Condition screening and process investigation of aldose transformation in borate-containing acidic phosphate buffer system under microwave irradiation

Article abstract of DOI:10.1039/c4ra02612h

The highly selective isomerization and dehydration of various carbohydrates (glucose, xylose, cellobiose and cellulose) are one-pot conversions conducted in a simple borate-containing phosphate buffer solution (PBS) under microwave irradiation. It is demonstrated that the key to glucose converting into 5-hydromethylfurfural (5-HMF) with a high selectivity is matching the isomerization and dehydration processes of glucose at the appropriate boron/glucose mole ratio (B/G) and pH of PBS with the aid of microwave irradiation. Moreover, the interaction between borate and glucose in the isomerization process has been demonstrated by both Raman spectra and theoretical calculations. Furthermore, the reusability of such a PBS system with borate has been accomplished by the successive addition of glucose and by the continual removal of 5-HMF in a biphasic system. These results not only deepen the understanding of the isomerization and dehydration behavior of glucose, but also provide the possibility for practical applications, owing to the appealing green, inexpensive and sustainable characteristics of such a catalytic system.

Full text of DOI:10.1039/c4ra02612h

RSC Advances
View Article Online
View Journal | View Issue
PAPER
Condition screening and process investigation of
aldose transformation in borate-containing acidic
phosphate buffer system under microwave
irradiation†
Cite this: RSC Adv., 2014, 4, 39453
Yani Yan, Qingbin Wu, Xiao Guo, Jinhua Lu, Zhen Hua Li,* Yahong Zhang* and Yi Tang
The highly selective isomerization and dehydration of various carbohydrates (glucose, xylose, cellobiose
and cellulose) are one-pot conversions conducted in a simple borate-containing phosphate buffer
solution (PBS) under microwave irradiation. It is demonstrated that the key to glucose converting into
5
-hydromethylfurfural (5-HMF) with a high selectivity is matching the isomerization and dehydration
processes of glucose at the appropriate boron/glucose mole ratio (B/G) and pH of PBS with the aid of
microwave irradiation. Moreover, the interaction between borate and glucose in the isomerization
process has been demonstrated by both Raman spectra and theoretical calculations. Furthermore, the
reusability of such a PBS system with borate has been accomplished by the successive addition of
glucose and by the continual removal of 5-HMF in a biphasic system. These results not only deepen the
understanding of the isomerization and dehydration behavior of glucose, but also provide the possibility
for practical applications, owing to the appealing green, inexpensive and sustainable characteristics of
such a catalytic system.
Received 25th March 2014
Accepted 5th August 2014
DOI: 10.1039/c4ra02612h
www.rsc.org/advances
as dimethylfuran, furandicarboxylic acid, levulinic acid (LA),
1
. Introduction
4
–6
and g-valerolactone.
Consequently, 5-HMF provides the
Interest in searching for green, renewable and sustainable possibility to synthesize chemicals and biofuels from carbo-
resources of energy and chemicals has increased signicantly hydrates, and so alleviates the demand pressure of fossil
in recent years, due to limited fossil resources, exhausted gas resources.
1
emissions and growing energy demands. Biomass resources
Generally, 5-HMF is mainly obtained from the dehydration
containing rich carbohydrates are one of the most promising of hexoses, such as fructose and glucose, or disaccharides, like
6,7
alternative carbon sources for the construction of chemicals sucrose and cellobiose. To date, most preparations of 5-HMF
and materials for human survival in the future. More impor- have been derived from fructose as the raw material, due to its
tantly, biomass is well-sourced, environmental friendly and higher selectivity and yield of 5-HMF compared to other
2
low cost. Therefore, study on the comprehensive utilization of hexoses. However, the relatively high cost of fructose greatly
biomass resources is attracting ever more attention. restricts its prospect in industrial applications. Unlike
Among the possibilities, 5-hydromethylfurfural (5-HMF), rstly fructose, glucose has more sources in the natural world, since
8
reported at the end of 19th century by heating inulin with it can be obtained from the hydrolysis of cellulose. However,
3
oxalic acid solution, is one of the most signicant biomass- unfortunately, glucose suffers from
a poor dehydration
derived platform compounds. As an important intermediate, performance in the preparation of 5-HMF, due to its much
9
5
-HMF can be converted into various useful compounds, such more stable aldose structure than fructose. Nevertheless, its
much lower cost indicates that the direct conversion of
glucose into 5-HMF is still the focus of scientic attention. In
Department of Chemistry, Shanghai Key Laboratory of Molecular Catalysis and addition to glucose, cellobiose and cellulose are the two most
Innovative Materials, Laboratory of Advanced Materials, Fudan University, desired raw materials to be converted directly into 5-HMF, also
Shanghai, 200433, China. E-mail: zhangyh@fudan.edu.cn; lizhenhua@fudan.edu.cn
due to their much lower cost and higher availability
compared with the other carbohydrates in nature. However,
in fact, the conversion of cellulose is much more difficult
than glucose and disaccharides, because its special b-1,4-
†
Electronic supplementary information (ESI) available: The data of the
10
conversion of glucose into 5-HMF in the PBS system without borate addition,
the conversions of glucose in the pure water under microwave irradiation and
conventional heating, the inuence of B/F ratio on the conversion of fructose
into 5-HMF, and two kinds of Gibbs free-energy proles and structures of the glycosidic bonds lead to poor solubility and difficult conver-
intermediates and transition states of the isomerization of glucose to fructose.
6
sion in aqueous solution. Thus, some extreme strategies,
See DOI: 10.1039/c4ra02612h
This journal is © The Royal Society of Chemistry 2014
RSC Adv., 2014, 4, 39453–39462 | 39453

View Article Online
RSC Advances
Paper
such as high acid concentration and reaction temperature or 2.2 Dehydration of carbohydrates
expensive ionic liquids, have to be used to swell and convert
All the PBSs with different components, as well as the acidic
solution, were prepared according to their stoichiometric ratios
in 50 mL asks. The catalytic reactions in the PBS containing
borate were performed in a 30 mL reactor under microwave
irradiation provided by Nova-2S from Preekem Scientic
Instruments Co., Ltd. In a typical procedure, glucose (1.2 g) and
borax (0.3 g) were dissolved by the ultrasonic bath in the above
PBS system, and the total volume of the reaction solution was
kept at 12 mL. Then, the mixture was put in the microwave-
11
cellulose. Therefore, the one-pot conversion of cellulose
into 5-HMF is an encouraging route but is still far from
practical.
In addition, it is worth mentioning that the main reason for
the unsatisfying results of aldose conversion into 5-HMF, based
on plenty of research results, is believed to be because a
pretreatment process is needed before the dehydration. Taking
glucose as an example, it has to rst isomerize into fructose,
7
which then dehydrates into 5-HMF. During the whole reaction,
ꢀ
assisted reactor at 160 C. For xylose and cellobiose conver-
the conversion rate of glucose mainly depends on its isomeri-
sion, the glucose was replaced by 1.2 g of xylose and cellobiose
in the above process. For cellulose conversion, a-cellulose (2.4 g)
and NaCl (3.0 g) were dissolved by the ultrasonic bath in 1.5 M
8,12,13
zation rate.
Moreover, the formation of 5-HMF in aqueous
solution is largely affected by some undesirable by-products
obtained from the rehydration of 5-HMF, such as LA and for-
mic acid (FA), as well as the insoluble/soluble polymeric
compounds, such as humins. To date, the conversion of glucose
into 5-HMF has not been achieved in a large-scale industrial
production yet, due to its low conversion rate and 5-HMF
selectivity, which is believed to be the greatest challenge
remaining.
H
3
PO
4
(10.0 g) and the mixture was placed in the microwave-
ꢀ
assisted reactor for 15 min at 160 C. Aer the reaction, the
solution was ltered, and NaH PO and borax were added to
2
4
maintain the H PO : NaH PO ratio ¼ 1 : 1 and B/G ¼ 0.5.
3
4
2
4
Finally, the mixture solution was put in the microwave-assisted
ꢀ
reactor for 30 min at 160 C. For the conventional heating
experiments, the above glucose reaction system was heated in
the oven at the same temperature. For the biphasic system, all
the procedures were the same as the above PBS system, but the
organic solvent consisting of MIBK and 2-butanol (volume ratio
Recently, an acidic phosphate buffer solution (PBS) using
sodium borate as a promoter was developed to highly selec-
tively catalyze glucose dehydration into 5-HMF under micro-
14
wave irradiation by our group. Herein, the evolution process
of the glucose conversion is highlighted in borate-
assistant PBS medium and the key factors, such as the effect
of pH, the buffer system, and the boron/glucose mole ratio
¼
7 : 3) was added before the microwave irradiation.
2.3 Computational details
(
5
B/G), on the conversion of glucose and the selectivity of
-HMF, as well as the effect of microwave irradiation, are
First-principles density functional theory (DFT) calculations
15,16
were performed with the M06-2x/6-31+G(d) method.
M06-2X
studied systematically. The highly selective conversion of
various carbohydrates (glucose, xylose, cellobiose and a-cellu-
lose) into 5-HMF/furfural is thereby accomplished in such a
system. Moreover, the interaction between borate and glucose,
as well as its effect on the glucose isomerization, are proved by
Raman spectra and theoretical calculation. Finally, a biphasic
system is employed to further improve the selectivity of 5-HMF
and to achieve the sustainability of the homogenous catalytic
system.
is by far one of the most accurate DFT functional methods in
predicting energy barriers. To account for the solvation effect,
15
the calculations were performed with the IEFPCM solvation
17
model with radii and non-electrostatic terms in Truhlar and
18
coworkers' SMD solvation model. All the geometry optimiza-
tions and vibrational frequency analyses were performed with
the solvation effect considered. The default geometry conver-
gence criterion of Gaussian 09 (ref. 19) was used. The integral
grid used for all the DFT calculation is a pruned (99 590) grid
(the “ultrane” grid as dened by Gaussian 09). To consider the
effect of pH while reducing the computational efforts to a
+
3
manageable degree, a solvated proton was simplied as H O
2
. Experimental
during the calculations. Since the computed harmonic vibra-
tional frequencies were systematically higher than the experi-
mental vibrational frequencies, the calculated vibrational
frequencies were scaled by a factor of 0.97 in order to directly
compare to experimental spectra. All these calculations were
2
.1 Chemicals
D-(+)-Glucose ($99.5%) was purchased from Sigma-Aldrich. D-
+)-Xylose ($99%), D-(+)-cellobiose (98.0%) and a-cellulose (25
mm) were purchased from Aladdin. Na B O $10H O ($99.5%),
20
(
2
4
7
2
methyl-isobutylketone (MIBK) ($99.0%) and 2-butanol
$98.0%) were purchased from Sinopharm Chemical Reagent
Co., Ltd. H PO (85.0%) was purchased from Shanghai Feida
Chemical Co., Ltd. NaH PO $2H O ($99.0%) and HCl (36.0–
8.0%) were purchased from Jiangsu Chemical Co., Ltd. NaOH Samples were taken at setting time and analyzed via an external
96.0%) was purchased from Shanghai Fourth Reactant Factory. standard method by high performance liquid chromatography
19
performed using the Gaussian 09 quantum soware package.
(
3
4
2.4 Products analysis
2
4
2
3
(
All the chemicals were used without further purication. (HPLC) equipped with a refractive index detector and a shodex
Deionized water produced from a laboratory water purication SH1011 sugar column purchased from Shimadzu Corporation.
ꢀ
system (RO DI Digital plus) was used to prepare all the aqueous
solutions.
The HPLC was detected at 40 C with 0.5 mM H
2
SO
4
mobile
ꢁ
1
phase in a ow rate of 1.0 mL min . Before being injected into
39454 | RSC Adv., 2014, 4, 39453–39462
This journal is © The Royal Society of Chemistry 2014

View Article Online
Paper
RSC Advances
the HPLC, the samples needed to be ltered through a micro- irradiation and borate, 62.0% of glucose is converted into 34.6%
ꢀ
syringe lter (VWR, 0.22 mm PTFE for the organic samples and of 5-HMF with a selectivity of 55.8% in 90 min at 160 C (Entry
PES for the aqueous samples). The retention times (HPLC) of all 1). However, the selectivity of 5-HMF displays a decrease with a
the reactants and products were determined with the standard prolonged reaction time, due to the formation of by-products
samples, and are as follows: glucose (7.0 min), fructose (7.4 such as LA, FA and humins. For example, with the prolonging
min), xylose (7.5 min), erythrose (8.4 min), glycolal (8.8 min), FA of the reaction time from 90 to 150 min, the conversion of
(9.4 min), AA (10.6 min), LA (11.7 min), 5-HMF (23.3 min) and glucose increases from 62.0% to 79.4%, while the selectivity of
furfural (35.7 min), respectively. Raman spectra data were 5-HMF decreases from 55.8% to 45.1%. Moreover, xylose
obtained by taking 400 mL of each sample interrogated in 8 mm displays the same trend as the glucose in conversion and
ꢁ
1
glass vials and detected by the Inspector785 (200–2000 cm
purchased from SciAps Inc.
)
selectivity (Entry 2). Therefore, the timely removal of product
has to be taken into consideration.
To avoid operating errors, the concentrations of products
Cellobiose and cellulose, which are composed of multi-
were corrected by an internal standard. The yields (%) and glucose molecules, are considered promising for application
selectivities (%) of the products were based on the initial in the preparation of 5-HMF, due to their low cost and adequate
concentration of carbohydrates and were calculated as follows: sources. Herein, under the same condition as glucose, the
conversion of cellobiose reaches 60.5% and produces as much
Product concentration
Yieldð%Þ ¼
ꢂ 100%
as 5-HMF of 33.2% in 120 min, which is quite good performance
for cellobiose in aqueous solution (Entry 3). For cellulose, it is
inspiring that a fast hydrolysis of a-cellulose is observed in a
NaCl-assisted H PO system and that a 25.9% yield of glucose
Theoretical product concentration
3
4
Conversionð%Þ ¼
ꢀ
ꢁ
(5.2 wt%) is obtained from 20 wt% a-cellulose solution without
Carbohydrate concentration
1
ꢁ
ꢂ 100% any pretreatment in 15 min under microwave irradiation (Entry
Initial carbohydrate concentration
4). Furthermore, a 48.8% selectivity of 5-HMF based on the as-
prepared glucose can be achieved from the above hydrolytic
solution of a-cellulose in such a PBS system (Entry 5). These
impressive results are believed to be attributed to the solution
acidity, the buffer system, and the effect of borate, which lead to
perfect matching of the isomerization and dehydration rates
during the glucose/xylose conversion.
Yield
ꢂ 100%
Selectivityð%Þ ¼
Conversion
3. Results and discussion
3.1 Dehydration of glucose, xylose, cellobiose and a-cellulose
3.2 Effect of PBS system
The dehydration condition of various carbohydrates in aqueous Different mineral acidic solutions at different pH values con-
solution is optimized by a series of experiments, including pH taining a xed amount of borate have been employed to study
of the PBS system and the borate content, which will be the isomerization and dehydration of glucose. As shown in Table
explained in detail later. Table 1 shows the results of glucose, 2, glucose in the PBS system at different pH values can be
xylose, cellobiose and a-cellulose conversion into 5-HMF/ transformed to various products. In the stronger acid system (pH
furfural in the optimum borate-assistant PBS system under ¼ 1.1), 5-HMF, LA, and FA are formed, indicating the instability
microwave irradiation. With the assistance of microwave of 5-HMF at such a pH value. When the pH reaches 4.7, glucose
a
Table 1 The dehydration results of glucose, cellobiose and a-cellulose in the PBS system under microwave irradiation
b
b
Entry
1
Aldose
Time (min)
90
Conv. (%)
Y (%)
S (%)
Glucose
62.0
71.3
79.4
56.8
86.1
51.7
60.5
—
34.6
37.1
35.8
27.6
37.1
30.5
33.2
25.9
21.0
55.8
52.0
45.1
48.6
43.1
59.0
54.9
—
1
1
20
50
30
c
2
3
Xylose
60
Cellobiose
90
20
15
30
1
d
4
5
a-Cellulose
As-prepared glucose (5.2 wt%)
e
43.0
48.8
a
ꢀ
Reaction conditions: 10 wt% glucose, cellobiose in 0.5 M H
3
PO
4
/0.5 M NaH
2
PO
4
PBS system, borax ¼ 0.3 g, and reaction temperature ¼ 160 C.
b
c
Yield and selectivity of 5-HMF for Entries 1, 3 and 5; yield and selectivity of furfural for Entry 2; yield of glucose for Entry 4. 10 wt% xylose in 0.5 M
d
H
3
PO
4
/0.5 M NaH
2
PO
4
PBS system with B/X ¼ 0.25. Reaction condition: 20 wt% a-cellulose in 1.5 M H
3 4
PO solution with 3 g NaCl in 15.4 g reaction
e
solution. Filtered solution from Entry 3 reacted in 1.5 M H
selectivity.
3
PO
4
/1.5 M NaH
2
PO
4
PBS system with B/G ¼ 0.5. Conv.: conversion, Y: yield, S:
RSC Adv., 2014, 4, 39453–39462 | 39455
This journal is © The Royal Society of Chemistry 2014

View Article Online
RSC Advances
Paper
Table 2 The conversion of glucose (10 wt%) and the corresponding product distributions in PBS at different pH values under microwave
a
irradiation
Solution
pH
Conv. (%)
Y5-HMF (%)
YFru. (%)
YLA (%)
YFA (%)
YAA (%)
YEry. (%)
1
0
1
0
.0 M H
.5 M H
.0 M NaH
.5 M NaH
3
3
PO
PO
4
1.1
2.1
4.7
6.8
37.0
62.0
93.1
94.8
9.7
34.6
6.6
0.0
3.8
0.0
0.0
3.1
0.7
0.0
0.0
16.0
11.1
11.3
14.1
0.0
0.0
3.9
4.2
0.0
0.0
8.3
4
/0.5 M NaH
2
PO
4
2
PO
PO
4
2
4
/0.5 M Na
2
HPO
4
0.0
36.9
a
ꢀ
Reaction conditions: reaction time ¼ 90 min, B/G ¼ 0.5, and reaction temperature ¼ 160 C. 5-HMF: 5-hydromethylfurfural; Fru.: fructose; LA:
levulinic acid; FA: formic acid; AA: acetic acid; Ery.: erythrose; Conv.: conversion; Y: yield.
seems to prefer other conversion routes rather than the dehy- inuences the glucose isomerization rate. Furthermore, in the
dration way, such as retro-aldol condensation, and hence strong acid solution, the 5-HMF formed is inclined to rehydrate
erythrose is detected in the reaction system. Furthermore, the into LA and FA, which leads to a poor selectivity of 5-HMF. On
formation of erythrose becomes predominant at a pH of 6.8, and the other hand, when the pH value is higher than 2.1, such as at
no 5-HMF could then be observed. Only at a pH of 2.1 does the 4.7 or 6.8, the increase in the glucose isomerization rate and
reaction display the best result. Furthermore, it is found that the other side-reaction rates causes a higher conversion of glucose
conversion of glucose decreases with the increasing acidity of and a lower selectivity of 5-HMF. This can also be demonstrated
the system, i.e., the lowest conversion is obtained at a pH of 1.1. by the conversion of xylose into furfural. As shown in Fig. 1A,
This is contrary to the results of fructose conversion in the the highest selectivity of furfural is achieved at a pH of 2.1, and
14
solution at different pH values. Moreover, it is surprising that high acidity (pH 1.1) leads to a low conversion of xylose in the
other reaction systems, such as the diluted strong acid system at borate-containing PBS system.
different pH values (Table 3) and PBS systems without borate
Since the pH value of the reaction system has a dramatic
addition (Fig. S1, ESI†), also display the same trend, not only for effect on the conversion of glucose, the use of a buffer system is
the selectivity of 5-HMF but also for the conversion of glucose, critical for its ability to maintain the pH value during the
i.e. the best performance of the reaction is at a pH of 2.1 and the reaction. A series of experiments were conducted to study the
lowest conversion of glucose is at a pH of 1.1. These results effect of the buffer system during glucose dehydration in the
clearly show the importance of acidity in the glucose dehydra- H SO solution with different pH values, similar to those in the
2
4
tion. On the one hand, when the pH value is lower than 2.1, the PBS systems. As shown in Table 3, aer being irradiated for 90
ring-structure of glucose is more stable and is not easy to min, glucose converts only 16.0% and the yield of 5-HMF is
isomerize into fructose, even with the existence of borate. It is merely 8.2%, which is far less than the results obtained in the
ꢁ
believed that B(OH)
4
can form stable chelate complexes with buffer system at the same pH value (2.1). It is obvious that the
carbohydrates and can catalyze the isomerization of aldoses to solution with the appropriate pH value only at the beginning is
2
1–24
ketoses.
However, in the strong acidic system, the concen- not enough for a good catalytic performance during the whole
ꢁ
tration of B(OH)
4
is quite low due to the limitation of the borate reaction. The generation of FA during the reaction at all pH
hydrolysis equilibrium, as shown by eqn (1).
values ranging from 1.1 to 6.8 will have an impact on the acidity
of the reaction system. Therefore, it is difficult for the diluted
acid solution consisting of H SO to maintain its pH value
3
BO
3
+ H
2
O # B(OH) + H
4ꢁ
+
(1)
H
2
4
during the whole reaction and so its acidity becomes stronger
alongside the generation of the organic acid. This can be
demonstrated by its product distributions at different pH
+
The higher H concentration suppresses the shi of equi-
librium from le to right at lower pH values, and thereby
Table 3 The conversion of glucose (10 wt%) and the corresponding
product distributions in H
microwave irradiation
2
SO
4
solution at different pH values under
a
pH
Conv. (%)
Y5-HMF (%)
YFA (%)
YEry. (%)
YGly. (%)
1
2
4
6
.1
.1
.7
.8
13.9
16.0
26.3
71.8
3.6
8.2
10.6
0.0
10.0
3.2
1.1
6.2
0.0
0.0
4.8
0.0
0.0
0.0
14.1
16.1
a
Fig. 1 Effects of pH value (A) and B/X ratio (B) on the conversion of
Reaction conditions: reaction time ¼ 90 min, B/G ¼ 0.5, and reaction
ꢀ
xylose (10 wt%) into furfural under microwave irradiation. For all, the
temperature ¼ 160 C. 5-HMF: 5-hydromethylfurfural; FA: formic acid;
Ery.: erythrose; Gly.: glycolal; Conv.: conversion; Y: yield.
ꢀ
reaction time is 30 min and the reaction temperature is 160 C. Fur.:
furfural; Conv.: conversion; Y: yield; S: selectivity.
39456 | RSC Adv., 2014, 4, 39453–39462
This journal is © The Royal Society of Chemistry 2014

View Article Online
Paper
RSC Advances
values, which regularly move to more acidic ones compared to selectivity of 5-HMF. Moreover, the yields of the by-products LA
those of PBS systems (Table 2). Moreover, the shi of pH values and FA derived from the rehydration of 5-HMF also show the
toward the stronger acidic system will also suppress the same trend as that of 5-HMF, indicating that the decrease of 5-
conversion of glucose, as indicated above. On the contrary, the HMF selectivity at the high B/G ratio could be assigned to the
PBS system is able to provide a self-regulation reaction envi- occurrence of other side-reactions before the dehydration of
ꢁ
ronment, and so the acidity of the solution is not infected fructose. Although B(OH)
4
, which is formed by the hydrolysis
notably by the generated organic acids. As a consequence, the of boric acid, can facilitate the isomerization of aldoses into
use of the PBS system is one of the key indispensables for a high ketoses by forming stable chelate complexes with aldoses/
ꢁ
selectivity of 5-HMF and a fast conversion of glucose.
ketoses, the strong complexation between B(OH)
4
and
glucose/fructose also blocks the elimination of water in the rst
step of the fructose conversion into 5-HMF and will induce
3.3 Effect of borate
24
other side-reactions. Therefore, an excessive amount of borate
will upset the dehydration of fructose, although it can also
accelerate the conversion of glucose into fructose, which can be
proved by the results of fructose dehydration with borate-
containing PBS system (Table S1, ESI†). As shown in Table
S1,† a high boron/fructose mole ratio (B/F) also results in a low
selectivity of 5-HMF. Moreover, the content of glucose is stable
with increasing B/F ratios during the reaction, which implies
that the high B/F ratio has a great inuence on the conversion of
fructose rather than on its isomerization toward glucose.
Furthermore, the similar conversion of fructose and the low
yield of 5-HMF further indicate the occurrence of other side-
reactions during the dehydration of fructose with the increase
of B/F ratio. In addition, the same trend is also observed in the
dehydration of xylose into furfural in Fig. 1B. With increasing B/
X ratios, the conversion of xylose into furfural is accelerated, but
the best result can be obtained at B/X ¼ 0.25. Consequently, it
can be concluded that a suitable B/G ratio is the key to the
isomerization of glucose and the dehydration of fructose, and
that a good yield of 5-HMF is only achieved with good matching
between them.
Table 4 displays the effect of B/G ratios on the conversion of
glucose into 5-HMF. Obviously, an increasing B/G ratio leads to
an accelerating conversion rate of glucose, but very diverse
yields and selectivities of 5-HMF within the same reaction time
when the pH value of the reaction system is xed at 2.1.
According to Table 4, without the addition of borate, the
conversion of glucose is only 43.0% in 90 min. On the contrary,
the glucose converts faster with the addition of borate, and the
conversion of glucose can then reach 86.0% at a 1.5 B/G ratio in
30 min. This result is consistent with those of Table 2 and
Fig. S1 (ESI†). It is clear that borate can accelerate the conver-
sion rate of glucose. Moreover, in the borate-containing reac-
tion system, fructose is detected during the reaction, which
indicates that borate has a positive effect on the isomerization
of glucose into fructose. Furthermore, this can be further sup-
ported by theoretical calculations (see below). During the reac-
tion, glucose rst isomerizes into fructose, which then
dehydrates into 5-HMF continuously with the assistance of
mineral acid and borate in the PBS system.
However, the selectivity of 5-HMF displays a quite different
trends with the increasing B/G ratio; i.e., the highest selectivity
of 5-HMF is obtained at B/G ¼ 0.5 with a relatively high
conversion of glucose. When the B/G ratio is lower, the accel-
erating effect of borate for the isomerization of glucose is not
enough, although there is an acceptable selectivity of 5-HMF.
When the B/G ratio becomes higher, such as a B/G of 1.5, the
reaction shows the fastest conversion of glucose and the lowest
To further study the effect of borate, a series of Raman
spectra are measured, as shown in Fig. 2A–C. Moreover, the
theoretical spectra are given in Fig. 2D to verify these Raman
analyses. For the pure borate solution (Fig. 2Aa), only the
ꢁ
ꢁ1
characteristic peaks of B(OH)
4 3 3
and H BO appear at 743 cm
ꢁ1
25
and 877 cm , respectively. In the pure glucose solution
a
Table 4 Influence of B/G ratio on the conversion of glucose into 5-HMF under microwave irradiation
B/G
0
Time (min)
Conv. (%)
Y
5-HMF (%)
S
5-HMF (%)
Y
Fru. (%)
Y
LA (%)
YFA (%)
90
90
90
43.0
48.2
62.0
71.3
71.9
69.3
83.2
82.7
91.7
86.0
93.9
98.6
22.0
24.8
34.6
37.1
36.6
33.7
39.5
34.9
37.1
21.6
23.0
16.3
51.1
51.5
55.8
52.0
50.9
48.6
47.5
42.2
40.5
25.1
24.5
16.5
0.0
0.3
3.8
4.0
5.5
7.4
3.9
5.1
3.9
4.0
2.1
0.0
1.1
0.7
0.7
1.8
1.3
0.3
0.7
0.0
0.5
0.0
0.0
0.0
7.0
5.6
11.1
17.7
17.5
0.0
0.0
0.6
1.8
0.0
0
0
.25
.5
1
20
90
60
0
1
.75
90
1
1
.25
.5
60
90
30
6
9
0
0
0.8
1.8
a
ꢀ
Reaction conditions: 10 wt% glucose in 0.5 M H
3
PO
4
/0.5 M NaH
2
PO
4
PBS system, B/G ¼ 0–1.5, and reaction temperature ¼ 160 C. 5-HMF: 5-
hydromethylfurfural; Fru.: fructose; LA: levulinic acid; FA: formic acid; Conv.: conversion; Y: yield; S: selectivity.
This journal is © The Royal Society of Chemistry 2014
RSC Adv., 2014, 4, 39453–39462 | 39457

View Article Online
RSC Advances
Paper
Fig. 2 Raman spectra of the reaction system. (A) Raman spectra of borate solution (a), 10 wt% glucose solution (b) and 10 wt% glucose solution
containing borate (c); (B) Raman spectra of 10 wt% glucose solution (a), 10 wt% glucose in PBS (b) and 10 wt% glucose in PBS containing borate
(
c); (C) Raman spectra of the system reacted for 20 min (a) and 10 min (b), and the original reaction system (c); (D) theoretical Raman spectra of
ꢁ
borate solution (a), glucose solution (b), and the ring-opening form of the B(OH)
4
–glucose chelate complex (c).
ꢁ1
ꢁ1
(
Fig. 2Ab), the peaks emerging at 912 cm and 851 cm are
To get a better understanding of the conversion process of
assigned to the in-plane stretching vibration of C–C]O, and the glucose, the reaction was determined by Raman spectra every 10
ꢁ
1
ꢁ1
signals at 1127 cm , 1065 cm are the peaks of the out-of- min under microwave irradiation. It is clear that, during the
26
ꢁ1
ꢁ1
plane stretching vibration of C–C]O. However, in the solu- reaction, some new peaks arise at 1664 cm , 1524 cm , 1263
tion of borate and glucose mixture (Fig. 2Ac), the in-plane cm , and 1200 cm , respectively in Fig. 2C, which can be
stretching vibration of C–C]O in the glucose molecule assigned to the generation of 5-HMF. In addition, the
ꢁ
1
ꢁ1
27
displays a great change compared to that of pure glucose solu- increasing intensity of the 5-HMF characteristic peaks and the
ꢁ
ꢁ
1
ꢁ1
ꢁ1
tion, and moves to lower wavenumber (781 cm and 871 cm ), decreasing glucose peak at 1126 cm also imply the conversion
which indicates the formation of B(OH)4 –glucose chelate of glucose into 5-HMF (Fig. 2C).
complexes. Moreover, this interaction could also be proved by
Moreover, these experimental Raman spectra agree well with
the theoretical calculations as well; i.e., borate does have an the calculation results, especially in the spectra of the borate
effect on the ring opening process of glucose during the reac- and glucose ring-opening process. The experimental results of
ꢁ
ꢁ1
ꢁ1
tion, as shown in Fig. 2Dc. In addition, the out-of-plane B(OH)
stretching vibration of its C–C]O also shows a clear change, consistent with the theoretical ones at 732 cm and 864 cm
although the change is slight. However, when glucose is dis- in Fig. 2Da. As shown in Fig. 2Ac, the characteristic peaks at 781
4
3 3
and H BO at 743 cm and 877 cm in Fig. 2Aa are
ꢁ1
ꢁ1
ꢁ
1
ꢁ1
ꢁ1
ꢁ1
solved in PBS medium consisting of 0.5 M H
3 4
PO and 0.5 M cm , 871 cm , 1060 cm , and 1130 cm of the glucose
NaH PO , the in-plane stretching vibration of C–C]O in solution containing borate correspond with the calculated ones
2
4
ꢁ
glucose is totally covered by the strong peaks caused by the of the ring-opening form of the B(OH)4 –glucose chelate
mineral acid and salt (Fig. 2Bb). Only the small change on the complex (Fig. 2Dc). Accordingly, it is believed that tetrahy-
ꢁ
out-of-plane stretching vibration of C–C]O could still be droxyborate [B(OH)
observed (Fig. 2Bc). zation of glucose.
4
] plays an important role in the isomeri-
39458 | RSC Adv., 2014, 4, 39453–39462
This journal is © The Royal Society of Chemistry 2014

View Article Online
Paper
.4 Effect of microwave irradiation
RSC Advances
3
bond in the ring with the help of a proton. The resulting
opening form of glucose (acyclic-glucose) undergoes an intra-
molecular hydrogen transfer to the opening form of fructose
The effect of microwave heating on the conversion of glucose was
also studied, and the results are shown in Fig. 3. Obviously,
microwave irradiationleadstoa fasterconversion ofglucoseanda
higher selectivity of 5-HMF compared with the results under oven
heating. For example, the reaction system under microwave
heating can achieve a glucose conversion of 62.0% with a 5-HMF
(
acyclic-fructose). Acyclic-fructose then undergoes a ring closing
process to form the ring-close form of fructose. The Gibbs free-
energy proles and the structures of the intermediates and
transition states of the isomerization process without borate are
presented in Fig. 4. Fig. 4 shows that the overall free-energy
barrier for the isomerization of glucose into fructose is 29.3
ꢀ
selectivity of 55.8% within 90 min at 160 C, but under oven
heating this just reaches a glucose conversion of 63.1% with a 5-
HMF selectivity of50.9% within 120 min at the same temperature.
The observed rate enhancement can be contributed to the nature
of the microwave direct, volumetric heating model and the reac-
ꢁ
1
kcal mol when borate is absent.
ꢁ
In the reaction system containing borate, B(OH)4 can form
the stable chelate complex, boroglucopyronose, with glucose, by
anchoring with two hydroxyl groups of glucose through the
removal of two water molecules. Similar to the results of Anders
2
8
tion system with a high ion concentration. This acidic PBS
solution containing borate contains an abundance of ions such as
2
4
ꢁ
2ꢁ
3ꢁ
ꢁ
+
+
Riisager et al., we nd that 4,6-boroglucopyronose and 3,4-
boroglucopyronose are the two most stable complexes between
H PO
2 4
, HPO
4
, PO
4
, B(OH) , H O , Na , which leads to a
4 3
rapid temperature rise under microwave irradiation, since ionic
conduction is one of the two main microwave heating models.
ꢁ
29
B(OH)
4
and glucose, and the former is slightly more stable
ꢁ
1
than the latter by 3.2 kcal mol at 298.15 K in Gibbs free
energy. Under the reaction conditions, these two complexes can
be viewed as if in equilibrium with each other. Starting from
Such instantaneous heating process results in not only a fast
conversionofglucose but alsoa highselectivityof5-HMF, because
an extended reaction time could cause the formation of by-
products and a decrease of selectivity. These results can also
explain the reason for the similar results of glucose conversion in
the pure water heated by conventional means and by microwave
irradiation (Fig. S2, ESI†). The ion concentration in the water is
very low, and so the instantaneous heatingeffect ofthe microwave
2
9
either 4,6-boroglucopyronose or 3,4-boroglucopyronose, a
similar isomerization pathway as that one without borate as
presented in Fig. 4 can be determined in Fig. 5 and S3 (ESI†).
However, it is found that the pathway starting with 4,6-bor-
oglucopyronose has a higher energy barrier (with an overall free-
ꢁ
1
energy barrier of 25.5 kcal mol , Fig. S3, ESI†) than that
eld becomes insignicant. Therefore, the difference in glucose
starting with 3,4-boroglucopyronose (Fig. 5). The latter one has
conversion in the pure water between conventional heating and
microwave heating cannot be clearly observed. Therefore, micro-
wave irradiation is also another inuence factor on the highly
selective conversion of glucose into 5-HMF in such a PBS system.
ꢁ
1
an overall free-energy barrier of just 17.9 kcal mol . As can be
seen from Fig. 4 and 5, the addition of borate can greatly lower
ꢁ
the energy barrier of the isomerization. Binding with B(OH)
4
increases the tension of the 6-membered ring of glucose, which
lowers its Gibbs free-energy barrier for the opening from the
closed-form to the open-form through the transition state TS1
from 22.8 kcal mol to just 17.9 kcal mol . The second step,
the hydrogen migration from –C HOH– to the terminal
3.5 Computational study
ꢁ
1
ꢁ1
Theoretical calculations were performed to understand the
isomerization mechanism of glucose to fructose in acidic
solution with or without borate. The calculations indicate that
the isomerization begins with the opening of the 6-member
ring-close form of glucose (glucopyranose) by breaking its C–O
(
2)
(
1)
–
C
H]O of acyclic-glucose to form acyclic-fructose, is also
greatly enhanced, with the free-energy-barrier lowering from
ꢁ
1
ꢁ1
1
9.9 kcal mol to 13.5 kcal mol
.
ꢁ
In the calculation, B(OH)
order to study if H PO
4
3 4
is also replaced by H PO , in
can accelerate the isomerization of
3
4
glucose as well. A similar pathway as presented in Fig. 5 is
computed (Fig. S4, ESI†). The overall Gibbs free-energy barrier
ꢁ
1
in the reaction system containing H PO is 25.1 kcal mol
,
3
4
ꢁ
which is higher than that with the help of B(OH) (17.9 kcal
mol , Fig. 5), but is lower than that with only protons but
without other additives (29.3 kcal mol , Fig. 4). This means
that H PO can also facilitate the isomerization of glucose, but
4
ꢁ
1
ꢁ1
3
4
its effect is weaker than that of borate. The results are consistent
with experimental data and prove the key role of borate in
accelerating the isomerization of glucose. In addition, it is
worth mentioning that the pathway of anchoring two glucose
ꢁ
molecules with one B(OH)4 is considered to mimic the effect of
borate concentration and changed the B/G ratio from 1 to 0.5.
When the borate content is decreased, the overall Gibbs free-
energy barrier of the glucose isomerization process increased
to 20.9 kcal mol . This implies that the isomerization is faster
with a B/G ratio of 1 than that of 0.5, which is also consistent
Fig. 3 The conversions of 10 wt% glucose in the borate-containing
PBS system under microwave irradiation and with conventional
ꢁ1
ꢀ
heating at 160 C. 5-HMF: 5-hydromethylfurfural; Conv.: conversion;
Y: yield; S: selectivity.
This journal is © The Royal Society of Chemistry 2014
RSC Adv., 2014, 4, 39453–39462 | 39459

View Article Online
RSC Advances
Paper
Fig. 4 Gibbs free energy profiles at 298.15 K and structures of the intermediates and transition states of the isomerization of glucose to fructose
ꢁ1
+
3
with the help of proton. The numbers are relative free energies (in kcal mol ) to the complex between glucose and H O .
Fig. 5 Gibbs free energy profiles at 298.15 K and structures of the intermediates and transition states of the isomerization of glucose to fructose
ꢁ1
via 3,4-boroglucopyronose with the help of protons. The numbers are relative free energies (in kcal mol ) of the complexes between D-glucose
+
3
and H O .
with the experimental results (Table 4). Besides, our calcula- evidence for the reliability of our theoretical calculations and
tions also indicate that the ring opening form of 3,4-bor- the proposed isomerization mechanism.
oglucopyronose is
a
stable intermediate during the
isomerization process: its relative free energy to 3,4-bor-
oglucopyronose is just 3.8 kcal mol (Fig. 5). Its conversion
ꢁ
1
3.6 Sustainability of PBS system with borate
back to 3,4-boroglucopyronose has a free-energy barrier of 14.1 Compared with the solid acid catalyst, one of the biggest
ꢁ1
kcal mol (transition state TS1), while its isomerization to disadvantages of liquid acids is their poor reusability. Solid acid
ꢁ1
acyclic-fructose through hydrogen-migration is 13.5 kcal mol
transition state TS2). Therefore, it is possible to observe this processes, while liquid acid catalysts normally cannot be recy-
intermediate during the isomerization process. In fact, the cled and used again aer the reaction. Moreover, the general
Raman spectrum of this intermediate agrees well with the mineral acids (HCl and H SO ) are highly corrosive for the
catalysts can be easily recycled and reused aer a series of
(
2
4
experimental Raman spectrum of glucose in the PBS solution reaction vessel. Compared to them, this PBS system consisting
containing borate (Fig. 2). This agreement is also strong of phosphoric acid/acid phosphate and borate is safer and less
30
corrosive. Furthermore, as indicated by Dumesic, such a PBS
system can also allow the possibility of recycling by a biphasic
39460 | RSC Adv., 2014, 4, 39453–39462
This journal is © The Royal Society of Chemistry 2014

View Article Online
Paper
RSC Advances
a
Table 5 The sustainability of the PBS system with borate
5-HMF involved in the matching of isomerization and dehy-
dration processes, but also present the possibility of achieving a
sustainable green process for carbohydrates conversion.
Glucose
wt%)
Time
(min)
Conv.
(%)
(
Y5-HMF (%)
S5-HMF (%)
Run 1
Run 2
Run 3
Total
10
5
5
90
90
90
270
90
57.8
51.3
43.9
74.0
74.9
34.6
30.5
26.0
44.7
39.6
60.0
59.5
59.2
60.4
52.9
Acknowledgements
This work is supported by 973 Program (2013CB934101,
20
2011CB808505), NSFC (21171041, 21273042), and STCMS
Original 20
(11JC1400400, 08DZ2270500 and 09DZ2271500).
a
ꢀ
Reaction conditions: reaction temperature ¼ 160 C and B/G ¼ 0.5 in a
biphasic system. An organic phase (MIBK : 2-butanol ¼ 7 : 3) with the
same volume as the aqueous phase is added into the above PBS Notes and references
system before reaction. 5-HMF: 5-hydromethylfurfural; Conv.:
conversion; Y: yield; S: selectivity.
1
T. Tuercke, S. Panic and S. Loebbecke, Chem. Eng. Technol.,
2
009, 32, 1815–1822.
Y. Roman-Leshkov, J. N. Chheda and J. A. Dumesic, Science,
006, 312, 1933–1937.
G. D u¨ ll, Chem.-Ztg., 1895, 19, 216–220.
Y. Nakagawa, M. Tamura and K. Tomishige, ACS Catal.,
2
2
system, that is, by continuously removing 5-HMF by organic
solvents and the continual feed of glucose. Herein, three step-
wise additions of glucose and organic phase were used to
investigate the possibility of the reusability of this system. As
shown in Table 5, the selectivity of 5-HMF could be maintained
aer three additions of glucose, and reaches 60% even when the
total glucose concentration is up to 20 wt%. Moreover, the
results of separately adding one is higher than not only that of
3
4
2
013, 3, 2655–2668.
J. C. Maria, C. Avelino and I. Sara, Green Chem., 2011, 13,
20–540.
5
6
7
8
9
5
A. A. Rosatella, S. P. Simeonov, R. F. M. Frade and
C. A. M. Afonso, Green Chem., 2011, 13, 754–793.
Y. J. Pag ´a n-Torres, T. Wang, J. M. R. Gallo, B. H. Shanks and
J. A. Dumesic, ACS Catal., 2012, 2, 930–934.
10 wt% glucose in pure PBS phase (Entry 1, Table 1) but also
from 20 wt% glucose in the biphasic system (Table 5), although
the conversion of glucose is almost at the same level. In addi-
tion, the PBS catalysts can also be recovered by neutralizing
T. St ˚a hlberg, M. G. Sørensen and A. Riisager, Green Chem.,
2010, 12, 321–325.
E. Nikolla, Y. Roman-Leshkov, M. Moliner and M. E. Davis,
ACS Catal., 2011, 1, 408–410.
0 S. Dutta, S. De and B. Saha, Biomass Bioenergy, 2013, 55, 355–
them into their salts such as Na
Na PO in water is poorer and more sensitive to temperature
compared to its acidic form, and can be easily recrystallized. For
3 4
PO . Generally, the solubility of
3
4
1
369.
example, if the PBS solution is neutralized to Na
addition of NaOH, 31 wt%, 53 wt%, and 74 wt% of Na
3
PO
4
by the
could
1
1
1 Y. B. Huang and Y. Fu, Green Chem., 2013, 15, 1095–1111.
2 M. Moliner, Y. Roman-Leshkov and M. E. Davis, Proc. Natl.
Acad. Sci. U. S. A., 2010, 107, 6164–6168.
3 H. E. Van Dam, A. P. G. Kieboom and H. Van Bekkum,
Starch-St a¨ rke, 1986, 3, 95–101.
3
PO
4
ꢀ
ꢀ
ꢀ
31
be recovered in theory at 20 C, 10 C, and 0 C, respectively.
According to our experiments, more than 90% of the theoretical
amount of Na PO could be obtained. These results suggest a
1
1
1
1
3
4
possible reusability of the PBS system for future practical
applications.
4 J. H. Lu, Y. N. Yan, Y. H. Zhang and Y. Tang, RSC Adv., 2012,
2
, 7652–7655.
5 Y. Zhao and D. G. Truhlar, J. Phys. Chem., 2006, 110, 5121–
129.
5
4
. Conclusions
6 R. Krishnan, J. S. Binkley, R. Seeger and J. A. Pople, J. Chem.
Phys., 1980, 72, 650–654.
In summary, PBS system containing borate with an appropriate
pH has been shown to be able to convert glucose, cellobiose and 17 G. Scalmani and M. J. Frisch, J. Chem. Phys., 2010, 132, 110–
a-cellulose into 5-HMF/furfural directly with a high selectivity. It 114.
is found that the key factor for the highly selective conversion of 18 A. V. Marenich, C. J. Cramer and D. G. Truhlar, J. Phys. Chem.
glucose to 5-HMF is the matching of the isomerization and B, 2009, 113, 6378–6396.
dehydration processes of glucose, i.e. matching among appro- 19 M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria,
priate pHs (2.1), B/G ratios (0.5) and the PBS system under
microwave irradiation. Only under that unique condition, can
glucose isomerize into fructose and then dehydrate into 5-HMF
continuously at the matched speed. Both the Raman spectra
and theoretical calculations prove the interaction between
borate and glucose, as well as the borate-assistant isomerization
process of glucose. Moreover, the PBS system with borate is
expected to achieve reusability and stable catalytic performance
in the biphasic system. These results not only deepen the
understanding of the one-pot conversion process of glucose to
M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone,
B. Mennucci, G. A. Petersson, H. Nakatsuji, M. Caricato,
X. Li, H. P. Hratchian, A. F. Izmaylov, J. Bloino, G. Zheng,
J. L. Sonnenberg, M. Hada, M. Ehara, K. Toyota,
R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda,
O. Kitao, H. Nakai, T. Vreven Jr J. A. Montgomery,
J. E. Peralta, F. Ogliaro, M. Bearpark, J. J. Heyd,
E. Brothers, K. N. Kudin, V. N. Staroverov, R. Kobayashi,
J. Normand, K. Raghavachari, A. Rendell, J. C. Burant,
S. S. Iyengar, J. Tomasi, M. Cossi, N. Rega, J. M. Millam,
This journal is © The Royal Society of Chemistry 2014
RSC Adv., 2014, 4, 39453–39462 | 39461

View Article Online
RSC Advances
Paper
M. Klene, J. E. Knox, J. B. Cross, V. Bakken, C. Adamo, 25 S. A. Devi, D. Philip and G. Aruldhas, J. Solid State Chem.,
J. Jaramillo, R. Gomperts, R. E. Stratmann, O. Yazyev, 1994, 113, 157–162.
A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, 26 X. Yang, A. Y. Zhang, D. A. Wheeler, T. C. Bond, C. Gu and
R. L. Martin, K. Morokuma, V. G. Zakrzewski, G. A. Voth, Y. Li, Anal. Bioanal. Chem., 2012, 402, 687–691.
P. Salvador, J. J. Dannenberg, S. Dapprich, A. D. Daniels, 27 T. Kim, R. S. Assary, L. A. Curtiss, L. Marshall and P. C. Stair,
O. Farkas, J. B. Foresman, J. V. Ortiz, J. Cioslowski and J. Raman Spectrosc., 2011, 42, 2069–2076.
D. J. Fox, Gaussian 09, Revision A.02, Gaussian, Inc., 28 H. J. Kitchen, S. R. Vallance, J. L. Kennedy, N. Tapia-Ruiz,
Wallingford CT, 2009.
0 I. M. Alecu, J. Zheng, Y. Zhao and D. G. Truhlar, J. Chem.
Theory Comput., 2010, 6, 2872–2887.
L. Carassiti, A. Harrison, A. G. Whittaker, T. D. Drysdale,
S. W. Kingman and D. H. Gregory, Chem. Rev., 2014, 114,
1170–1206.
2
21 J. Verchere and M. Hlaibi, Polyhedron, 1987, 6, 1415–1420.
22 J. F. Mendicino, J. Am. Chem. Soc., 1960, 82, 4975–4979.
23 K. B. Hicks, E. V. Symanski and P. E. Pfeffer, Carbohydr. Res.,
29 C. O. Kappe, Angew. Chem., Int. Ed., 2004, 43, 6250–6284.
30 Y. Roman-Leshkov, C. J. Barrett, Z. Y. Liu and J. A. Dumesic,
Nature, 2007, 447, 982–985.
1
983, 112, 37–50.
31 Lange's Chemistry Handbook, ed. J. A. Dean, Revision 13,
Science Press, China, 1991, pp. 10–19, Table 10–2.
2
4 T. St ˚a hlberg, S. Rodriguez-Rodriguez, P. Fristrup and
A. Riisager, Chem. –Eur. J., 2011, 17, 1456–1464.
39462 | RSC Adv., 2014, 4, 39453–39462
This journal is © The Royal Society of Chemistry 2014