6] a) A. Fukuoka, P. L. Dhepe, Angew. Chem. Int. Ed. 2006, 45, 5161–5163;
Angew. Chem. 2006, 118, 5285–5287; b) C. Luo, S. Wang, H. Liu, Angew.
Chem. Int. Ed. 2007, 46, 7636–7639; Angew. Chem. 2007, 119, 7780–
7783; c) S. Van de Vyver, J. Geboers, M. Dusselier, H. Schepers, T. Vosch,
L. Zhang, G. Van Tendeloo, P. A. Jacobs, B. F. Sels, ChemSusChem 2010,
3, 698–701; d) M. Liu, W. Deng, Q. Zhang, Y. Wang, Y. Wang, Chem.
Commun. 2011, 47, 9717–9719; e) A. M. Ruppert, K. Weinberg, R. Palko-
vits, Angew. Chem. Int. Ed. 2012, 51, 2564–2601; Angew. Chem. 2012,
124, 2614–2654.
[7] a) N. Ji, T. Zhang, M. Zheng, A. Wang, H. Wang, X. Wang, J. G. Chen,
Angew. Chem. Int. Ed. 2008, 47, 8510–8513; Angew. Chem. 2008, 120,
8638–8641; b) Y. Liu, C. Luo, H. Liu, Angew. Chem. Int. Ed. 2012, 51,
3249–3253; Angew. Chem. 2012, 124, 3303–3307; c) Z. Tai, J. Zhang, A.
Wang, M. Zheng, T. Zhang, Chem. Commun. 2012, 48, 7052–7054; d) A.
Wang, T. Zhang, Acc. Chem. Res. 2013, 46, 1377–1386.
[8] M. Dusselier, P. Van Wouwe, A. Dewaele, E. Makshina, B. F. Sels, Energy
Environ. Sci. 2013, 6, 1415–1442.
[9] a) R. Datta, M. Henry, J. Chem. Technol. Biotechnol. 2006, 81, 1119–1129;
b) L. Shen, E. Worrell, M. Patel, Biofuels Bioprod. Bioref. 2010, 4, 25–40.
[10] S. Enthaler, J. von Langermann, T. Schemidt, Energy Environ. Sci. 2010, 3,
1207–1217.
[11] a) B. Loges, A. Boddien, H. Junge, M. Beller, Angew. Chem. Int. Ed. 2008,
47, 3962–3965; Angew. Chem. 2008, 120, 4026–4029; b) T. C. Johnson,
D. Morris, M. Wills, Chem. Soc. Rev. 2010, 39, 81–88; c) A. Boddien, D.
Mellmann, F. Gꢃrter, R. Jackstell, H. Junge, P. J. Dyson, G. Laurenczy, R.
Ludwig, M. Beller, Science 2011, 333, 1733–1736; d) Q. Y. Bi, X. L. Du,
Y. M. Liu, Y. Cao, H. H. He, K. N. Fan, J. Am. Chem. Soc. 2012, 134, 8926–
8933.
[12] R. Wçlfel, N. Taccardi, A. Bçsmann, P. Wasserscheid, Green Chem. 2011,
13, 2759–2763.
[13] J. Albert, R. Wçlfel, A. Bçsmann, P. Wasserscheid, Energy Environ. Sci.
2012, 5, 7956–7962.
[14] J. Li, D. J. Ding, L. Deng, Q. X. Guo, Y. Fu, ChemSusChem 2012, 5, 1313–
1318.
[15] M. S. Holm, S. Saravanamurugan, E. Taarning, Science 2010, 328, 602–
605.
Characterization
ESR spectroscopic measurements were performed at RT by using
a Bruker EMX-10/12 ESR spectrometer operated at X-band frequen-
cy. After the O2 was removed, the sample solution was transferred
into a capillary tube with an inner diameter of 0.8 mm. The height
of the solution was typically 20 mm. The parameters for ESR meas-
urements were as follows: microwave frequency, 9.8 GZ; micro-
wave power, 20 mW; modulation frequency; 100 kHz; attenuator,
10 dB. 51V NMR spectra were recorded by using a Bruker AVII 400
NMR spectrometer operated at 105.2 MHz at RT with 0.5 cm3
sample solution in an NMR tube. The chemical shift of V was mea-
sured relative to an external reference, VOCl3. The parameters for
the NMR measurements were as follows: pulse width, 308; spectral
width, 100 kHz; acquisition time, 0.33 s; line broadening, 40 HZ;
frequency domain size, 64 K data points.
[16] F. de Clippel, M. Dusseloer, R. Van Rompaey, P. Vanelderen, J. Dijkmans,
E. Makshina, L. Giebeler, S. Oswald, G. V. Baron, J. F. M. Denayer, P. P. Pes-
carmona, P. A. Jacobs, B. F. Sels, J. Am. Chem. Soc. 2012, 134, 10089–
10101.
Acknowledgements
[17] F. Chambon, F. Rataboul, C. Pinel, A. Cabiac, E. Guillon, N. Essayem,
Appl. Catal. B 2011, 105, 171–181.
[18] F. F. Wang, C. L. Liu, W. S. Dong, Green Chem. 2013, 15, 2091–2095.
[19] Y. Wang, W. Deng, B. Wang, Q. Zhang, X. Wan, Z. Tang, Y. Wang, C. Zhu,
G. Wang, H. Wan, Nat. Commun. 2013, 4, 2141.
This work was supported by the National Natural Science Foun-
dation of China (Nos. 21173172, 21103143, and 21033006), the
Program for Changjiang Scholars and Innovative Research Team
in Chinese Universities (No. IRT1036), and the Research Fund for
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 2014, 7, 1557 – 1567 1566

CHEMSUSCHEM
FULL PAPERS
www.chemsuschem.org
[20] S. J. Kays, S. F. Nottingham, Biology and Chemistry of Jerusalem Arti-
choke: Helianthus Tuberosus L., CRC Press, Boca Raton, 2007.
[21] S. A. Nolen, C. L. Liotta, C. A. Eckert, R. Glꢃser, Green Chem. 2003, 5,
663–669.
[22] a) D. M. West, D. A. Skoog, Anal. Chem. 1959, 31, 583–586; b) A. But,
J. L. Nꢄtre, E. L. Scott, R. Wever, J. P. M. Sanders, ChemSusChem 2012, 5,
1199–1202; c) T. Hirao, Chem. Rev. 1997, 97, 2707–2724; d) I. K. Meier, J.
Schwartz, J. Org. Chem. 1990, 55, 5619–5624.
[29] a) J. Ma, Z. Du, J. Xu, Q. Chu, Y. Pang, ChemSusChem 2011, 4, 51–54;
b) Z. Du, J. Ma, F. Wang, J. Liu, J. Xu, Green Chem. 2011, 13, 554–557;
c) Z. Du, J. Ma, H. Ma, J. Gao, J. Xu, Green Chem. 2010, 12, 590–592;
d) Z. Du, J. Ma, H. Ma, M. Wang, Y. Huang, J. Xu, Catal. Commun. 2010,
11, 732–735; e) J. Ma, W. Yu, M. Wang, X. Jia, F. Lu, J. Xu, Chin. J. Catal.
2013, 34, 492–507.
[30] a) A. M. Khenkin, R. Neumann, Angew. Chem. Int. Ed. 2000, 39, 4088–
4090; Angew. Chem. 2000, 112, 4254–4256; b) A. M. Khenkin, L. Weiner,
Y. Wang, R. Neumann, J. Am. Chem. Soc. 2001, 123, 8531–8542; c) R.
Neumann, Inorg. Chem. 2010, 49, 3594–3601.
[23] J. Liu, Z. Du, Y. Yang, T. Lu, F. Lu, J. Xu, ChemSusChem 2012, 5, 2151–
2154.
[24] T. S. Smith II, R. LoBrutto, V. L. Pecoraro, Coord. Chem. Rev. 2002, 228, 1–
18.
[31] a) B. M. Weckhuysen, D. E. Keller, Catal. Today 2003, 78, 25–46; b) K.
Chen, A. T. Bell, E. Iglesia, J. Catal. 2002, 209, 35–42.
[25] S. E. O’Donnell, M. T. Pope, J. Chem. Soc., Dalton Trans. 1976, 2290–
2297.
[32] F. Wang, J. L. Dubois, W. Ueda, J. Catal. 2009, 268, 260–267.
[26] I. Bertini, P. Turano, A. J. Vila, Chem. Rev. 1993, 93, 2833–2932.
[27] A. A. Shatalov, D. V. Evtuguin, C. P. Neto, Carbohydr. Polym. 2000, 43,
23–32.
Received: January 30, 2014
[28] D. Rehder, Dalton Trans. 2013, 42, 11749–11761.
Published online on May 2, 2014
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 2014, 7, 1557 – 1567 1567