Conversion of hemicellulose to furfural and levulinic acid using biphasic reactors with alkylphenol solvents

Article abstract of DOI:10.1002/cssc.201100608

Plans with plants: The hemicellulose fraction of lignocellulosic biomass is converted to furfural and levulinic acid using biphasic reactors with alkylphenol solvents that selectively partition furanic compounds from acidic aqueous solutions. The furfural and levulinic acid products are valuable compounds for a variety of chemical applications, and they serve as precursors for the synthesis of liquid transportation fuels. Copyright

Full text of DOI:10.1002/cssc.201100608

DOI: 10.1002/cssc.201100608
Conversion of Hemicellulose to Furfural and Levulinic Acid using Biphasic
Reactors with Alkylphenol Solvents
Elif I. Gꢀrbꢀz, Stephanie G. Wettstein, and James A. Dumesic*[a]
Diminishing fossil fuel resources and the increasing impact of
global climate change have driven research towards the utiliza-
tion of lignocellulosic biomass resources as renewable feed-
stocks for the production of energy, fuels, and chemicals. The
conversion of lignocellulosic biomass into fuels and chemicals
requires effective utilization of the C5 and C6 sugars present in
hemicellulose and cellulose, respectively, by either processing
these fractions together or separating and processing them
separately. While simultaneous processing, such as in gasifica-
tion or pyrolysis, offers the potential for simplicity of operation,
the fractionation of hemicellulose and cellulose allows the
processing of each fraction to be tailored to take advantage of
the different chemical and physical properties of these frac-
tions, and provides increased flexibility of operation. For exam-
ple, chemical processing methods can be employed to convert
C5 sugars into fuels/chemicals in hemicellulose, while employ-
ing recent advances in biological conversions allows to convert
the C6 sugars in cellulose into fuels and/or chemicals.[1,2] One
can also take advantage of the physical properties of cellulose
for pulp and paper applications.
addition, while zeolite catalysts (i.e., ZSM-5) can be used to re-
place resin catalysts and can be regenerated with a calcination
treatment following deactivation,[13] employing these catalysts
results in significantly lower LA yields, especially when in-
creased LA concentrations are desired in the product stream.
Considering the aforementioned challenges for processing
hemicellulose, we present a new biorefining strategy for con-
verting the hemicellulose portion of lignocellulosic biomass to
FuAL and LA by utilizing biphasic systems that consist of an
extractive organic layer and an aqueous layer that contains a
mineral acid. These biphasic systems achieve high concentra-
tions of FuAL and LA, enabling the recovery of both products
at the top of distillation columns, and eliminating issues relat-
ed to deactivation and regeneration of solid acid catalysts.
Three organic solvents, 2-sec-butylphenol (SBP), 4-n-hexylphe-
nol (NHP) and 4-propyl guaiacol (PG), are demonstrated to be
effective extracting agents for the production of FuAL and LA
in these biphasic systems. Information on the toxicity and
availability of these alkylphenol solvents is given in the Sup-
porting Information. The use of these solvents is particularly
advantageous because they (i) have high partition coefficients
for extraction of FuAL, FuOH, and LA; (ii) do not extract signifi-
cant amounts of mineral acids from aqueous solutions;
(iii) have higher boiling points than the final product; and
(iv) could potentially be synthesized directly from biomass
(i.e., lignin), such that these solvents would not have to be
transported to the site of the biomass conversion steps.
Herein, we show that the hemicellulose fraction of lignocel-
lulosic biomass can be converted into furfural and levulinic
acid by using biphasic reactors with alkylphenol solvents that
selectively partition furanic compounds from acidic aqueous
solutions. These furfural and levulinic acid products are
valu !->able compounds for a variety of chemical applica-
tions,[3,4] and they serve as precursors for the synthesis of
liquid transportation fuels.[5–7]
For the first step of this biorefining strategy (Figure 1), solid
biomass (i.e., corn stover) is subjected to mild pretreatment in
a dilute-acid, aqueous solution to solubilize the hemicellulose
as xylose. After filtering the solution from the solid cellulose
and lignin, an organic solvent (i.e., SBP) is added to the aque-
ous solution, and these liquids are heated in a biphasic reactor
to achieve dehydration of xylose to FuAL, which is a valuable
chemical intermediate.[3] FuAL can be distilled from SBP and
sold as a chemical or, as depicted in Figure 1, converted to LA
by first hydrogenating FuAL to FuOH over a metal-based cata-
lyst (e.g., copper)[14,15] and then reacting the FuOH with water
in a biphasic reactor to form LA. Similar to FuAL, the LA prod-
uct can be distilled from the organic solvent and sold as a
chemical.
The conversion of cellulose to chemicals and liquid fuels has
been demonstrated through the formation of several platform
molecules, such as glucose, 5-hydroxymethylfurfural, and levu-
linic acid (LA), utilizing chemical routes;[6–9] however, fewer
studies address the conversion of hemicellulose into chemicals
and fuels.[10,11] Previous studies for the production of furfural
(FuAL) from C5 sugars (i.e., xylose) suffer from the low concen-
trations of FuAL in the product stream due to low xylose con-
centrations (1–2 wt%) obtained from hemicellulose decon-
struction.[10,11] In addition, even though the production of LA
from furfuryl alcohol (FuOH) has been reported with good
yields over ion-exchange resin catalysts (e.g., Amberlyst),[12,13]
the regeneration of these catalysts following deactivation by
deposition of solid humins during reaction is problematic. In
Xylose dehydration to FuAL has been demonstrated with
high yields (ca. 90%) in several previous studies using mineral
acids and salts in biphasic systems with organic solvents, such
as methyl isobutyl ketone (MIBK), 2-butanol, and tetrahydrofur-
an (THF).[9,11,16] However, the low partition coefficients for ex-
traction of FuAL in these systems (i.e., the ratio of the FuAL
concentration in the organic solvent to the FuAL concentration
in aqueous solution) required the use of large amounts of or-
ganic solvent relative to the aqueous xylose solution, resulting
[a] E. I. Gꢀrbꢀz, Dr. S. G. Wettstein, Prof. J. A. Dumesic
Chemical and Biological Engineering Department
University of Wisconsin
Madison, WI 53706 (USA)
Fax: (+1)608-262-5434
E-mail: dumesic@engr.wisc.edu
Supporting Information for this article is available on the WWW under
http://dx.doi.org/10.1002/cssc.201100608.
ChemSusChem 2012, 5, 383 – 387
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
383

tive for extracting FuAL from
acid aqueous solutions, with an
exceptionally high partition coef-
ficient of ca. 50 in the absence
of salt in the aqueous phase and
increasing to >90 when the
aqueous phase is saturated with
NaCl (these values were mea-
sured for a mass ratio of organic
to aqueous layer equal to 2).
This high partition coefficient of
FuAL in the water–SBP system
allows for high yields of FuAL in
the organic phase, even when
small amounts of the SBP sol-
vent are used in the biphasic re-
actor (i.e., when the mass ratio
of the aqueous solution to SBP
is increased). Therefore, higher
concentrations of FuAL can be
obtained in a single stage com-
pared to the starting xylose con-
centration. As shown in Table 1,
entry 1, a 1.5 wt% xylose feed in
an aqueous solution containing
0.1m HCl and saturated with
NaCl results in high yields of
FuAL (78%), with 90% of FuAL
partitioning to SBP, when a small
amount of SBP is employed
(xylose solution/SBP mass ratio
of 6.67). The final organic phase
Figure 1. Schematic representation of the conversion of hemicellulose to furfural (FuAL) through the biphasic de-
hydration of xylose to furfural using an organic solvent, followed by the production of levulinic acid (LA) by reduc-
tion of furfural to furfuryl alcohol (FuOH) over a metal catalyst and further reaction of furfuryl alcohol with water
in a biphasic reactor.
in low concentrations of FuAL in the organic phase compared
to the starting concentration of xylose in the aqueous phase.
This problem is exacerbated by the low concentrations of
xylose (e.g., 1–2 wt%) typically obtained in aqueous solutions
from dilute acid treatment of real biomass feedstocks (e.g.,
corn stover). Thus, it is important to identify new methods to
produce higher concentrations of FuAL in the organic phase of
biphasic reactor systems to improve the efficiency of down-
stream processing options, such as distillation or further up-
grading reactions.
contains 4.1 wt% FuAL in SBP. Another advantage of using SBP
as the organic solvent is that it extracts only negligible
amounts of ClÀ ions (13 ppm, detected via ion chromatogra-
phy in Galbraith Laboratories) residing in the aqueous phase
originating both from HCl and NaCl. In addition, no chlorinated
organic compounds were detected within our detection limits.
To produce an aqueous solution of xylose from the hemicel-
lulose fraction of a real biomass feedstock, corn stover was
treated for 5 h at 363 K in an aqueous solution containing
0.1m HCl and saturated with NaCl (see Supporting Informa-
tion). This corn stover feed (1.1 wt% xylose) resulted in a maxi-
mum yield of ca. 70% for production of FuAL (Table 1, entry 2),
with again ca. 90% of FuAL partitioning into SBP, using our bi-
We have recently identified a new extracting solvent, SBP,
for the effective extraction of LA from acidic aqueous solu-
tions.[17] We now show for the first time that SBP is also effec-
Table 1. Results of xylose dehydration experiments carried out at 443 K in a biphasic reactor system (10 mL glass reactors), using SBP as the extracting sol-
vent in a 6.67:1 aqueous/organic mass ratio, with aqueous solutions containing 0.1m or 0.25m HCl and saturated with NaCl. For all experiments, approxi-
mately 90% of total furfural is partitioned into the SBP phase.
Entry
Xylose
[wt%]
HCl conc.
[M]
t
Xylose conv.
[%](Æ2)
Furfural select.
[%](Æ2)
Furfural yield
[%](Æ2)
Final furfural
in SBP [wt%](Æ0.5)
[min]
1
2
1.5
0.1
0.1
0.25
0.1
0.25
20
30
15
15
15
98
95
92
92
95
80
74
82
77
75
78
70
75
71
71
4.1
3.2
3.2
11.6
5.3
1.1 (from corn stover)
3
4
5
2.1 (2 cycles from corn stover)
384
www.chemsuschem.org
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 2012, 5, 383 – 387

phasic reactor system. The small
decrease in yield compared to
the simulated feed could be due
to inhibiting effects of byprod-
ucts produced in the hemicellu-
lose deconstruction step; how-
ever, the overall yield from this
biomass derived feed can be in-
creased to ca. 75% by increasing
the HCl concentration to 0.25m
(entry 2). It should be noted that
increasing the HCl concentration
for the corn stover deconstruc-
Table 2. Yields to levulinic acid starting from furfuryl alcohol solutions in 2-sec-butylphenol (SBP) in a biphasic
reactor system with water or in a monophasic system with water at 398 K. All reactions were carried out in a
batch reactor system. LA and SA correspond to levulinic acid and 1m sulfuric acid, respectively. Furfuryl alcohol
conversion is complete for all experiments.
Entry
Furfuryl
alcohol [wt%]
Organic
solvent
Org./aq.
[mLmLÀ1
Catalyst
t
[h]
Yield to
LA [%]
LA in
org. [%]
]
1
2
3
4
5
6
1
1
1
1
1
–
–
SBP
SBP
SBP
SBP
–
–
1
1
2
1
ZSM-5
SA
SA
SA
SA
1
1
1
3
1
1
15Æ1
32Æ1
68Æ4
72Æ4
66Æ4
31Æ2
–
–
67Æ5
67Æ5
81Æ5
66Æ2
10
SA
tion step results in the same xylose yields by decreasing the re-
action time to 2 h. It is shown in entry 3 with a simulated
5 wt% xylose feed that the overall furfural yield (ca. 71%) and
the partitioning into SBP are not altered when the xylose con-
centration is increased. Thus, the xylose concentration in the
aqueous phase can be increased by adding corn stover to the
HCl solution in progressive stages. With two cycles of corn
stover deconstruction, the concentration of xylose is doubled
(2.1 wt%), reaching similar overall yields. Using this feed for bi-
phasic dehydration (entry 4), a maximum FuAL yield of ca.
71% is obtained with ca. 90% of FuAL partitioning into SBP, re-
sulting in 5.3 wt% FuAL in SBP. Finally, scaling up the biphasic
reaction from a 10 mL glass reactor to a 450 mL Parr reactor
leads to a similar overall yield to FuAL (66%).
The strategy employed in this paper is to utilize biphasic re-
actors to achieve a low concentration of reactive species in
acidic aqueous solutions to increase LA yields by minimizing
undesirable polymerization reactions, and yet to achieve a
high concentration of the final product to facilitate product
separation and purification. To apply this concept, a biphasic
reactor with SBP was employed to form LA, as seen in
Figure 1, in a manner analogous to the case of xylose dehydra-
tion to FuAL. In the case of FuOH conversion to LA, the organ-
ic extracting solvent partitions the reactant out of the reactive
aqueous phase (thereby maintaining a low concentration in
the aqueous solution), whereas in the case of xylose dehydra-
tion to FuAL, the product is partitioned out of the reactive
aqueous phase.
As shown in Figure 1, the FuAL can now be distilled out of
SBP (see the Supporting Information for distillation calcula-
tions) and sold as a chemical, or it can be hydrogenated with
nearly quantitative yields over a metal catalyst (e.g., copper) in
the vapor phase to form FuOH, an important chemical in the
polymer industry.[14,15] Importantly, as explored below, this
FuOH intermediate can be used in another biphasic reactor
system to produce LA, another attractive platform molecule
from which fine chemicals (e.g., d-aminolevulinic acid, diphe-
nolic acid) and fuel additives (e.g., levulinate esters, MTHF) can
be produced.[18]
Due to a high partition coefficient (7.5) of FuOH in the SBP–
water system, the FuOH reactant remains mostly in the organic
phase, decreasing the FuOH concentration in the acidic aque-
ous medium and, thus, decreasing the rates of degradation re-
actions accordingly. When the reaction is carried out in a bi-
phasic reactor, mineral acids can be used instead of solid acid
catalysts, and these mineral acids can be recovered and recy-
cled,[17,19] eliminating any issues of deactivation and regenera-
tion of possible solid acid catalysts. It should be noted that an-
other important function of the extracting solvent for produc-
tion of LA is to extract the majority of LA to the organic layer
to enable its separation from the mineral acid in the aqueous
layer. It can be seen in Table 2 that up to ca. 72% yield of LA
can be obtained using a biphasic system containing aqueous
1m H2SO4 solution and SBP (entry 3 and 4), while only 32%
yield of LA is obtained in a single aqueous phase medium with
1m H2SO4 and 1 wt% FuOH feed (Table 2, entry 2). Approxi-
mately 67% of LA can be retained in the SBP layer when the
volume ratio of organic to aqueous layer is 1 (entry 3). Howev-
er, by decreasing the amount of aqueous layer to obtain a
ratio of 2 (entry 5), a higher amount of LA (ca. 81%) can be re-
covered in the organic layer, while still reaching approximately
66% total yield, compared to 68%, shown in entry 3. (The LA
that remains in the aqueous solution after phase separation of
the solvents will be essentially inert when the aqueous phase
is used in subsequent biphasic processing of furfuryl alcohol,
and thus this LA can be recovered in the process.)
The conversion of FuOH in alcohol solvents to produce levu-
linate esters has been reported in the literature with high
yields;[13] however, when LA is the desired product, the conver-
sion of FuOH to LA in aqueous acidic solutions is plagued by
polymerization reactions that lead to low selectivities at even
modest concentrations (e.g., 1 wt%) when compared to pro-
duction of levulinate esters. For example, we have achieved an
ethyl levulinate yield of 85% starting with 1 wt% FuOH in eth-
anol using Amberlyst-15 as the acid catalyst, as opposed to a
lower yield of 55% for production of LA in water at 398 K with
equal amount of catalyst and reaction time (see the Support-
ing Information). The same trend is also seen for ZSM-5. We
achieved 64% yield to ethyl levulinate in alcohol compared to
15% yield to LA (Table 2, entry 1) in water at the same reaction
conditions. This behavior is observed because the rates of un-
desirable reactions in acidic aqueous solutions increase more
rapidly with reactant concentration than the rates of desirable
reactions, such that the selectivity for desirable reactions be-
comes lower at higher reactant concentrations.
While use of the SBP solvent leads to a significant increase
in the yield of LA compared to monophasic reaction in water,
it can be seen in Table 2, entry 6 that increasing the concentra-
ChemSusChem 2012, 5, 383 – 387
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemsuschem.org
385

tion of FuOH from 1 to 10 wt% in SBP leads to a decrease in
the LA yield (ca. 31%). An additional strategy to ensure that
the concentrations of the reactant and the intermediates
always remain low versus time is to employ semi-batch opera-
tion, similar to Lange et al.[13] (see Supporting Information), in
which the FuOH solution in SBP is slowly fed to the reactor
that includes a heated solution of 1m H2SO4 solution. This
semi-batch operation resulted in LA yields of around 65%
starting with 10–20 wt% FuOH solutions (Table 3, entries 3–4),
and more than 70% of the total LA is partitioned into the or-
ganic layer. In contrast to these high yields obtained using SBP
and semi-batch operation, the yields of LA are still low
(Table 3, entries 1 and 2) in a semi-batch mode with a single
aqueous phase using ZSM-5, these yields being ca. 45% and
ca. 35% starting with feeds containing 10 and 20 wt% FuOH
concentrations, respectively. Thus, the high partition coefficient
of FuOH in the SBP-water system allows for high yields of LA
even at high feed concentrations that cannot be obtained in
monophasic water systems. The solution of LA in SBP can then
be processed further as presented in our earlier work to form
GVL over a Ru–Sn catalyst.[17]
higher than those obtained in monophasic aqueous phase sys-
tems with ZSM-5.
The biorefining strategy outlined here offers the production
of FuAL and LA from the hemicellulose portion of lignocellulo-
sic biomass utilizing biphasic systems with new solvent sys-
tems. In the case of xylose dehydration, the presence of an ex-
tractive organic solvent enables the continuous removal of the
highly reactive product (FuAL) from the acidic aqueous
medium to prevent further degradation. In the case of FuOH
hydrolysis, the biphasic system avoids high concentrations of
the highly reactive reactant and/or intermediates in the acidic
aqueous medium to prevent oligomerization reactions of
FuOH that result in the formation of solid humins. In both of
these reaction systems, the organic solvent extracts the majori-
ty of the FuAl and LA products, enabling the separation of
these valuable products from the mineral acid in the aqueous
layer. Use of solvents such as SBP, NHP and PG allows for the
production of FuAL and LA at higher concentrations compared
to monophasic reactions in water, leading to more efficient
separation of these products at the top of distillation columns.
In addition, these three new solvents may possibly be pro-
duced from lignin, leading to new research directions for the
creation of sustainable biorefineries that utilize effectively the
hemicellulose, cellulose, and lignin fractions of lignocellulosic
biomass.
Since the boiling point of LA (519 K) is higher than that of
SBP (500 K), LA cannot be removed from the SBP solvent at
the top of a distillation column. For this reason, two possible
alkylphenol solvents with higher boiling points were studied.
One alkylphenol solvent is 4-n-hexyl phenol (NHP; boiling
point 560 K), which as seen in Table 3 (entry 5), resulted in
good yields of LA (ca. 70%) with a 10 wt% FuOH feed and 1:1 Experimental Section
volume ratio of the organic solvent to the acidic aqueous solu-
Dehydration of xylose and hydrolysis of FuOH were carried out in
tion, with a slightly lower partitioning of LA (60% being in
NHP).
biphasic systems that consisted of an aqueous mineral acid solu-
tion (HCl and H2SO4, respectively) and an organic extracting solvent
(SBP, NHP, and PG). The experiments with solutions of xylose and
FuOH (1 wt%) were carried out in 10 mL glass reactors kept at con-
stant temperature (443 K and 398 K, respectively) in a pre-heated
oil bath using magnetic stirring. Semi-batch experiments for higher
FuOH feed concentrations were carried out in a 50 mL Parr reactor
by slowly feeding FuOH solutions to the preheated aqueous H2SO4
solutions. Upon completion of the reactions, the two phases were
separated and analyzed to quantify FuAL, FuOH, LA, and xylose
using GC (Shimadzu GC 2060, equipped with a DB-5 column
(Restek) and an FID) and HPLC (Waters 2695 system with a Bio-Rad
Another possible alkylphenol solvent for biphasic reactor op-
eration is 4-propyl guaiacol (PG), which has a boiling point of
537 K. Importantly, this compound has been reported to be
isolated from lignin degradation.[20] It can be seen in entry 6
that good overall yields of LA (ca. 69%) can be obtained with
2:1 organic to aqueous volume ratio, with ca. 60% being re-
tained in PG. As shown in entries 7 and 8, higher concentra-
tions of FuOH result in lower overall yields, however, the yields
of LA in these biphasic reactor systems are still considerably
Table 3. Yields to levulinic acid starting from furfuryl alcohol solutions in 2-sec-butylphenol (SBP), 4-n-hexylphenol (NHP), or 4-propyl guaiacol (PG) in a bi-
phasic reactor system with water or in a monophasic system with water at 398 K. All reactions were carried out in a semi-batch reactor system. LA and SA
correspond to levulinic acid and 1m sulfuric acid, respectively. Furfuryl alcohol conversion is complete for all experiments.
Entry
Furfuryl
alcohol [wt%]
Organic
solvent
Org./aq.
[mLmLÀ1
Catalyst
WHSV
[hÀ1
Yield to
LA [%]
LA in
org. [%]
LA in
soln. [wt%]
]
]
1
2
3
4
5
6
7
8
10(20)[a]
20(40)[a]
10(20)[a]
20[b]
–
–
–
–
1
1
1
2
2
2
ZSM-5
ZSM-5
SA
SA
SA
SA
SA
SA
0.24
0.48
0.08
0.16
0.08
0.16
0.32
0.48
45Æ1
35Æ1
67Æ2
65Æ2
70Æ2
69Æ3
59Æ3
55Æ3
–
–
5.7Æ0.2
10.0Æ0.2
5.2Æ0.2
10.2Æ0.2
4.7Æ0.2
4.6Æ0.4
8.8Æ0.6
13.0Æ0.6
SBP
SBP
NHP
PG
PG
PG
72Æ1
67Æ1
60Æ2
58Æ2
61Æ2
60Æ2
10[b]
10[b]
20(40)[a]
30(60)[a]
[a] Experiments carried out by starting with solvent and catalyst and feeding a concentration of furfuryl alcohol (shown in parenthesis) to reach a final fur-
furyl alcohol concentration (value not in parenthesis). [b] Experiments carried out by starting with aqueous SA solutions and feeding the given concentra-
tion of furfuryl alcohol.
386
www.chemsuschem.org
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 2012, 5, 383 – 387

[5] G. W. Huber, J. N. Chheda, C. J. Barrett, J. A. Dumesic, Science 2005, 308,
1446.
[6] J.-P. Lange, R. Price, P. M. Ayoub, J. Louis, L. Petrus, L. Clarke, H. Gosse-
link, Angew. Chem. 2010, 122, 4581; Angew. Chem. Int. Ed. 2010, 49,
4479.
Aminex HPX-87H column and a UV detector) for organic and aque-
ous phase analyses, respectively. All the yields reported in
Tables 1–3 are based on GC and HPLC analyses and are not isolat-
ed yields. Details of the experimental set-up and procedures are
given in the Supporting Information.
[7] J. Q. Bond, D. M. Alonso, D. Wang, R. M. West, J. A. Dumesic, Science
2010, 327, 1110.
[8] M. Mascal, E. B. Nikitin, Angew. Chem. 2008, 120, 8042; Angew. Chem.
Int. Ed. 2008, 47, 7924.
Acknowledgements
[9] Y. Roman-Leshkov, J. N. Chheda, J. A. Dumesic, Science 2006, 312, 1933.
[10] J. N. Chheda, Y. Roman-Leshkov, J. A. Dumesic, Green Chem. 2007, 9,
342.
[11] R. Xing, A. V. Subrahmanyam, H. Olcay, W. Qi, W. G. P. van, H. Pendse,
G. W. Huber, Green Chem. 2010, 12, 1933.
[12] W. D. Van De Graaf, J.-P. Lange, WO2007023173A1, 2007.
[13] J.-P. Lange, W. D. van de Graaf, R. J. Haan, ChemSusChem 2009, 2, 437.
[14] B. M. Nagaraja, A. H. Padmasri, B. D. Raju, K. S. R. Rao, J. Mol. Catal. A:
Chem. 2007, 265, 90.
[15] R. S. Rao, R. T. K. Baker, M. A. Vannice, Catal. Lett. 1999, 60, 51.
[16] R. M. West, Z. Y. Liu, M. Peter, J. A. Dumesic, ChemSusChem 2008, 1, 417.
[17] D. M. Alonso, S. G. Wettstein, J. Q. Bond, T. W. Root, J. A. Dumesic, Chem-
SusChem 2011, 4, 1078.
This work was supported in part by the US Department of Energy
Office of Basic Energy Sciences, and by the DOE Great Lakes Bio-
energy Research Center (http://www.greatlakesbioenergy.org),
which is supported by the US Department of Energy, Office of Sci-
ence, Office of Biological and Environmental Research, through
Cooperative Agreement DE-FC02–07ER64494 between The Board
of Regents of the University of Wisconsin System and the US De-
partment of Energy. We thank Sercan Murat Sen for the help
with ASPEN simulations.
[18] J. J. Bozell, L. Moens, D. C. Elliott, Y. Wang, G. G. Neuenschwander, S. W.
Fitzpatrick, R. J. Bilski, J. L. Jarnefeld, Resour. Conserv. Recycl. 2000, 28,
227.
[19] E. I. Gꢀrbꢀz, D. M. Alonso, J. Q. Bond, J. A. Dumesic, ChemSusChem
2011, 4, 357.
Keywords: furfural · hemicelluloses · heterogeneous catalysis ·
levulinic acid · sustainable chemistry
[1] B. C. Saha, J. Ind. Microbiol. Biotechnol. 2003, 30, 279.
[2] J. Zaldivar, J. Nielsen, L. Olsson, Appl. Microbiol. Biotechnol. 2001, 56, 17.
[3] A. S. Mamman, J.-M. Lee, Y.-C. Kim, I. T. Hwang, N.-J. Park, Y. K. Hwang,
J.-S. Chang, J.-S. Hwang, Biofuels, Bioprod. Biorefin. 2008, 2, 438.
[4] H. Mehdi, V. Fabos, R. Tuba, A. Bodor, L. T. Mika, I. T. Horvath, Top. Catal.
2008, 48, 49.
[20] M. Phillips, Science 1931, 73, 568.
Received: September 29, 2011
Published online on January 24, 2012
ChemSusChem 2012, 5, 383 – 387
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemsuschem.org
387