Synthesis of furfural from xylose by heterogeneous and reusable nafion catalysts

Article abstract of DOI:10.1002/cssc.201100023

Nafion 117 has been proven as a robust and reusable heterogeneous catalyst for the dehydration of 9.1% (w/w) xylose in dimethyl sulfoxide (DMSO) to yield 60% furfural in 2 h at 150°C. The catalytic high activity promoted shorter reaction times to limit the formation of side-products which otherwise would lead to decreased yields. Within the allowable operating temperature range of Nafion (125 to 175°C), the reaction was kinetically controlled. In corroboration with AFM and SEM imaging, ATR-FTIR confirmed that the Nafion catalytic activity remained unchanged after 15 repeated uses. With excellent chemical and thermal stability under the conditions for xylose dehydration compared to existing solid acid catalysts, this reusable Nafion system could be a step towards the more economical production of furfural from renewable biomass, an intermediate chemical for the preparation of value-added chemicals.

Full text of DOI:10.1002/cssc.201100023

DOI: 10.1002/cssc.201100023
Synthesis of Furfural from Xylose by Heterogeneous and
Reusable Nafion Catalysts
Edmond Lam, Ehsan Majid, Alfred C. W. Leung, Jonathan H. Chong, Khaled A. Mahmoud,
[a]
and John H. T. Luong*
Nafion 117 has been proven as a robust and reusable hetero-
geneous catalyst for the dehydration of 9.1% (w/w) xylose in
dimethyl sulfoxide (DMSO) to yield 60% furfural in 2 h at
imaging, ATR–FTIR confirmed that the Nafion catalytic activity
remained unchanged after 15 repeated uses. With excellent
chemical and thermal stability under the conditions for xylose
dehydration compared to existing solid acid catalysts, this re-
usable Nafion system could be a step towards the more eco-
nomical production of furfural from renewable biomass, an in-
termediate chemical for the preparation of value-added chemi-
cals.
1508C. The catalytic high activity promoted shorter reaction
times to limit the formation of side-products which otherwise
would lead to decreased yields. Within the allowable operating
temperature range of Nafion (125 to 1758C), the reaction was
kinetically controlled. In corroboration with AFM and SEM
Introduction
Owing to diminishing petrochemical resources, focus has
turned to abundant biomass sources such as agricultural
crops, forestry by-products and plant wastes rich in hemicellu-
lose. Hemicellulose is a branched, heterogeneous polymer that
comprises a backbone of pentose sugars such as xylose and
mannose. Hemicellulose derived from hardwood trees and
grass tends to be rich in xylose, while hemicellulose derived
from softwood trees is rich in mannose. Organic acids and
other sugars (such as arabinose, glucose, and rhamnose) can
substitute for some of the pentose sugars in the hemicellulose
and HCl) that are highly toxic, corrosive, and difficult to
handle. Other major drawbacks to mineral acids include: the
occurrence of extensive side reactions from long reaction
times that reduce furfural yield; neutralization of the spent
acid; and technological challenges in the development, use,
and maintenance of equipment exposed to the corrosive liq-
uids. As part of a four-step process for the formation of jet-
[3]
and diesel-fuel-range alkanes, Xing et al. reported the synthe-
sis of furfural from HCl-catalyzed biphasic dehydration of hemi-
cellulose extracts at 87% yield. However, they acknowledge
that the high cost of producing such alkanes could be reduced
by using solid acids in place of HCl.
[
1]
backbone.
Hemicellulose-rich biomass can serve as alternative sustaina-
ble supplies for non-petroleum-derived chemicals. Furfural, an
intermediate for the preparation of value-added products in-
cluding plastics (furan-based polymers), pharmaceuticals, fuels,
Various materials have been tried as alternatives for the non-
recoverable homogeneous catalysts currently used in furfural
production. These heterogeneous catalysts are often solid
acids characterized by strong Brønsted acidity, high surface
area, and are thermally stable at the high temperatures re-
quired for rapid xylose dehydration to furfural. Zeolites gener-
[
2]
and fine chemicals, is produced from the acid-catalyzed de-
hydration of hemicellulose (Scheme 1).
To date, both batchwise and continuous acid hydrolysis are
able to hydrolyze hemicellulosics to xylose, followed by dehy-
dration of the latter to form furfural. At 170–1858C, the reac-
[4–6]
ally produce lower reaction yields
whereas heteropolyacids
require lower substrate-to-catalyst loading to obtain moderate
[
2]
[7]
tion is complete within 3 h, yielding 40–50% furfural. Current
yields (63%). Sulfonic-acid-functionalized MCMs (Mobil cata-
commercial processes use concentrated inorganic acids (H SO4
lytic materials) have 91% xylose conversion and 82% furfural
selectivity, but require 24 h of reaction with diminishing cata-
2
[8]
lyst activity for consecutive runs. In all cases, xylose dehydra-
tion is carried out at lower initial xylose concentrations (<
3.3 wt%), requiring the use of large solvent volumes on an in-
dustrial scale and costly isolation and distillation of furfural.
[
a] Dr. E. Lam, E. Majid, Dr. A. C. W. Leung, Dr. J. H. Chong, Dr. K. A. Mahmoud,
Prof. Dr. J. H. T. Luong
Biotechnology Research Institute
National Research Council Canada
6
100 Royalmount Avenue, Montreal, QC H4P 2R2 (Canada)
Fax: (+1)514 496-6265
Scheme 1. Transformation of biomass into value-added chemical products.
E-mail: john.luong@cnrc-nrc.gc.ca
ChemSusChem 2011, 4, 535 – 541
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
535

J. H. T. Luong et al.
There is an urgent need to improve the cost-effectiveness of
furan-based chemical industries.
This study demonstrates and optimizes the catalytic perfor-
mance and reusability of Nafion, a sulfonated tetrafluoroethy-
[
9]
lene-based fluoropolymer-copolymer. Nafion has been known
to withstand high temperatures (up to 1908C) without any sig-
nificant chemical degradation. As a superacid, Nafion has been
used to catalyze various organic transformations including acy-
[
10]
lation, alkylation, and isomerization reactions. Indeed, Nafion
has been used in the dehydration of fructose (hexose sugar) to
[11]
produce hydroxymethylfurfural (HMF) with a 50% yield. To
our knowledge, this study is the first systematic attempt to
convert concentrated xylose (pentose sugar) to furfural using
Nafion as a heterogeneous and reusable catalyst to facilitate
product recovery. Mechanistic considerations towards selectivi-
ty and catalyst poisoning are suggested, and experimental
data reveal the important role of water and the pore size of
the catalyst in the formation of byproducts.
Results and Discussion
Initial screening of Nafion materials
Nafion polymers are commercially available in several forms,
including beads, membranes, and solutions (in water/alcohol
+
ꢀ1
mixtures). Nafion 117 (polymeric membrane, 0.91 meqH g )
+
and SAC-13 (Nafion polymer on amorphous silica, 0.12 meqH
ꢀ
1
g ) were selected because these physical forms can easily be
recovered and reused for subsequent reactions. The two
Nafion catalysts were used in the dehydration of 0.67 m xylose
in DMSO. This solvent was specifically chosen as the test sol-
vent to compare Nafion with other reported solid acid catalysts
used in furfural synthesis. DMSO, being hygroscopic, is effec-
tive in suppressing condensation side products and the rehy-
dration of sugars, resulting in improved selectivities and dehy-
Figure 1. a) Reaction profiles for the dehydration of xylose with Nafion 117
(
5 wt% loading vs. xylose) at different temperatures in DMSO. b) Reaction
profiles for the dehydration of xylose with SAC-13 (5 wt% loading vs. xylose)
at different temperatures in DMSO.
[
12]
dration rates. Without the catalyst, only a 1% furfural yield
was obtained. Table 1 shows the effects of reaction tempera-
ture on xylose dehydration.
As expected, the rate of xylose conversion increases with in-
creasing temperature. An induction period was observed for
the conversion of xylose in Nafion SAC-13 at 1258C. Although
the effect was less pronounced with Nafion 117 at 1258C,
more than 3 h was required for either catalyst to convert 50%
of the starting xylose. When the
Figure 1 shows representative reaction profiles for
Nafion 117 and SAC-13 at 5 wt% loading (vs. xylose content).
reaction temperature was in-
[
a]
Table 1. Dehydration of xylose with Nafion catalysts in DMSO.
creased from 1258C to 1508C,
the turnover frequency (TOF) for
[
a]
Entry Catalyst Catalyst loading
T
t
Conversion Selectivity Yield TOF
catꢀ
1
ꢀ1
ꢀ1 ꢀ1
xylose consumption also in-
creased (Table 1). Above 1508C,
xylose conversion was far more
rapid, where 90% of the initial
xylose had already been convert-
ed to furfural and/or byproducts
after 3 h. This effect was mir-
rored in the formation of furfural
for both catalysts at 5 wt% load-
ing versus xylose. Furfural pro-
duction was relatively slow at
[
wt%](vs. xylose) [8C] [h] [%]
[%]
[%] [mmolg
h
]
[mmolmeq h ]
1
2
3
4
5
6
7
8
9
0
117
117
117
117
117
117
SAC-13
SAC-13
SAC-13
SAC-13 38
SAC-13 38
SAC-13 38
5
5
5
20
20
20
5
125
150
175 0.5 89
125
150
175 0.5 91
125
150
175 0.5 84
125
150
175 0.5 90
8
2
87
84
67
61
58
62
66
65
63
50
45
57
57
56
58
51
52
56
60
59
55
39
37
52
49
51
31
131
212
16
49
59
34
143
233
18
53
65
111
1155
1954
25
172
263
8
2
91
91
8
2
87
78
13
5
5
139
234
3
21
32
1
8
2
92
93
1
1
12
1
1
258C compared to 150 and
758C, with maximum yields of
[a] Initial xylose concentration of 9.1 wt% [b] TOF for 1258C calculated at 1 h, other temperatures at 0.5 h.
5
36
www.chemsuschem.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 2011, 4, 535 – 541

Furfural from Xylose by Heterogeneous and Reusable Nafion Catalysts
5
8% and 55% at 8 h for Nafion 117 and SAC-13, respectively
mer with the silanol groups, which would decrease the availa-
[14]
(Table 1, entries 1 and 7). Peak furfural yields were achieved
bility of the sulfonic acid groups to participate in catalysis.
within 3 h at 150 and 1758C, but began decreasing with pro-
longed reaction time.
Notably, furfural, with a boiling point of 1628C, could be sepa-
rated from DMSO (boiling point 1898C) by distillation, allowing
the reuse of this solvent.
Figure 2 shows the reaction profiles for Nafion 117 and SAC-
1
3 at 20 and 38 wt% loading versus xylose. The xylose conver-
The reaction profiles of Nafion 117 (5 wt% loading versus
xylose) and SAC-13 (38 wt% loading versus xylose) at 1758C
sion rate was faster at these higher catalyst loadings. Similar to
+
were compared on the basis of equivalent number of H sites
ꢀ
3
+
(
4.55ꢁ10 meqH , determined by ion-exchange-capacity ti-
trations) in Figure 3.
Figure 3. Reaction profiles for the dehydration of xylose with Nafion 117
(
5 wt% catalyst loading versus xylose) versus Nafion SAC-13 (38 wt% cata-
lyst loading versus xylose) at 1758C.
Similar yields were achieved for Nafion 117 and SAC-13
when equivalent protons were compared. For Nafion SAC-13,
the TOF for xylose consumption expressed on a weight basis
was lower than that observed for Nafion 117 (Table 1, en-
tries 1–3,10–12). However, if the TOF was expressed as a rela-
tionship of the catalyst acid content, the activities of
Nafion 117 and SAC-13 were very similar. In all cases, the TOF
increased with increasing temperature. Nafion 117 for the de-
hydration of xylose to furfural was compared with other re-
ported heterogeneous catalysts (Table 2). Although the furfural
yield of Nafion 117 (20 wt% loading vs. xylose) was similar to
that of Amberlyst-15 or the heteropolyacid H PW O , the re-
Figure 2. a) Reaction profiles for the dehydration of xylose with Nafion 117
(20 wt% loading vs. xylose) at different temperatures in DMSO. b) Reaction
3
12 40
profiles for the dehydration of xylose with Nafion SAC-13 (38 wt% loading
vs. xylose) at different temperatures in DMSO.
action was performed at higher initial xylose concentrations, at
a higher xylose/catalyst ratio, and with shorter reaction times.
Consequently, the remaining work in this study was conducted
with Nafion 117 because it achieved the highest yields.
reactions conducted at 5 wt% loading, a decrease in furfural
yield was observed at 150 and 1758C for the higher catalyst
loading with extended reaction times. Nafion 117 (20 wt%
loading vs. xylose) produced the highest furfural yield at 60%,
with a selectivity of 66% in 2 h at 1508C (Table 1, entry 5), al-
though a yield of 59% was also achieved at 1758C after only
Reactivity of the Nafion 117 catalyst
Acid-catalyzed dehydration of xylose is postulated to involve
successive protonation of the xylose hydroxyl groups, leading
to the release of three water molecules and the formation of
0.5 h with 20 wt% catalyst loading versus xylose (Table 1,
[15]
entry 6). The higher yield of Nafion 117 with increasing catalyst
loading can be attributed to the greater number of available
furfural. The cluster-channel model for Nafion shows hydro-
[16]
phillic clusters of sulfonic acid groups (4 nm in diameter)
that are dispersed through the hydrophobic semicrystalline
[
13,14]
sulfonic acid sites
per unit weight relative to SAC-13, the
[9,17]
weight of which mainly comprises the amorphous silica sup-
port. Also, the reduced yield of SAC-13 compared to
Nafion 117 might be due to the interaction of the Nafion poly-
perfluorocarbon matrix.
Apparently, xylose migrates to
these hydrophilic clusters where sulfonic acid groups catalyze
the xylose dehydration.
ChemSusChem 2011, 4, 535 – 541
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemsuschem.org
537

J. H. T. Luong et al.
Table 2. Comparison of heterogeneous catalysts used in the dehydration of xylose to furfural.
Catalyst
Solvent
T
t
[h]
Conversion
[%]
Selectivity
[%]
Yield
[%]
Initial xylose
conc. [wt%]
Xylose/catalyst
[w/w]
[
8C]
[
a]
HY faujasite
Water/toluene (1:3)
Water/toluene (3:7)
DMSO
DMSO
DMSO
175
150
150
175
140
150
0.83
4
4
4
4
66
90
94
91
87
91
64
53
67
82
68
66
42
48
63
75
59
60
3.3
2.7
2.7
2.7
2.7
9.1
3.75
1.5
1.5
1.5
1.5
5
[
b]
Delaminated H-Nu6(2)
[
c]
H
3
PW12
O
40
[
d]
MCM-41-SO
3
Hc
[
d]
Amberlyst-15
Nafion 117
DMSO
2
[a] Reference [4]. [b] Reference [5]. [c] Reference [7]. [d] Reference [8].
The pore size of the Nafion catalysts may play a significant
role in the formation of byproducts. The size of a furfural mole-
[
18]
cule along its longest axis is 0.57 nm. The relatively large
pore sizes of Nafion 117 (>10 nm) could allow for furfural to
have a substantial residence time within the catalyst, leading
to condensation reactions with xylose intermediates to form
[
19]
furfural-based dimers, trimers, and oligomers. As such, mor-
denites with 0.7 nm pore size tend to exhibit higher selectivity
[
4]
than faujasites with 1.3 nm pore sizes. For heterogeneous
[
6]
zeolite ZSM-5 (1.2 nm pore size) and faujasites, the deposi-
tion of furfural oligomers inside catalyst pores is the primary
source of catalyst deactivation. However, even with the deposi-
tion of carbonaceous materials on the surface of the Nafion,
the reactivity and reusability of the catalyst remained high
after repeated runs (vide infra).
Figure 4. Arrhenius plots of xylose conversion using the Nafion 117 catalyst
at 5 wt% and 20 wt% catalyst loading versus xylose.
A maximum furfural yield was achieved in 2 h at 1508C with
91% xylose conversion for Nafion 117 (20 wt% loading vs.
xylose). However, as the reaction was prolonged, oligomeric
side products began to form, decreasing the yield from 60%
to 45–50% after 6 h of reaction. The reaction of furfural in
DMSO (0.67 m) with Nafion 117 (20 wt% loading vs. xylose,
[
15,22]
through coordination to the polyanionic Nafion network.
Solid catalysts can be subjected to diffusion and mass transfer
[23]
effects.
However, the linear Arrhenius plots demonstrated
1508C) was conducted to determine whether the formation of
that between 125 to 1758C, the conversion of xylose by Nafion
by-products was due to the self-polymerization of furfural.
After 6 h of reaction, >99% of the starting furfural was recov-
ered, confirming that the presence of the starting xylose was
was not affected by mass-transfer-related effects and that the
[23]
reaction was kinetically controlled. The optimal reaction tem-
perature range for using Nafion catalysts for xylose dehydra-
tion is unlikely to be outside of these temperatures. At reaction
temperatures below 1258C, the rate of xylose dehydration was
slow, and at temperatures above 1758C, particularly>1908C,
the degradation of the Nafion polymer would occur, hamper-
ing its reusability.
[15]
required for the byproduct formation process to occur.
In xylose dehydration, the conversion rate may be deter-
mined by the rate of transport of the xylose involved in the
diffusion of the bulk fluid to the active site of the Nafion sul-
fonic acid sites, located either on the surface or within the
pores of the catalyst. The rate of xylose conversion is inde-
Batch processes using mineral acids for the production of
furfural are often inefficient because furfural can react with
xylose intermediates in the liquid phase, forming unwanted
byproducts. In addition to using biphasic solvent systems,
[
15]
pendent of furfural loss reactions. For Nafion catalysts, in-
creasing temperatures produce an exponential rise in the initial
[
7]
rate of xylose conversion, similar to heteropolyacids.
[3]
The kinetics for xylose conversion was carried out using the
most notably water and tetrahydrofuran, processes such as
[15]
method of initial rates. The ln k vs. 1/T Arrhenius plots
Suprayield aim to produce high yields of furfural by rapidly
removing furfural vapor away from xylose as it is being formed
by adiabatic distillation. Presently, only orthophosphoric acid is
used as the catalyst in the Suprayield process due to corrosion
problems associated with other mineral acids. Solid acids could
be used as an alternative to orthophosphoric acid for such a
process, providing the benefits of a reusable, nontoxic catalyst.
However, the ability of solid acids to interact with slurries of
biomass material to produce furfural is currently unknown.
(
Figure 4) produced activation energy values of 86.4 and
ꢀ
1
8
9.3 kJmol , respectively, for Nafion 117 at 5% and 20% cata-
lyst weight loading. These energy values are lower than other
ꢀ
1 [20]
liquid acid catalysts such H SO (134 kJmol )
and HCl
2
4
ꢀ1
[21]
(
(
142 kJmol ), and the heterogeneous catalyst zeolite ZSM-5
ꢀ1 [6]
134 kJmol ) used in xylose conversion. Nafion 117 could ac-
celerate the reaction due to its high acidity as a superacid and
its ability to stabilize the protonated xylose transition state
5
38
www.chemsuschem.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 2011, 4, 535 – 541

Furfural from Xylose by Heterogeneous and Reusable Nafion Catalysts
Effect of water content on Nafion 117
[a]
Table 3. The effects of different catalyst poisons on Nafion 117
Water can lead to undesirable condensation side products and
Poison
Conversion [%]
Selectivity [%]
Yield [%]
[
6,7,12]
the rehydration of furfural.
Indeed, water played an impor-
None
NaCl
92
86
45
63
15
3
58
13
1
tant role in xylose dehydration by Nafion 117, as shown in
Figure 5. Nafion 117 could tolerate a water content of up to
NEt
3
[
a] Dehydration of 0.25 m xylose in 9:1 v/v DMSO/H
2
O, 20 wt% catalyst
loading vs. xylose, at 1508C in 2 h. Poison=0.07 mmol, xylose=
+
0
.25 mmol, available H =0.007 mmol
fural yield, 13% and 1% respectively, compared to 58% for the
control. Poisons can play a role in reducing the effectiveness of
the catalyst, particularly in real-world industrial applications
where the xylose source material may be alkaline or contain
significant quantities of cations.
Catalyst recyclability of Nafion 117
A series of experiments (20 wt% loading versus xylose, 2h,
1508C) was conducted to assess the reusability of Nafion 117.
Figure 5. Effects of water content on dehydration of xylose with Nafion 117
The catalyst was removed from the reaction mixture after each
run by filtration and reused without any treatment. After each
run, the Nafion catalyst slowly changed in color, from colorless
to brown, with minute amounts of organic residue coated on
the catalyst surface.
(
20 wt% loading versus xylose, 1508C, 2 h) in DMSO.
1
0% in DMSO without significant loss of activity, as an average
furfural yield of 56% was obtained with 89% xylose conver-
sion. However, the furfural yield decreased significantly to 36%
and 29% when the water content increased to 15 and 20%.
The dehydration of xylose to furfural releases three equivalents
of water for every equivalent of xylose, that is, the maximum
water content added to the reaction mixture would be around
Scanning electron microscopy—energy dispersive X-ray
(
SEM–EDX) analysis was conducted on the pristine catalyst and
after three reaction runs. A cross-sectional image of the pris-
tine catalyst (Figure 6a) shows striations along the edge of the
membrane. After three runs, the roughness along the edge of
the membrane surface has been significantly reduced, possibly
due to the deposition of organic residues (Figure 6b ). Minimal
membrane swelling was observed after each run. EDX analysis
3
.5%. The experiments conducted in this study with varying
water content intend to show that water tolerance granted by
the catalyst is acceptable within the range of water produced
by the dehydration reaction. Notice also that many other solid
(
Figure 6c) provided further information about the elemental
[22]
composition of the Nafion material. A slight increase in carbon
acid catalysts are also affected by the presence of water. In
the case of Nafion, water may also lead to the solvation of sul-
fonic acid groups which reduces the number of participating
acidic sites, thereby reducing the rate at which xylose dehydra-
[
24]
tion occurs.
Catalyst poisoning of Nafion 117
Deprotonation of the sulfonic acid groups of Nafion 117 would
deactivate the catalyst by reducing the number of available
acid sites for xylose dehydration. Monovalent and divalent cat-
+
ions can readily exchange with the H of the sulfonic acid
[
25]
groups of Nafion. Sulfonic acid groups can also react with
bases, such as triethylamine (NEt ) to form triethylammonium
3
[
26]
sulfonate salts, which may also result in a similar catalyst-de-
activating effect.
To confirm the effects of sulfonic acid exchange and neutral-
ization on the Nafion catalyst, NaCl and NEt were added to
3
2
.5 m aqueous solutions of xylose (Table 3). These solutions
Figure 6. SEM images of Nafion 117 a) before, and b) after reaction. c) EDX
spectrum of Nafion 117 membrane after reaction. d) IR spectra of pristine
Nafion 117 and post-reaction Nafion 117 (after three recycling runs).
were diluted in DMSO to yield 10% water and 3.8% initial
xylose content. Both NaCl and NEt adversely affected the fur-
3
ChemSusChem 2011, 4, 535 – 541
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemsuschem.org
539

J. H. T. Luong et al.
and oxygen content, along with a decrease in sulfur and fluo-
rine content after the reaction runs, showed that non-Nafion
carbonaceous matters were deposited on the membrane’s sur-
face (Table 4).
Table 4. Elemental analyses of pristine and post-reaction Nafion 117.
Element
Pristine Nafion 117 [wt%]
Post-reaction Nafion 117 [wt%]
C
O
S
22.7
5.4
1.0
23.9
5.6
0.9
F
70.8
70.6
Figure 8. Furfural yield from recycling of Nafion 117 (20 wt% loading versus
xylose, 1508C, 2 h).
For the attenuated total reflectance–infrared (ATR–IR) spec-
tra of the membrane samples (Figure 6d), characteristic peaks
for Nafion are listed in Table 5. The lack of peak shifts between
even in the presence of organic impurities that could block
the catalytically active acidic sites, leading to catalyst deactiva-
tion. This is in contrast to other solid acid catalysts such as sul-
fonic-acid-functionalized MCM or delaminated zeolites, which
employed catalyst regeneration procedures and still experi-
enced reduced activity with consecutive runs. Although
Nafion 117 activity remains high after 15 runs, treatment with
dilute hydrogen peroxide and acid could regenerate the cata-
Table 5. ATR–IR absorption peaks of pristine and post-reaction
Nafion 117
ꢀ1
ꢀ1 [a]
Peak
Pristine Nafion 117 [cm
]
Post-Reaction Nafion 117 [cm ]
n
n
n
n
n
as(CF
2
)
1201
1146
1057
982
1203
1146
1052
981
s
s
s
s
(CF
2
)
(SO)
[28]
[29]
(CFRCF
(COC)
3
)
lyst back to a tan-colored membrane, if desired. Annealing
969
968
is another means to improve catalyst longevity by altering the
physical properties of Nafion, such proton conductivity, water
permeability and water sorption, a subject of future endeavors.
[a] After 3 runs.
the pristine and used membrane show that there are no signif-
icant changes to the chemical environment of the used Conclusions
[
27]
Nafion, even after thermal treatment at 1508C in DMSO over
three consecutive runs. The ATR–IR spectrum of the used cata-
lyst was also clear of any bands that would suggest furfural
was coated on the surface of the catalyst. Atomic force micros-
copy (AFM) images of the fresh and used catalyst are shown in
Figure 7. In Figure 7a, the surface of the Nafion is smooth, but
in Figure 7b, the surface of the catalyst is covered with local-
ized regions of nanoparticle-sized organic residue deposits.
After 15 consecutive runs at the optimized reaction condi-
tions, the Nafion membrane remained intact with furfural
yields ranging from 58 to 62% (Figure 8). Xylose conversion
was >90% in all 15 runs after 2 h. Such a result clearly demon-
strates the stability and reusability of the Nafion 117 catalyst
Nafion 117 is a robust and reusable heterogeneous catalyst for
the dehydration of concentrated xylose to yield 60% furfural in
2
h at 1508C. A negligible mass transfer effect is noted at 125
to 1758C, that is, the reaction is virtually kinetic-controlled. The
catalyst is reused multiple times without loss of activity, a sig-
nificant achievement over existing solid acid catalysts. Consid-
ering its chemical and thermal stability under the reaction con-
ditions for xylose dehydration, this reusable catalyst could be a
potential catalytic system for the industrial production of furfu-
ral, an important chemical with diversified applications, from
renewable xylose-containing biomass. From a practical view-
point, it is desirable to evaluate the applicability of Nafion for
the synthesis of furfural from different sources of hemicellulose
including xylan; a subject of future endeavor.
Experimental Section
Materials: d-(+)-Xylose (BDH), Nafion SAC-13, furfural, dimethyl
sulfoxide (DMSO), phloroglucinol (Sigma–Aldrich), acetonitrile (J. T.
Baker), hydrochloric acid (Fisher), and glacial acetic acid (EMD)
were used as received. Nafion 117 (Sigma–Aldrich) was purified by
treatment with hot hydrogen peroxide (3%) to remove organic im-
purities and 1m hot sulfuric acid to remove metallic/ionic impuri-
[28]
ties. Catalytic acid sites of the Nafion species were determined
[30]
Figure 7. AFM images of Nafion 117 a) before and b) after reaction.
by titration.
5
40
www.chemsuschem.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 2011, 4, 535 – 541

Furfural from Xylose by Heterogeneous and Reusable Nafion Catalysts
Analytical methods: Furfural concentration was determined by
high-performance liquid chromatography (HPLC) at 258C using a
spectrophotometer (Waters Lambda-Max Model 487) set at
Keywords: biomass · heterogeneous catalysis · Nafion ·
renewable resources · solid acids
2
50 nm, equipped with a pump (Waters Model 590), and a Beck-
man Ultrasphere (235330) 5 mm column. An acetonitrile/water
[
1] D. Fengel, G. Wegener, Wood: Chemistry, Ultrastructure, Reactions, Walter
de Gruyter, Berlin, 1984, pp. 107–131.
ꢀ1
(
10:90 v/v) mixture was used as the mobile phase at 2.0 mLmin .
The furfural concentration was estimated from a calibration curve.
The unreacted xylose content was determined by a colorimetric
[2] A. Corma, S. Iborra, A. Velty, Chem. Rev. 2007, 107, 2411–2502.
[
3] R. Xing, A. V. Subrahmanyam, H. Olcay, W. Qi, G. P. van Walsum, H.
Pendse, G. W. Huber, Green Chem. 2010, 12, 1933–1946.
[31]
assay using a Beckman DU 640 spectrophotometer.
xylose samples were reacted with phloroglucinol in acidic media at
008C for 4 min. Absorbances were measured at 554 nm with the
xylose concentration calculated from a calibration curve. Xylose
In brief,
[
4] C. Moreau, R. Durand, D. Peyron, J. Duhamet, P. Rivalier, Ind. Crops Prod.
1
998, 7, 95–99.
1
[
[
5] S. Lima, M. Pillinger, A. A. Valente, Catal. Commun. 2008, 9, 2144–2148.
6] R. O’Neill, M. N. Ahmad, L. Vanoye, F. Aiouache, Ind. Eng. Chem. Res.
[32]
conversion [Eq. (1)] and furfural selectivity [Eq. (2)] were calculat-
ed as:
2
009, 48, 4300–4306.
[
[
7] A. S. Dias, M. Pillinger, A. A. Valente, Appl. Catal. A 2005, 285, 126–131.
8] A. S. Dias, M. Pillinger, A. A. Valente, J. Catal. 2005, 229, 414–423.
Conversion ¼ ðinitial xylose ½molꢁ
ð1Þ
[9] K. A. Mauritz, R. B. Moore, Chem. Rev. 2004, 104, 4535–4585.
10] a) G. Gelbard, Ind. Eng. Chem. Res. 2005, 44, 8468–8498; b) P. Wipf, En-
cyclopedia of Reagents for Organic Synthesis, John Wiley & Sons, New
York, 2005, pp. 167.
[11] D. Mercadier, L. Rigal, A. Gaset, J.-P. Gorrichon, J. Chem. Technol. Biotech-
nol. 1981, 31, 489–496.
[
ꢀ
unreacted xylose ½molꢁÞ=initial xylose ½molꢁ
Selectivity ¼ furfural ½molꢁ=ðinitial xylose ½molꢁ
unreacted xylose ½molꢁÞ
ð2Þ
ꢀ
[
12] J. N. Chheda, Y. Romꢂn-Leshkov, J. A. Dumesic, Green Chem. 2007, 9,
42–350.
13] X. Qi, M. Watanabe, T. M. Aida, R. L. Smith Jr. , Green Chem. 2009, 11,
327–1331.
3
Procedure for the conversion of xylose to furfural: d-(+)-Xylose
100 mg), DMSO (1 mL), and an acid catalyst, were loaded into a
mL vial with a magnetic stir bar and sealed. The mixture was
[
(
5
1
[
[
14] P. Botella, A. Corma, J. M. Lopez-Nieto, J. Catal. 1999, 185, 371–377.
15] K. J. Zeitsch, The Chemistry and Technology of Furfural and Its By-Prod-
ucts, Elsevier, Amsterdam, 2000.
stirred at 300 rpm and heated in a constant temperature oil bath
after a certain reaction time. In general, three replicates were taken
for each data point. Zero time was taken to be the instant the vial
was immersed in the oil bath. After reaction, the vial was cooled to
room temperature and the sample was analyzed for furfural and
unreacted xylose by HPLC and UV/vis spectroscopy, respectively.
[16] a) M. Jeske, C. Ellenberg, C. Soltmann, M. Willhelm, D. Koch, G. Grath-
wohl, J. New Mater. Electrochem. Systems. 2008, 11, 31–36; b) T. D.
Gierke, G. E. Munn, F. C. Wilson, J. Poly. Sci.: Poly Phys. Ed. 1981, 19,
1687–1704.
[
17] Q. Deng, Y. Hu, R. B. Moore, C. L. McCormick, K. A. Mauritz, Chem. Mater.
1997, 9, 36–44.
Characterization of Nafion 117: SEM–EDX analysis was performed
on a Hitachi S 2600N SEM (Hitachi Scientific Instruments, Tokyo,
Japan) equipped with a microanalysis detector for EDX (Inca x-act,
Oxford Analytical Instruments, Abington, UK). Nafion samples (pris-
tine and post-reaction) were glued on double-sided carbon tape
and mounted on a sample holder. EDX spectra were collected at
[18] E. Sjçman, M. Mꢃnttꢃri, M, Nystrçm, H. Koivikko, H. Heikkilꢃ, J. Membr.
Sci. 2007, 292, 106–115.
[
[
19] A. Gandini, M. N. Belgacem, Prog. Polym. Sci. 1997, 22, 1203–1379.
20] N. Bhandari, D. G. Macdonald, N. N. Bakhshi, Biotechnol. Bioeng. 1984,
2
6, 320–327.
[
[
[
21] J. M. Smuk, J. F. Harris, L. L. Zoch, J. Phys. Chem. 1966, 70, 71–77.
22] T. Okuhara, Chem. Rev. 2002, 102, 3641–3665.
23] a) C. N. Satterfield, Mass Transfer in Heterogenous Catalysis, Robert E.
Krieger Publishing Company, Malabar, Florida, 1981; b) R. N. Mukherjea,
P. Bhattacharya, T. Gangopadhyay, B. K. Ghosh, Biotechnol. Bioeng. 1980,
22, 543–553.
3
08 angle, 20 kV accelerating voltage and 20 nm working distance.
EDX results were analyzed using incorporated Inca, Point and Ana-
lyze software. SEM micrographs acquired in backscattered electron
mode were obtained at low vacuum of 15 Pa, 25 kV of accelerating
voltage with a 12 mm working distance.
ATR–FTIR spectra were collected (Bruker Tensor 27 FTIR spectro-
[24] Y. Liu, E. Lotero, J. G. Goodwin Jr. , J. Catal. 2006, 242, 278–286.
[25] a) H. L. Yeager, A. Steck, Anal. Chem. 1979, 51, 862–865; b) A. Steck,
H. L. Yeager, Anal. Chem. 1980, 52, 1215–1218.
ꢀ
1
ꢀ1
photometer) from 4000 to 600 cm for 64 scans and 4 cm reso-
lution using a zinc selenide (ZnSe) crystal. AFM (atomic force mi-
croscopy) measurement was performed using tapping mode at
scan rates of 0.5 Hz with 512ꢁ512 pixels on a Dimension 3100
with a NanoScope IV controller (Veeco Digital Instruments, Santa
Barbara, CA). All measurements were made using a silicon cantile-
[
26] a) L. Holub, K. Je rˇ ꢂbek, J. Mol. Catal. A: Chem. 2005, 231, 21–26; b) B. R.
Einsla, Y.-T. Hong, Y. S. Kim, F. Wang, N. Gundaz, J. E. McGrath, J. Polym.
Sci., Part A: Polym. Chem. 2004, 42, 862–874.
[
27] Z. Liang, W. Chen, J. Liu, S. Wang, Z. Zhou, W. Li, G. Sun, Q. Xin, J.
Membr. Sci. 2004, 233, 39–44.
ꢀ1
ver (MPP-11100, spring constant ca. 40 Nm , resonance frequency
ca. 300 kHz, NanoDevices, CA).
[28] C. Yang, S. Srinivasan, A. B. Bocarsly, S. Tulyani, J. B. Benziger, J. Membr.
Sci. 2004, 237, 145–161.
[
29] J. E. Hensley, J. D. Way, S. F. Dec, K. D. Abney, J. Membr. Sci. 2007, 298,
90–201.
[30] D. E. Lꢄpez, J. G. Goodwin Jr. , D. A. Bruce, J. Catal. 2007, 245, 381–391.
1
Acknowledgements
[
31] T. J. Eberts, R. H. B. Sample, M. R. Glick, G. H. Ellis, Clin. Chem. 1979, 25,
440–1443.
32] J. Haber, Pure Appl. Chem. 1991, 63, 1227–1246.
1
This work was conducted as part of the National Bioproducts
Program of Canada to promote the development of sustainable
energy and materials. We thank Pierre Bouvrette, Sabahudin Hra-
povic (SEM–EDX), and Yali Liu (AFM) for technical support.
[
Received: January 18, 2011
Published online on March 17, 2011
ChemSusChem 2011, 4, 535 – 541
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemsuschem.org
541