Isomerization of hexoses from enzymatic hydrolysate of poplar sawdust using low leaching K2MgSiO4 catalysts for one-pot synthesis of HMF

Article abstract of DOI:10.1039/c5ra17132f

A cost-effective K_1.5Mg/SiO₂ catalyst with a K₂MgSiO₄ component was employed to help transfer several hexoses into hydroxylmethylfurfural (HMF) in a biphasic water/butanol system. The solid catalysts with rod-like structures and moderate base sites could promote the aldose-ketose isomerization of glucose and have potential for the synthesis of lactic acid via a retro-aldol condensation. In the presence of sulfonated titania/mordenite solid acid, the K_1.5Mg/SiO₂ catalyst presented heat-stability and prolonged activity in the one-pot conversion of glucose to HMF, leading to an optimal yield of 62.2%. The one-pot catalysis of saccharides from the enzymatic hydrolysate of poplar sawdust achieved a high HMF yield of 56.9% at a glucose conversion of 91.6%.

Full text of DOI:10.1039/c5ra17132f

RSC Advances
View Article Online
View Journal | View Issue
PAPER
Isomerization of hexoses from enzymatic
hydrolysate of poplar sawdust using low leaching
K MgSiO catalysts for one-pot synthesis of HMF
Cite this: RSC Adv., 2015, 5, 96990
2
4
ab
c
a
a
b
Xiang Shen, Yanxin Wang,* Birgitte K. Ahring,* Hanwu Lei, Qiang Gao
and Hui Liu
c
2 2 4
A cost-effective K1.5Mg/SiO catalyst with a K MgSiO component was employed to help transfer several
hexoses into hydroxylmethylfurfural (HMF) in a biphasic water/butanol system. The solid catalysts with
rod-like structures and moderate base sites could promote the aldose–ketose isomerization of glucose
and have potential for the synthesis of lactic acid via a retro-aldol condensation. In the presence of
Received 24th August 2015
Accepted 26th October 2015
2
sulfonated titania/mordenite solid acid, the K1.5Mg/SiO catalyst presented heat-stability and prolonged
activity in the one-pot conversion of glucose to HMF, leading to an optimal yield of 62.2%. The one-pot
catalysis of saccharides from the enzymatic hydrolysate of poplar sawdust achieved a high HMF yield of
56.9% at a glucose conversion of 91.6%.
DOI: 10.1039/c5ra17132f
www.rsc.org/advances
from the hydrolysate with a broad range of compounds having
inhibitory effects on the microorganisms for subsequent
Introduction
Current agricultural and forestry residues and municipal solid fermentation, the pentoses being difficult to ferment, and least
waste including abundant lignocellulosic biomass could exibility to treat different raw materials. For these reasons,
provide low cost and sustainable resources for the production of there is a great interest in developing new catalytic technolo-
1–3
13–15
various valuable chemicals containing HMF. Among these gies.
The route to effectively hydrolyze lignocellulosic
options, intensively-grown poplars have gained great attention biomass by integrating solid base and acid catalysts could easily
in the midwestern United States due to continuously increasing overcome above-mentioned problems.
4
,5
productivity and cold/drought tolerance. Enzymatic hydro-
As an important intermediate, HMF can be produce from
16–18
lysis of poplars can produce various sugars, followed by glucose by heterogenous catalysis,
which would be used for
fermentation to fuel-grade ethanol. The treatment with biodegradable surfactants, pharmaceuticals, furanic polyesters,
6,7
19–22
enzymes, bacteria, and fungi took place in mild conditions polyamides, and biofuels.
However, glucose only contained
and spotlighted on their potential to simplify and improve key 1% of furanose tautomers in aqueous solutions, which gave
process steps via modern biotechnology. In fact, as main poor HMF yield in direct dehydration due to side reactions.
components of lignocellulosic biomass, hemicellulose was Integrating isomerization of hexose to ketose and subsequent
formed by the linkage of ve different sugar monomers of D- ketose dehydration to HMF was considered as an effective
8,9
xylose, L-arabinose, D-galactose, D-mannose, and D-glucose,
and cellulose consisted of D-glucose linked by 1,4-b-glycosidic combining metal chloride salts (e.g., AlCl
bonds.
method. For this purpose, some research teams focused on
, CrCl , SnCl , LaCl
The pretreatment operation, such as steam explo- with a Brønsted acid (e.g. HCl) to perform the glucose
sion, ammonia ber expansion, aqueous ammonia recycle, isomerization/dehydration reactions, leading to high HMF yield
3
3
4
3
)
10,11
12
23,24
dilute sulfuric acid, lime, and sulte had to be applied prior to of 67 mol%.
However, metal halides were generally inactive
the biological steps to overcome the natural recalcitrance from in neutral aqueous solution, the use of strong mineral acids
lignin sheath and cellulose. However, several problems derived incurred environmental problems, and the separation of HMF
was difficult. Interestingly, Davis group found that a Sn-beta
zeolite could isomerize glucose to fructose with high activity
25,26
a
in water under acidic conditions.
The unique performance
Bioproducts Science and Engineering Laboratory, Washington State University,
Richland, WA 99352, USA. E-mail: bka@tricity.wsu.edu; hlei@tricity.wsu.edu; Fax: of the Sn-beta solid Lewis acid created new opportunities for the
+
1 5093726782; Tel: +1 5093726782
facile integration of isomerization with dehydration steps.
Nikolla group developed the synergistic catalysis of Sn-beta
with HCl for glucose conversion in a water/butanol biphasic
system, leading to 55 mol% HMF selectivity under a glucose
conversion of 75%. However, Sn-beta is not a commercially
b
Engineering Research Center of Nano-Geo Materials of Ministry of Education, China
27
University of Geosciences, Wuhan 430074, China. E-mail: xiang.shen@tricity.wsu.edu
c
State Key Laboratory of Biogeology and Environmental Geology
& School of
Environmental Studies, China University of Geosciences, Wuhan 430074, China.
E-mail: yx.wang@cug.edu.cn
96990 | RSC Adv., 2015, 5, 96990–96996
This journal is © The Royal Society of Chemistry 2015

View Article Online
Paper
RSC Advances
available zeolite and complex preparation leading to high cost. drying at the same temperature by the addition of 2 mL
In addition, the enzymatic hydrolysate of poplar sawdust hydrouoric acid (40%). At last, 2 mL nitric acid was added to
included glucose, cellobiose, xylose, mannose, and galactose; dissolve the residue. As an analytical sample, 0.1 mL of the
efficient conversion of these saccharides to HMF will be digestion solution was suctioned with a pipette into a volu-
a substantial challenge.
metric ask to 25 mL by addition of ultra pure water. Blank tests
Given the fact, the work focused on the aldose–ketose and parallel experiments were conducted to improve the
isomerization using low leaching solid catalysts for improving measurement accuracy and precision. Thermogravimetric
the selectivity of HMF in a one-pot system. A combination of analysis (TG) and differential thermal analysis (DTA) was
sulfonated titania/mordenite (TM) with different K
catalysts could afford high HMF yield under a moderate reac- from 25 to 500 C at a ramp of 10 C min under a ow of N for
x
Mg/SiO
2
carried out on a PerkinElmer Diamond TG/DTA instrument
ꢁ
ꢁ
ꢀ1
2
tion temperature. The one-pot reaction using heterogeneous evaluating the thermal stability of prepared catalysts.
catalysts would provide the best solution to process the complex
saccharides from poplar sawdust.
Catalytic tests
In order to achieve fast separation of product and reactants, one-
pot catalysis of saccharides from the enzymatic hydrolysate of
Experimental
Preparation and characterization of solid catalysts
poplar sawdust was performed in water/butanol biphase system
using a 250 mL thick-walled glass reactor under nitrogen (20 mL
min ) with a reux-condenser. An oil bath was used to heat the
reactor rapidly to the desired reaction temperature, with
temperature and stirring being controlled by a digital laboratory
hot plate magnetic stirrer (IKA™ RCT Basic IKAMAG™ Safety
The optimal titania/mordenite solid-acid was prepared by sol–
gel-hydrothermal method according to our previous report.
The sulfonated titania/mordenite catalysts were denoted as
TM0.2–0.5 (ratio of mordenite to titania is 20, 30, 40 and 50 wt%,
respectively).
ꢀ
1
28
x 4
Control). In a typical experiment, K MgSiO catalyst of 0.4 g and
The K MgSiO catalysts were prepared by precipitation-gel
x
4
10 mL saccharides solution (or 0.6 g glucose dissolved in 10 mL
method. In a typical procedure, 8.8 g of Mg(NO
.3 g CTAB was dissolved with 10 mL mixed solution of deion-
ized water and ethanol (50% by volume). The solution was
added into 20 mL acidic industrial silica sol (31 wt% SiO , pH
3 2 2
) $6H O and
saturated sodium chloride solution) were transferred to the
reactor and then poured 30 mL butanol into it. The reactor was
0
ꢁ
immersed in the oil bath of 120 C and stirred at 600 rpm for 1 h.
2
ꢁ
Next, 0.6 g solid-acid was added to the reaction system at 140 C
of 3) under vigorous stirring for 15 minutes. The stoichiometric
potassium carbonate was dissolved with about 10 mL deionized
water and added into the above mixed sol under homoge-
neously stirring for 30 minutes. The forming gel was aged
overnight at 403 K and then washed with ethanol. The resultant
precursors were dried and calcined at 973 K for 2 h. Thermally
for 2 h. The reaction was quenched by soaking the reactor in ice
water. The resultant aqueous and organic phases were separated
for subsequent analysis. The determination of HMF, organic
acids and saccharides was performed by Dionex UltiMate 3000
high performance liquid chromatography system (HPLC)
equipped with Aminex HPX-87H Column (9 mm, 300 ꢂ 7.8 mm)
x 2
activated catalysts were denoted as K Mg/SiO (as K/Si molar
ꢁ
and Dionex Refractive Index (RI) detector at 65 C. The mobile
ratio, x ¼ 1.0, 1.2, 1.5, 1.7).
ꢀ1
phase was 5 mM sulfuric acid at a ow rate of 0.6 mL min
.
The microstructure of prepared solid catalysts was analyzed
Each sample was diluted with ultrapure water of pH 5–7 before
analysis. Glucose conversion and HMF yield were calculated
from the product of the aqueous and organic phase concentra-
tion obtained by above test and their corresponding volumes
aer reaction since the value of Vorg/Vaq was changed aer
reaction. The conversion of glucose was dened as the moles of
glucose reacted divided by the moles of initial glucose; the
selectivity of HMF was dened as the moles of HMF produced
divided by the moles of glucose reacted; and the yield of HMF
was dened by integrating the product of the glucose conversion
ꢀ1 and the HMF selectivity.
by a FEI Sirion 200 eld emission scanning electron microscope
(
(
FESEM), FEI Tecnai G2 20 Transmission Electron Microscope
TEM) and X'Pert PRO DY2198 X-ray diffractometer (XRD). The
mean crystallite size of samples was calculated by the XRD line
width of 220 peak using the Scherrer formula. The alkalinity of
samples was studied with TPD using CO
Each sample was treated at 573 K with N
for 1 h and then in a reacting CO ow of 15 mL min at 323 K
for 30 min. The physical adsorbed CO was eliminated by pure
nitrogen. The oven temperature of the TPD was programmed
2
as molecular probe.
ꢀ1
2
ow of 30 mL min
ꢀ
1
2
2
ꢁ
ꢀ1
from 323 K to 1073 K at 10 C min in He ow of 15 mL min
and desorbed CO was detected using TP-5075 apparatus
Tianjin Xianquan Co.) with thermal conductivity detector. The
2
(
Results and discussion
potassium and magnesium analyses were carried out by using
inductively coupled palama-atomic emission spectrometer
Composition and surface properties of solid alkali
2 4
(Thermo IRIS Interpid XSP ICP-AES, USA). Refer to digestion The characteristic diffraction peaks of crystalline K MgSiO ,
procedure, about 0.1 g catalyst sample was transferred into (220), (311), (400) and (422) at ca. 2q ¼ 32.77, 38.65, 47, 58.41,
a Teon vessel containing 5 mL hydrochloric acid (37%) and respectively, were observed in the X-ray patterns (Fig. 1). As the
ꢁ
heated on a hot plate at 120 C until the residual liquid was 2–3 K/Si mole ratio increased, the peak of amorphous silica dimin-
2 4
mL. Next, 5 mL nitric acid (69%) was added into the Teon ished whereas K MgSiO characteristic peaks signicantly
vessel to obtain colorless solution by evaporation, followed by enhance, implying a change in composition and structure of
This journal is © The Royal Society of Chemistry 2015
RSC Adv., 2015, 5, 96990–96996 | 96991

View Article Online
RSC Advances
Paper
catalysts due to the addition of potassium carbonate. Another, it
seemed that small amount of active component of the catalysts
was KNO ((021), (012), (112), (221), (041) and (313) at ca. 2q ¼
3
2
3.92, 29.49, 33.93, 41.3, 41.97, 68.76) and K O ((220) and (700) at
2
ca. 2q ¼ 19.81, 40.44). This could be because the clustering or
unevenly dispersed KNO on the catalyst formed in the hydro-
thermal process and only part of the loaded KNO decomposed
into K
3
3
2
O ((220), (700) at ca. 2q ¼ 19.81, 40.44) under the thermal
29
activation condition. The characteristic diffraction peaks of
silica and magnesia were not found, implying the formation of
new active component of K MgSiO by the precipitation-gel
2
4
process.
To provide additional evidence, morphologies of prepared
catalysts were analyzed. The TEM images of K1.5Mg/SiO showed
4
rodlike crystal (Fig. 2f–h) and exhibited small round-shape
particles on the surface (Fig. 2e). Comparing with Fig. 2c and i
scanning micrograph, it could be seen that the crystalline
particles decreased gradually with the increase of potassium
amount in prepared catalysts. Electron micrographs (Fig. 2b, d, f
and k) showed that the K1.5Mg/SiO catalysts presented a cluster
2
of well-developed rodlike structure with a certain length–diam-
eter ratio different from other catalysts, implying a difference in
catalytic activity owing to their different composition. And the
2
K1.5Mg/SiO catalysts can be well dispersed in reaction medium,
which would be benecial to the one-pot reaction. We noted that
the K1.5Mg/SiO catalyst with lower amount led to incomplete
2
crystallization and an unclear image (Fig. 2a), implying an
unmatched synergism between the active components and the
silica support. Another, too much potassium loading could
increase the moisture absorbency of prepared K1.7Mg/SiO
catalysts leading to agglomerated particles.
2
2
Advancing a step, we investigated CO -TPD proles (Fig. 3) to
validate the correlation between composition of prepared
catalysts and their alkalinities. The broadening peaks corre-
sponding to different temperature demonstrate the different
basic sites on the surface of catalysts. The desorption peaks at
about 373 K or 573 K demonstrated the moderate basic sites on
Fig. 2 FESEM images of prepared solid catalysts: (a) K1.5Mg/SiO
a lower magnesium amount, (b) K1.0Mg/SiO , (c) K1.2Mg/SiO
2
with
, (e)
2
2
K
K
1.5Mg/SiO
1.2Mg/SiO
2
and (i) K1.7Mg/SiO
, ((f) 100 nm, (g) 50 nm, (h) 200 nm) K1.5Mg/SiO
2
; TEM (nm scale) images: ((d) 100 nm)
and ((k)
2
2
1
2
00 nm) K1.7Mg/SiO .
the catalysts, which are associated with the surface OH– groups
Fig. 1 XRD patterns of prepared solid bases: (a) K1.0Mg/SiO
2
, (b)
(A)
30,31
adsorbed on the K–O or Mg–O species.
2 4
The K MgSiO
K
1.2Mg/SiO
2
,
(c)
K
1.5Mg/SiO
2
,
(d)
K
1.7Mg/SiO
2
;
(>) KNO
3
,
K
2
MgSiO , (ò) K
4
2
O.
component of catalyst would be the source of the moderate
96992 | RSC Adv., 2015, 5, 96990–96996
This journal is © The Royal Society of Chemistry 2015

View Article Online
Paper
RSC Advances
Fig. 4 Composition of products resulted from the conversion of
saccharides from poplar sawdust by employing AlCl with different TM
3
Fig. 3 CO
2
-TPD profiles of prepared solid catalysts: (a) K1.0Mg/SiO
2
,
(
b) K1.2Mg/SiO
2
, (c) K1.5Mg/SiO , (d) K1.7Mg/SiO
2
2
.
solid acids; reaction conditions:10 mL of saccharides solution (pH z 5)
ꢁ
at 140 C for 180 min using 4 wt% anhydrous AlCl
catalysts; organic to aqueous volume ratio of 3.
3
and 6 wt% TM
basicity, which would induce excellent catalytic activity. Never-
theless, excess KNO adsorbed on the catalyst could increase the
moisture absorbency of prepared K1.7Mg/SiO catalysts leading
to agglomerated particles, which is unfavourable to the isom-
erization reaction.
3
2
relatively low concentration, which could improve the interac-
tion of active sites on catalysts with reactants. Considering the
similar conditions, the catalytic activity corresponding to HMF
yield increased as follows: K1.0Mg/SiO
2 2
< K1.2Mg/SiO < K1.7Mg/
Catalytic activities
SiO < K1.5Mg/SiO . Therefore, it was reasonable to assume that
2
2
ꢁ
Saccharides conversion studies were carried out at 140 C by the moderate base sites played a signicant role in determining
integrating TM solid-acid and K
butanol biphase system under the atmosphere of nitrogen.
The composition of saccharides solution (pH of ca. 5) from HMF by integrating different K Mg/SiO catalyst with TM0.3
x
Mg/SiO
2
catalysts in a water/ the conversion of hexoses to HMF.
Furthermore, we conducted the conversion of glucose to
x
2
enzymatic hydrolysate of poplar sawdust included glucose of catalysts (Fig. 5). The HMF selectivity increased with increasing
ꢀ
1
ꢀ1
ꢀ1
4
6.96 g L , cellobiose of 3.81 g L , xylose of 6.49 g L
,
glucose conversions, and a maximum of 65.98% with K1.5Mg/
mannose of 1.11 g L and galactose of 0.87 g L . Several SiO and TM0.3 catalysts at a glucose conversion of 94.32% was
chemicals such as acetic acid of 14.88 g L and furfural of achieved. This coincided with the trend that was for catalyst
ꢀ1
ꢀ1
2
ꢀ1
ꢀ
1
2
.26 g L from the enzymatic process were also found. Fig. 4 with relatively high base sites and Brønsted acid sites. Addi-
depicts distribution of products for the conversion of saccha- tionally, the high selectivity was attributed to the mesoporous
rides to HMF by combining anhydrous AlCl with different TM structure of solid catalysts, which increased active sites of inner
3
catalysts. It was noted that the glucose conversion was low, surface of catalysts. It was noted that lactic acid as a by-product
which could be resulted from inactive AlCl
oligosaccharides coating on the surface of catalysts. In the of K
3
under pH of 5 and was observed for all experiments due to the moderate base sites
32,33
x
Mg/SiO
2
catalyst.
2 4
The catalyst with K MgSiO compo-
presence of the same Lewis acid, TM0.3 catalysts showed nent could have potential for the synthesis of lactic acid from
a higher activity comparing to other TM catalysts. It also found glucose. From Table 1, it can be seen that glucose can be con-
that nearly all arabinose, galactose, and furfural disappeared verted in the presence of the K1.5Mg/SiO catalyst, and the
2
due to forming water soluble and humic species. Interestingly, selectivity of HMF was only 7% at 88% glucose conversion. In
amount of acetic acid dropped sharply, indicating that acetic the absence of either the K1.5Mg/SiO catalyst or TM solid-
2
0.3
acid could facilitate the side reactions leading to humic and acid, the yield of HMF was low (run 5 and 6). It could mean
oligomeric species (Fig. 4). that the K1.5Mg/SiO catalyst was in favor of the conversion of
2
In order to mitigate the negative effect of oligosaccharides glucose, and TM0.3 solid-acids with mesoporous structure hel-
and acetic acid, we chose the low concentration of enzymatic ped improve the selectivity of HMF. The surface of TM0.3 cata-
4
+
hydrolysate for the same reaction using TM0.3 solid-acid and lyst was in the presence of Lewis acid sites due to Ti and Si–O–
different K Mg/SiO catalysts. The composition of saccharides Al bond of mordenite skeleton, and Brønsted acid sites were
solution included glucose of 19.57 g L , acetic acid of 3.59 g due to the sulphation process. The Lewis acid and Brønsted acid
x
2
ꢀ
1
ꢀ
1
ꢀ1
ꢀ1
L
0
, cellobiose of 1.44 g L , xylose of 3.28 g L , mannose of sites facilitated the dehydration reaction of fructose. The addi-
ꢀ
1
ꢀ1
.86 g L and furfural of 1.32 g L . The yield of HMF was tion of K Mg/SiO catalysts improved signicantly the conver-
x 2
signicantly improved (Table 1). Moreover, the competitive sion of glucose, as shown in runs 7–10, which led to an optimal
adsorption was weakened due to less reactant molecules at HMF yield of 62.24% with a considerable reaction time. The
This journal is © The Royal Society of Chemistry 2015
RSC Adv., 2015, 5, 96990–96996 | 96993

View Article Online
RSC Advances
Paper
Table 1 Results of conversion of glucose and saccharides to HMF in a biphasic system
Conversion of
glucose (mole%)
Yield of HMF
(mole%)
Selectivity
of HMF (mole%)
Run
Catalyst
Feedstock
1
2
3
4
5
6
7
8
9
K
K
K
K
K
1.5Mg/SiO
1.7Mg/SiO
1.2Mg/SiO
1.0Mg/SiO
1.5Mg/SiO
2
2
2
2
2
+ TM0.3
+ TM0.3
+ TM0.3
+ TM0.3
Enzymatic hydrolyzate
Enzymatic hydrolyzate
Enzymatic hydrolyzate
Enzymatic hydrolyzate
Glucose
Glucose
Glucose
Glucose
Glucose
91.62
81.45
76.14
52.78
88.17
35.14
94.32
90.05
83.53
74.4
56.94
47.59
32.78
33.36
7.58
16.98
62.24
53
62.15
58.43
43.05
63.2
8.6
TM0.3
48.32
65.98
58.86
45.09
49.38
49.22
60.62
K
K
K
K
K
K
1.5Mg/SiO
2
2
2
2
2
2
+ TM0.3
+ TM0.3
+ TM0.3
+ TM0.3a
+ TM0.3
1.7Mg/SiO
1.2Mg/SiO
1.0Mg/SiO
1.5Mg/SiO
1.5Mg/SiO
39.17
36.74
34.6
1
1
1
0
1
2
Glucose
Glucose
Glucose
70.3
96.5
b
+ TM0.3
58.5
a
b
Amount of K1.5Mg/SiO
2
catalyst was set to 0.3 g. Amount of K1.5Mg/SiO
2
catalyst was set to 0.5 g.
Table 2 ICP-AES analysis of the fresh and used K1.5Mg/SiO
2
catalyst
saccharides through a retro-aldol condensation reaction. In
a subsequent step glyceraldehyde was dehydrated to pyr-
uvaldehyde, which subsequently rearranged to lactic acid.
K (ppm)
Leaching ratio Mg (ppm) Leaching ratio
K
1.5Mg/SiO
2
66.72 ꢃ 0.5
13.36
(fresh)
The stability of K1.5Mg/SiO
The leaching test was designed to verify the stability of bonding
potassium and magnesium of K1.5Mg/SiO catalyst. In agree-
2
catalyst
KMg-1
KMg-2
KMg-3
KMg-4
KMg-5
2.51 ꢃ 0.5 3.76%
1.79 ꢃ 0.5 2.68%
0.87 ꢃ 0.5 1.3%
0.38 ꢃ 0.5 0.57%
0.26 ꢃ 0.5 0.39%
60.96 ꢃ 0.5
0.019
0.016
0.007
0.007
—
0.14%
0.12%
0.05%
0.05%
—
2
ment with reaction conditions, 0.4 g catalyst was added into the
ꢁ
K1.5Mg/SiO
2
13.25
water/butanol biphase system and reuxed at 140 C for
(used)
120 min. Aer reaction, water phase was separated from the
reactor for a following analysis (mark as KMg-1). The solid
catalyst was retained in the reactor for subsequent four leaching
experiments according to the same procedure and the water
phase samples of each step was marked as KMg-2, KMg-3, KMg-
2
effectiveness of the K1.5Mg/SiO catalyst for the conversion of
glucose to fructose can be understood in terms of the intrinsic
properties of base sites and dispersed rodlike microstructure.
The product distribution was also inuenced by the amount of
4
and KMg-5, respectively. Finally, the used catalyst was recov-
ered for digestion and analysis. The amount of potassium and
the K1.5Mg/SiO
catalyst from 0.4 g to 0.3 g resulted in a declining yield of HMF
34%) with 49% selectivity, ascribed to a decrease in fructose
formation by base-catalyzed isomerization (run 11). In contrast,
increasing the amount of the K1.5Mg/SiO catalyst to 0.5 g
2 2
catalyst. Decreasing the amount of K1.5Mg/SiO
(
2
improved the HMF yield to 58% with 60% selectivity (run 12). It
was very interesting to note that the moderate base sites and
anisotropic structure would promote the aldose–ketose isom-
erization and the formation of lactic acid. Initially, the forma-
tion of D-fructose and D-mannose was observed. It was possible
that the K1.5Mg/SiO
2
catalyst or Lewis acid sites promoted the
isomerisation of D-glucose to the intermediate of D-mannose
34
and D-fructose. The formation of HMF could follow a mecha-
nism, in which the open-chain form of fructose was dehydrated
at the C-2 position using TM0.3 solid acid, forming a carbocation
which reacted with the hydroxyl group at C-5 position, forming
furaldehyde intermediate followed by further dehydration to
form HMF. Lactic acid, the main by-product, was likely formed
from the hexoses via a complex reaction pathway without the
Fig. 5 Content of HMF and lactic acid resulted from the conversion of
glucose by using different K
x
Mg/SiO
2
catalyst and TM0.3 solid acids;
32,35
formation of furaldehyde intermediate.
The rst step was
ꢁ
reaction conditions: 6 wt% glucose aqueous solution at 140 C for 180
believed to consist of the formation of glyceraldehyde from the min using 4 wt% K Mg/SiO2 catalyst and 6 wt% TM0.3 catalysts, and
x
organic to aqueous volume ratio of 3.
96994 | RSC Adv., 2015, 5, 96990–96996
This journal is © The Royal Society of Chemistry 2015

View Article Online
Paper
RSC Advances
mesoporous structure and strong Brønsted acid sites helped
improve the selectivity of HMF. The one-pot catalysis of
saccharides from enzymatic hydrolysate of poplar sawdust was
in favor of low substrate concentration to avoid the competitive
adsorption of oligosaccharides on solid catalysts. The optimal
HMF yield of 56.9% at a glucose conversion of 91.6% was ach-
ieved for the complex saccharides system, which represented
a considerable step toward the goal of developing a sustainable
process for lignocelluloses conversion.
Acknowledgements
This work is nancially supported by the National Natural
Science Foundation of China (40830748), the China Post-
doctoral Science Foundation (201104497, 20090451092), the
open foundation (CHCL10003) from MOE Key Laboratory of
Catalysis and Materials Science of the State Ethnic Affairs
Commission of Hubei in China and the Fundamental Research
Founds for National University, China University of Geosciences
2
Fig. 6 TG-DTA curves of K1.5Mg/SiO catalyst.
magnesium of prepared liquid samples were determined by
ICP-AES. The analysis results (Table 2) showed that the content
of K of the K1.5Mg/SiO catalyst decreased slightly aer each
2
leaching test and the leaching ratio of potassium was less than
(Wuhan) (Innovative Team, Grant CUG120115).
0
.5% in the h cycle experiment. And the content of Mg
maintained a constant aer each leaching test, implying the
stability of the K1.5Mg/SiO
2
catalyst. The low leaching ratio was Notes and references
correlated to the good reactivity of prepared catalyst corre-
sponding to the high yield.
1
2
3
M. FitzPatrick, P. Champagne, M. F. Cunningham and
R. A. Whitney, Bioresour. Technol., 2010, 101, 8915–8922.
J. N. Chheda, G. W. Huber and J. A. Dumesic, Angew. Chem.,
Int. Ed., 2007, 46, 7164–7183.
G. Centi, P. Lanzafame and S. Perathoner, Catal. Today,
2011, 167, 14–30.
Further, the thermal behavior of the dried K1.5Mg/SiO2
catalyst was shown in Fig. 6. Several ambiguous endothermic
peaks were detected from on the DTA curve. The rst endo-
thermic peak is due to the release of adsorbed water of the
ꢁ
catalyst sample around 150 C resulting in 1.76% weight loss.
ꢁ
4 R. S. Zalesny, R. B. Hall, J. A. Zalesny, B. G. McMahon,
W. E. Berguson and G. R. Stanosz, BioEnergy Res., 2009, 2,
106–122.
The second peak at ca. 380 C is attributed to the decomposition
of the small amount of potassium nitrate. From TG plot, the
weight remained almost unchanged above 600 C, indicating
the stability of the structure of solid catalyst.
ꢁ
5
Z. J. Wang, J. Y. Zhu, R. S. Zalesny and K. F. Chen, Fuel, 2012,
5, 606–614.
9
Solid catalyst's stability played an important role in deter-
mining the economical application for large-scale production.
6
M. Cantarella, L. Cantarella, A. Gallifuoco, A. Spera and
F. Alfani, Biotechnol. Prog., 2004, 20, 200–206.
36
The previous work reected massive potassium salt was
soluble in the water of ethanol system and potassium leaching
from the catalyst resulted in reducing the activity of catalyst.
Here, using the silica gel as supporter is necessary to stabilize
the active sites and to prevent the leaching procedure. The
excellent catalytic properties conrmed that the components of
the solid catalyst are synergetically assembled by the
precipitation-gel-thermally activated process.
7
8
M. K. Bhat, Biotechnol. Adv., 2000, 18, 355–383.
P. Maki-Arvela, T. Salmi, B. Holmbom, S. Willfor and
D. Y. Murzin, Chem. Rev., 2011, 111, 5638–5666.
9
P. L. Dhepe and R. Sahu, Green Chem., 2010, 12, 2153–2156.
1
1
1
1
0 Z. Fang, F. Zhang, H. Y. Zeng and F. Guo, Bioresour. Technol.,
2011, 102, 8017–8021.
1 M. X. Cheng, T. Shi, H. Y. Guan, S. T. Wang, X. H. Wang and
Z. J. Jiang, Appl. Catal., B, 2011, 107, 104–109.
2 R. Kumar, G. Mago, V. Balan and C. E. Wyman, Bioresour.
Technol., 2009, 100, 3948–3962.
3 Z. H. Zhang and Z. B. Zhao, Bioresour. Technol., 2011, 102,
3970–3972.
Conclusions
Cost-effective conversion of saccharides from poplar sawdust to
HMF was successfully conducted using a biphasic system in the 14 G. Gliozzi, A. Innorta, A. Mancini, R. Bortolo, C. Perego,
presence of TM0.3 solid-acid and K1.5Mg/SiO catalyst at M. Ricci and F. Cavani, Appl. Catal., B, 2014, 145, 24–33.
a moderate reaction temperature. The integrating synthetic 15 S. De, S. Dutta, A. K. Parra, B. S. Rana, A. K. Sinha, B. Saha
methodology offered the opportunity to utilize acid or base sites and A. Bhaumik, Appl. Catal., A, 2012, 435, 197–203.
separately in a one-pot reaction system. The low leaching 16 S. Zhao, M. X. Cheng, J. Z. Li, J. A. Tian and X. H. Wang,
K1.5Mg/SiO catalyst with moderate base sites could promote Chem. Commun., 2011, 47, 2176–2178.
the aldose–ketose isomerization of glucose and has potential 17 H. F. Xiong, T. F. Wang, B. H. Shanks and A. K. Datye, Catal.
2
2
for the synthesis of lactic acid. The TM0.3 solid-acids with
Lett., 2013, 143, 509–516.
This journal is © The Royal Society of Chemistry 2015
RSC Adv., 2015, 5, 96990–96996 | 96995

View Article Online
RSC Advances
Paper
1
1
2
2
2
8 J. H. He, Y. T. Zhang and E. Y. X. Chen, ChemSusChem, 2013, 27 E. Nikolla, Y. Roman-Leshkov, M. Moliner and M. E. Davis,
, 61–64. ACS Catal., 2011, 1, 408–410.
9 Y. Roman-Leshkov, C. J. Barrett, Z. Y. Liu and J. A. Dumesic, 28 X. Shen, Y. X. Wang, C. W. Hu, K. Qian, Z. Ji and M. Jin,
Nature, 2007, 447, 982–985. ChemCatChem, 2012, 4, 2013–2019.
0 J. P. Ma, Z. T. Du, J. Xu, Q. H. Chu and Y. Pang, 29 M. S. Kotwal, P. S. Niphadkar, S. S. Deshpande, V. V. Bokade
ChemSusChem, 2011, 4, 51–54. and P. N. Joshi, Fuel, 2009, 88, 1773–1778.
1 J. M. R. Gallo, D. M. Alonso, M. A. Mellmer and J. A. Dumesic, 30 D. Wang, X. Zhang, W. Wei and Y. Sun, Catal. Commun.,
Green Chem., 2013, 15, 85–90. 2012, 28, 159–162.
2 T. Buntara, S. Noel, P. H. Phua, I. Melian-Cabrera, J. G. de 31 J. Di Cosimo, V. Dıez, M. Xu, E. Iglesia and C. Apesteguıa, J.
6
Vries and H. J. Heeres, Angew. Chem., Int. Ed., 2011, 50,
083–7087.
3 Y. Yang, C. W. Hu and M. M. Abu-Omar, Green Chem., 2012,
4, 509–513.
4 Y. J. Pagan-Torres, T. F. Wang, J. M. R. Gallo, B. H. Shanks
and J. A. Dumesic, ACS Catal., 2012, 2, 930–934.
5 Y. Roman-Leshkov, M. Moliner, J. A. Labinger and
M. E. Davis, Angew. Chem., Int. Ed., 2010, 49, 8954–8957.
6 M. Moliner, Y. Roman-Leshkov and M. E. Davis, Proc. Natl.
Acad. Sci. U. S. A., 2010, 107, 6164–6168.
Catal., 1998, 178, 499–510.
32 X. Y. Yan, F. M. Jin, K. Tohji, T. Moriya and H. Enomoto, J.
Mater. Sci., 2007, 42, 9995–9999.
33 A. Onda, T. Ochi, K. Kajiyoshi and K. Yanagisawa, Catal.
Commun., 2008, 9, 1050–1053.
34 M. Watanabe, Y. Aizawa, T. Iida, R. Nishimura and
H. Inomata, Appl. Catal., A, 2005, 295, 150–156.
35 M. Bicker, S. Endres, L. Ott and H. Vogel, J. Mol. Catal. A:
Chem., 2005, 239, 151–157.
7
2
2
2
2
1
36 G. Arzamendi, E. Argui n˜ arena, I. Campo, S. Zabala and
L. M. Gand ´ı a, Catal. Today, 2008, 133–135, 305–313.
96996 | RSC Adv., 2015, 5, 96990–96996
This journal is © The Royal Society of Chemistry 2015