Direct, high-yield conversion of cellulose into biofuel

Article abstract of DOI:10.1002/anie.200801594

(Chemical Equation Presented) Fueling up with furans: Cellulose can be converted into furanic biofuels in unprecedented yields using an inexpensive, simple process involving concurrent hydrolysis, dehydration, and chlorine substitution reactions coupled with continuous extraction into an organic phase (see scheme). Furanic ethers, such as those that can be derived from the products above, are known diesel additives.

Full text of DOI:10.1002/anie.200801594

Communications
DOI: 10.1002/anie.200801594
Biofuels
Direct, High-Yield Conversion of Cellulose into Biofuel**
Mark Mascal* and Edward B. Nikitin
These are days of great incentive in the field of bioenergy
research. The stakes, of course, are immense—economic
independence from politically unstable, petroleum-exporting
countries, the remediation of greenhouse gas levels in the
atmosphere and their potential effect on the climate, and
mitigation of the economic consequences of our imminent
arrival at the Peak Oil point, particularly in light of the pace of
industrialization of emerging economic superpowers in Asia.
It may be said that the final answer to the global energy issue
will lie most credibly in ultraclean technologies based on
hydrogen and solar energy.[1] However, few would deny a
more immediate future to carbon-based fuels, in viewof the
prevailing automotive infrastructure based on the internal
combustion engine, as well as the fact that the chemical
industry will always require feedstocks for the production of
organic materials and chemicals, regardless of what is being
used for energy.[2]
The challenge to a newcarbon-based fuel economy, as it
emerges, is twofold: First, the carbon source must ultimately
be atmospheric carbon dioxide, which is most practically
harvested by the photosynthetic production of cellulose,
hemicellulose, starch, and simple sugars, and second, these
saccharides must be efficiently converted into molecules
which are ambient temperature liquids of low volatility and
high energy content. To some extent, the above challenge is
currently being met by the production of ethanol from either
starch-derived glucose or cane sugar, but this has largely been
an issue of expediency, making use of mature technologies
(agriculture and brewery/distillery) that were established long
before energy became an issue, and the approach is now
considered by many to be transitional.[3]
economy of glucose fermentation. Glucose is utilized by
microorganisms according to the equation C6H12O6 !
2C2H5OH + 2CO2. Even assuming quantitative efficiency
both in the derivation of glucose from cellulose as well as the
fermentation process, one third of the available carbon is
expelled as carbon dioxide, 9.6 g of which is produced for
every 10 g of ethanol. Effective approaches to biomass
utilization which avoid fermentation altogether and exploit
all of the available carbon present would thus be extremely
valuable. Certainly, this point has not escaped the attention of
researchers at the forefront of biomass conversion chemis-
try.[4] One promising direction this research has taken is
towards “furanics”, that is, high-energy, furan-based organic
liquids. A high-profile contribution by Dumesic and co-
workers in this area showed that fructose could be efficiently
converted, via 5-hydroxymethyfurfural (HMF, 1), into a range
of substituted furan and tetrahydrofuran products.[5] If,
however, this approach is to find broader application, it
cannot rely on fructose as the source of HMF (1). Interest-
ingly, a concurrent publication by Zhang and co-workers
described the conversion of glucose into HMF (1) in record
yield,[6] and taken together these two studies point towards a
workable nonfermentive process for the conversion of
glucose into biofuel. However, closer inspection of the latter
of these papers shows that the expensive 1-ethyl-3-methyl-
imidazolium chloride ionic liquid is used as the solvent which,
along with chromium(II) catalyst, produces HMF in about
70% yield, determined not by isolation but HPLC analysis.
Along the same lines, Dumesic and co-workers have pub-
lished a study in which HMF (1) is derived from glucose with
53% selectivity at high conversion in 60% aqueous DMSO in
a biphasic reactor. While also promising, the separation of
DMSO from HMF (1) remains an issue.[7]
Herein, we report that not only glucose, but cellulose
itself, is converted into furanic products in isolated yields of
greater than 80% by conversion mainly into 5-(chloro
methyl)furfural (CMF, 2), a hydrophobic molecule which is
easily sequestered into organic solvents in a two-phase
reaction medium. The experimental setup we use to this
purpose is a standard apparatus for continuous extraction of
an aqueous solution with a solvent of greater density than
water. Thus, microcrystalline cellulose was added to a stirred
solution of lithium chloride (5 wt%) in concentrated hydro-
chloric acid to give a homogeneous mixture, which was
introduced into a reaction chamber containing 1,2-dichloro-
ethane. The solvent was heated to reflux and the aqueous
slurry was kept at 658C with continuous mechanical stirring
and extracted for 18 h. At this point, a further solution of LiCl
in concentrated hydrochloric acid was added to the reaction
chamber and extraction continued for another 12 h. The
combined organic extracts were distilled to recover the
solvent, and the residual liquid was chromatographed. The
Since cellulose is by far the major form of photosyntheti-
cally fixed carbon, it can be argued that it should be the
principal focus of any emerging carbon-fuel technology. The
difficulty, from the point of viewof ethanol production, is that
fermentable sugars are not easily liberated from this material.
The current model for cellulose utilization involves saccha-
rification with immobilized enzymes, but despite recent
advancements, this remains a slowand expensive process.
Our own interest in this area had less to do with the
problems of cellulose hydrolysis than the poor carbon
[*] Prof. M. Mascal, Dr. E. B. Nikitin
Department of Chemistry and Bioenergy Research Group
University of California, Davis
Davis, CA 95616 (USA)
Fax: (+1)530-752-8995
E-mail: mascal@chem.ucdavis.edu
[**] This work was supported by a grant from Chevron Technology
Ventures. We also thank Prof. Brian Higgins for helpful discussions.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.200801594.
7924
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2008, 47, 7924 –7926

Angewandte
Chemie
following products were isolated: 5-(chloromethyl)furfural (2,
71%), 2-(2-hydroxyacetyl)furan (3, 8%), 5-(hydroxymethyl)
furfural (1, 5%), and levulinic acid (4, 1%). Filtration of the
furfural (MF, 6, b.p. 1878C), thus avoiding the need for
ethanol in the derivation of a useable biofuel from 2
(Scheme 1).[12] Hydrogenation of CMF (2) using a Pd/C
catalyst is reported to give 2,5-dimethylfuran, b.p. 948C, in
high yield,[13] while complete reduction would give 2,5-
dimethyltetrahydrofuran, b.p. 928C. Both of these latter
molecules are key products in the dehydration/reduction of
HMF (5) reported by Dumesic and co-workers,[5] and are
reduced relative to 2 by three and five equivalents of H2,
respectively. The facile hydrogenation of such molecules
makes them credible media for “hydrogen storage”, consid-
ering that the difference of five moles of H2 between 2 and
2,5-dimethyl-tetrahydrofuran represent 10% of the mass of
the latter. While a chromatographic separation is important to
establish the actual chemical yields of the products in this
study, treatment of the crude extracts containing 1–4 with
ethanol or H2/catalyst and direct isolation 5 and 6 by
distillation is equally feasible.[14]
Since there have been general concerns about the
potential health hazards of oxygenated fuels, it is worth
briefly commenting here on the toxicity of 5 and 6. A material
safety data sheet (MSDS) for EMF (5) has not been
published, but 5 would most likely be hydrolyzed in vivo
into HMF (1) and ethanol. HMF (5) is a flavor component in
many foods,[15] and is even considered a potential anti-tumor
agent.[16] It is listed in its MSDS as a lowhealth hazard
(Category 1) and a mild irritant. An MSDS has been
published for MF (6), where it too is listed as a Category 1
hazard. Furthermore, it is important to note that the vapor
pressures of 5 and 6 are far lower than that of toxic fuel
oxygenates such as methyl tert-butyl ether, and so the
principal means of exposure for most people (during auto-
mobile refueling) would be less of a concern in this case. On
the positive side of this issue, in a future biofuel-100-
compatible vehicle market, the use of oxygenated fuels
which have no (known) potential for abuse as inebriants
may actually prove an advantage.
remaining aqueous layer gave a small quantity of fine, black,
humic material (5% by mass). The total, isolated yield of
simple organics (1–4) is thus a remarkable 85%. Applying the
same reaction conditions to glucose gave 2 (71%), 3 (7%), 1
(8%), 4 (3%), and humic material (4% by mass), for a total
organic yield of 89%. Finally, we were also interested in
applying this method to sucrose, since it is currently the
principal rawmaterial used globally to produce bioethanol. [8]
Here, we isolate 2 (76%); 3 (6%); 1 (4%); 4 (5%), and
humic material (3% by mass), for a total organic yield of
91%.
Systematic variation of the time, temperature, and LiCl
concentration of the reaction have shown that these param-
eters are very nearly optimal for glucose, sucrose, and
cellulose using the present experimental setup. The quantity
of HCl/LiCl solution and dichloroethane solvent used was a
matter of convenient handling and has not been optimized.[9]
The nearly identical results obtained for cellulose and glucose
suggest that the rate-limiting component of the process is not
the hydrolysis of cellulose, but the dehydration of glucose or
the dropwise extraction of the furan products. A more
aggressive extraction protocol would almost certainly
reduce the reaction times, and we will address the optimiza-
tion of this variable in a future report.
While CMF (2) is itself not a biofuel candidate, it is
converted into ethoxymethylfurfural (EMF, 5) in nearly
quantitative yield by stirring it in ethanol solution at room
temperature (Scheme 1). Interestingly, EMF (5), a liquid with
a boiling point of 2358C, is already considered to be a
promising alternative fuel, the energy density of which is
reported to be 8.7 kWhLꢀ1, substantially higher than that of
ethanol (6.1 kWhLꢀ1), and comparable to that of standard
gasoline (8.8 kWhLꢀ1) and diesel fuel (9.7 kWhLꢀ1).[10] EMF
(5) has been tested in blends with diesel fuel by Avantium
Technologies, a spin-off of Royal Dutch Shell, who noted that
“the test yielded positive results for all blends tested. The
engine ran smoothly for several hours. Exhaust analysis
uncovered a significant reduction of soot.”[11]
Although reports of CMF (2) are relatively uncommon in
the chemical literature, it was described as early as 1901 as a
product from the action of dry hydrogen chloride on
cellulose.[17] While the conversion was low (12%), a related
study in which anhydrous HBr was employed showed that the
bromo analogue of 2 could be produced from cellulose in up
to 48% yield, although glucose itself underwent the reaction
in only 11% yield.[18] Finally, a number of other reports
address the preparation of CMF (2) from fructose,[19] which is
consistent with the related, facile conversion of the latter into
HMF (5).[20] Nowhere in the literature is glucose described
giving CMF (2) in good yield.
In conclusion, glucose, sucrose, and cellulose are readily
converted into 5-(chloromethyl)furfural (2), a stable, hydro-
phobic organic liquid. Small quantities of related furans HMF
(1) and 2-(2-hydroxyacetyl)furan (3)[21] as well as levulinic
acid (4)[22] are also isolated, bringing the
Alternatively, CMF (2) could be hydrogenated in 88%
yield using PdCl2 under mild conditions to give 5-methyl-
total yield of simple organic compounds to
85–91%. CMF (2) is easily converted into
the known biofuel EMF (5) or catalytically
hydrogenated to MF (6). The method, as
Scheme 1. Conversion of CMF (2) into EMF (5) and MF (6).
reported here, 1) constitutes the highest
Angew. Chem. Int. Ed. 2008, 47, 7924 –7926ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
7925

Communications
[8] J. Doerfler, H. V. Amorim, Zuckerindustrie 2007, 132, 694.
[9] In practice this may not be a crucial parameter. Since the only
reagent consumed in this process is hydrogen chloride, a flow
reactor could be envisaged where the saccharide is continuously
introduced, 1–4 are continuously removed, and the concentra-
tion of HCl is held constant by introduction of the gas into the
reactor. A preliminary modeling study of this process is included
in the Supporting Information.
[10] G. J. M. Gruter, F. Dautzenberg, Eur. Pat. Appl., 2007,
1834950A1.
[11] Avantium Technologies press release: http://www.avantium.
com/index.php?p = 115&n = 2.
yielding conversion of cellulose into simple, hydrophobic
organic molecules yet described; 2) significantly exceeds the
usable carbon yield of glucose/sucrose fermentation but,
unlike fermentation, is able to utilize cellulose directly; and
3) provides a solution to the long-standing glucose to HMF
(1) conversion problem,[6,7] given that CMF (2) can be easily
hydrolyzed to HMF (1).[20b] While future reports will address
further optimization, scaleup, and applications of the method
to rawbiomass, these preliminary results suggest that this
simple, efficient approach to cellulose deconstruction has the
potential, at the very least, to complement fermentation as a
means to produce biomass-derived automotive fuels, and to
establish furanics both as a renewable energy source and
industrial chemical feedstock of the future.
[12] The literature yield for this reaction is up to 95%, depending on
which solvent is used: K. Hamada, G. Suzukamo, K. Fujisawa,
Eur. Pat. Appl., 1982, 44186A1.
[13] Jap. P. 58013576, 1983.
[14] A reviewer brought the importance of processing these materials
to negligible halogen levels to our attention. To this end, we have
determined that no residual solvent or CMF (2) is present in the
distillates to the detection limits of 1H and 13C NMR, and
analysis by ASTM D4208 (oxygen bomb combustion selective
electrode method) shows that the chlorine content is < 0.5%.
[15] J. M. de Man, Principles of Food Chemistry, 3rd ed., Springer,
Heidelberg, 1999.
Received: April 5, 2008
Published online: August 1, 2008
Keywords: biofuels · carbohydrates · cellulose · furans ·
.
sustainable chemistry
[16] K. Michail, V. Matzi, A. Maier, R. Herwig, J. Greilberger, H.
Juan, O. Kunert, R. Wintersteiger, Anal. Bioanal. Chem. 2007,
387, 2801.
[1] N. S. Lewis, D. G. Nocera, Proc. Natl. Acad. Sci. USA 2006 103,
15729.
[2] A. J. Ragauskas, C. K. Williams, B. H. Davison, G. Britovsek, J.
Cairney, C. A. Eckert, W. J. Frederick, Jr., J. P. Hallett, D. J.
Leak, C. L. Liotta, J. R. Mielenz, R. Murphy, R. Templer, T.
Tschaplinski, Science 2006, 311, 484.
[17] H. J. H. Fenton, M. Gostling, J. Chem. Soc. Trans. 1901, 79, 807.
[18] H. Hibbert, H. S. Hill, J. Am. Chem. Soc. 1923, 45, 176.
[19] For example: K. Sanda, L. Rigal, A. Gaset, J. Chem. Technol.
Biotechnol. 1992, 55, 139.
[3] For a thought-provoking article which examines both sides of
this issue, see Chem. Eng. News, 2007, 8, 12. See also: a) J. Hill,
E. Nelson, D. Tilman, S. Polasky, D. Tiffany, Proc. Natl. Acad.
Sci. USA 2006, 103, 11206; b) R. Hammerschlag, Environ. Sci.
Technol. 2006, 40, 1744; c) A. E. Farrell, R. J. Plevin, B. T.
Turner, A. D. Jones, M. OꢀHare, D. M. Kammen, Science 2006,
311, 506; d) D. Pimentel, T. W. Patzek, Nat. Resour. Res. 2005, 14,
65.
[20] B. F. M. Kuster, H. J. C. Van der Steen, Staerke 1977, 29, 99.
HMFand CMFare also readily interconvertable: a) K. Sanda, L.
Rigal, A. Gaset, Carbohydr. Res. 1989, 187, 15; b) W. N.
Haworth, W. G. M. Jones, J. Chem. Soc. 1944, 667.
[21] 2-(2-Hydroxyacetyl)furan (3) is a well-known aroma component
of several natural substances, for example, toasted oak in wine
barrels: I. Cutzach, P. Chatonnet, R. Henry, D. Dubourdieu, J.
Agric. Food Chem. 1999, 47, 1663.
[4] J. N. Chheda, G. W. Huber, J. A. Dumesic, Angew. Chem. 2007,
119, 7298; Angew. Chem. Int. Ed. 2007, 46, 7164.
[5] Y. Romµn-Leshkov, C. J. Barrett, Z. Y. Liu, J. A. Dumesic,
Nature 2007, 447, 982.
[6] H. Zhao, J. E. Holladay, H. Brown, Z. C. Zhang, Science 2007,
316, 1597.
[22] Levulinic acid is a well-studied derivative of cellulosic biomass
and is considered a promising building block for bio-derived
polymers and fuel additives; see: Biorefineries—Industrial
Processes and Products: Status Quo and Future Directions
(Eds.: B. Kamm, P. R. Gruber, M. Kamm), Wiley-VCH,
Weinheim, 2007.
[7] J. N. Chheda, Y. Romµn-Leshkov, J. A. Dumesic, Green Chem.
2007, 9, 342.
7926 www.angewandte.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2008, 47, 7924 –7926