Direct microwave-assisted hydrothermal depolymerization of cellulose

Article abstract of DOI:10.1021/ja4056273

A systematic investigation of the interaction of microwave irradiation with microcrystalline cellulose has been carried out, covering a broad temperature range (150 → 270°C). A variety of analytical techniques (e.g., HPLC, ¹³C NMR, FTIR, CHN analysis, hydrogen-deuterium exchange) allowed for the analysis of the obtained liquid and solid products. Based on these results a mechanism of cellulose interaction with microwaves is proposed. Thereby the degree of freedom of the cellulose enclosed CH₂OH groups was found to be crucial. This mechanism allows for the explanation of the different experimental observations such as high efficiency of microwave treatment; the dependence of the selectivity/yield of glucose on the applied microwave density; the observed high glucose to HMF ratio; and the influence of the degree of cellulose crystallinity on the results of the hydrolysis process. The highest selectivity toward glucose was found to be ～75% while the highest glucose yield obtained was 21%.

Full text of DOI:10.1021/ja4056273

Communication
pubs.acs.org/JACS
Direct Microwave-Assisted Hydrothermal Depolymerization of
Cellulose
Jiajun Fan,† Mario De bruyn,† Vitaliy L. Budarin,† Mark J. Gronnow,† Peter S. Shuttleworth,‡
Simon Breeden,† Duncan J. Macquarrie,† and James H. Clark*,†
†Green Chemistry Centre of Excellence, Department of Chemistry, University of York, York, YO10 5DD, U.K.
‡Departamento de Física de Polímeros, Elastom
́ ́
eros y Aplicaciones Energeticas, Instituto de Ciencia y Tecnología de Polímeros,
CSIC, Madrid, 28006, Spain
S
* Supporting Information
water is an effective microwave energy absorber.6,9 The
beneficial effect of microwaves toward cellulose hydrolysis has
been reported previously be it only in the presence of strong
acid catalysts.10 Here we report our preliminary investigations
on the use of hydrothermal microwave processing, moderating
only the temperature and microwave field density to convert
cellulose to sugars without additives, making it more
industrially favorable.
Initially these experiments were performed in a high energy
density laboratory (35 mL) CEM Discover SP (CD)
microwave with a maximum pressure of 300 psi (230 °C,
accuracy verified previously).11 Experiments were run in
dynamic mode allowing for a system with a controlled variable
power input to achieve the desired temperature. During the
initial water heating phase, a maximum power input of 300 W
was recorded. Each experiment was repeated in triplicate, and
the obtained glucose yields were within 7%. The resulting
microwave hydrolysis products were analyzed by NMR, HPLC,
GC, and electronspray ionization mass spectrometry (ESI).
This allowed for both qualitative and quantitative analysis of the
formed products (Figures 1 and S1).
ABSTRACT: A systematic investigation of the interaction
of microwave irradiation with microcrystalline cellulose has
been carried out, covering a broad temperature range (150
→ 270 °C). A variety of analytical techniques (e.g., HPLC,
13C NMR, FTIR, CHN analysis, hydrogen−deuterium
exchange) allowed for the analysis of the obtained liquid
and solid products. Based on these results a mechanism of
cellulose interaction with microwaves is proposed. There-
by the degree of freedom of the cellulose enclosed
CH2OH groups was found to be crucial. This mechanism
allows for the explanation of the different experimental
observations such as high efficiency of microwave
treatment; the dependence of the selectivity/yield of
glucose on the applied microwave density; the observed
high glucose to HMF ratio; and the influence of the degree
of cellulose crystallinity on the results of the hydrolysis
process. The highest selectivity toward glucose was found
to be ∼75% while the highest glucose yield obtained was
21%.
HPLC analysis facilitated the identification and quantification
of glucose, fructose, xylose, cellobiose, and levoglucosan present
in the hydrolysates (Figure 1A). The presence of xylose is
attributed to residual hemicellulose in the cellulose samples.12
As can be seen in Figure 1B/C the total yield of sugar shows a
distinct maximum at 220 °C where 14% of the cellulose is
converted with a 75% selectivity to glucose (11% yield). 13C
NMR and ESI characterization confirmed the presence of 5-
hydroxymethylfurfural (HMF) as a major secondary byproduct
(see Supporting Information (SI) Figure S1a−h). Note that it
is well-known that in subcritical water (100 °C < T < 373.9 °C)
HMF is one of the major products from glucose degradation.13
The total cellulose depolymerization in the CD microwave
experiment was limited to 14% suggesting that only a small
fraction of the cellulose is available for hydrolysis under these
conditions. Microcrystalline cellulose consists of both amor-
phous (13%) and crystalline regions (87%) with significantly
different structural and thermal properties (Figure S11).
Thermal transformations in amorphous cellulose take place
between 180 and 230 °C,14 which is consistent with the
he 21st century has generated a demand for new sources
T
of refinery feedstocks due to depleting oil reserves.1 One
such feedstock is sugars from plants which can be converted to
chemicals and fuels through fermentation and/or microbial
processes.2 However, sugar is also a food source making its use
controversial. Cellulose, the most abundant and readily
renewable source of biomass on the planet, has potential to
be converted to nonfood-competing sugar, and hence,
methodologies that optimize its breakdown to simple sugars
are highly sought after.3 As such, various catalytic, thermal, and
enzymatic approaches have been explored.4 Enzymatic
depolymerization tends to need long residence times and
high dilution ratios, but is highly selective, while thermal
processes are limited by poor energy efficiency and thermal
conductivity.5,6 Furthermore, the high temperatures required
for thermal treatment limits glucose selectivity, with the main
products being anhydrosugars (e.g., levoglucosan) and
secondary breakdown products (e.g., HMF and phenols).7
Hydrothermal processing offers a route to hydrolyze cellulosic
biomass into simple sugars using elevated pressure and
temperature.8 Combining this technique with microwave
heating presents a potentially faster, more efficient, and
selective method for the thermal treatment of biomass, as
Received: June 5, 2013
© XXXX American Chemical Society
A
dx.doi.org/10.1021/ja4056273 | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Communication
structure, but can be broken at elevated temperatures.17 To
understand the activation of the amorphous cellulose region on
a molecular level, a proton/deuterium exchange experiment was
performed at 220 °C under both conventional and microwave
conditions. Using ATR-FTIR spectroscopy the accessibility of
the different hydrogen bonds could be measured as the ratios
between the areas of the OH and OD peaks in the respective
3400 and 2500 cm−1 regions (Figure 2B).
It was found that in the remaining cellulose, after
conventional hydrothermal treatment, the intrachain O(2)−
H···O(6) (3416 and 2529 cm−1) hydrogen bond is the most
accessible followed respectively by the O(3)−H···O(5) (3337
and 2480 cm−1) and O(3)···H−O(6) (3222 and 2427 cm−1)
hydrogen bonds (see Figure 2C). In contrast, the CH2O(6)
related hydrogen bonds in the remaining cellulose after
microwave (CD) treatment, i.e. intrachain O(2)−H···O(6)
and interchain O(3)···H−O(6), were found to be more prone
to proton/deuterium exchange than the intrachain O(3)−
H···O(5) hydrogen bond. It can thus be concluded that after
microwave-assisted heating the protons associated with the
CH2O(6)H group are more accessible. The mentioned values
and assignments are in agreement with the literature.18
At temperatures below 180 °C the CH2OH groups are
hindered from interacting with microwaves while they are
strongly involved in hydrogen bonding within both the
amorphous and crystalline regions (Figure 3A).19
Figure 1. HPLC analysis of the sugars formed during hydrolysis of
cellulose in the CD laboratory microwave (1 min holding time): (A)
HPLC trace of the hydrolysate obtained at 220 °C; (B) two-
dimensional trace of products formed over the temperature range
170−230 °C; (C) sugar mixture composition as a function of
temperature.
temperature at which the CD hydrothermal microwave
experiment yields the maximal amount of glucose (around
220 °C). Therefore it can be concluded that the crystalline
region of cellulose was not involved throughout the process, as
major structural changes in this region only occur well above
220 °C.15
Hydrogen bonding is a critical factor in polysaccharide
structures, and it is well-known that the depolymerization of
cellulose depends strongly on the structure of the hydrogen
bond network.16 More specifically the structure of cellulose
involves intersheet, interchain, and intrachain hydrogen bonds
(Figure 2A). These impart rigidity and stability to the cellulose
Figure 3. Schematic representation of the cellulose-microwave
interaction as a function of temperature: (A) mechanism of CH2OH
group activation; (B) scheme of cellulose degradation toward acids
and aldehydes.
Above the softening temperature (180 °C) these CH2OH
groups could be involved in a localized rotation in the presence
of microwaves.20 As such they could act similarly to “molecular
radiators” allowing for the transfer of microwave energy to their
surrounding environment21
In view of the limited presence of water inside the rigid
cellulose framework, this is likely to involve collisions between
the CH2OH groups and the anomeric C1 of the same glucose
ring thus forming levoglucosan.22 The latter can easily
hydrolyze to glucose (Figure 3B).23 The data obtained from
both conventional and microwave based experiments (see also
Figure S2) confirm the molecular explanation of cellulose
decomposition and its direct relationship to microwave
activation of the CH2OH pendant groups. Indeed, as shown
Figure 2. (A) Schematic representation of cellulose and its hydrogen
bond network; (B) ATR-FTIR spectrum of original cellulose
(bottom), the deuterated cellulose after conventional (middle) and
microwave (top) treatment respectively; (C) ratios of either
conventional (bottom) or microwave (top) O−D to O−H
deconvoluted peak areas.
B
dx.doi.org/10.1021/ja4056273 | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Communication
in Figure 4A, from 190 °C onward microwave heating is found
to be markedly more efficient toward the hydrolysis of cellulose
than conventional heating.
35 W L−1. As CD involves monomode operation, and CM6
multimode, this difference could be even further accentuated.
The dependence of the glucose yield on the microwave
power density suggests that the microwave activation has a
strong kinetic dimension. Two competitive processes deter-
mine the speed of the CH2OH group rotation: (i) acceleration
by interaction with microwave photons and (ii) deceleration
through interaction (e.g., collision, electromagnetic) with
neighboring groups. The dominance of either process depends
on the degree of freedom of the CH2OH groups. For
depolymerization of cellulose to occur the CH2OH groups
need to acquire the activation energy necessary to provoke the
above proposed SN2 reaction. At high microwave power
densities (800 W·L−1) this can already be achieved below 230
°C, while upon the use of lower microwave densities (35
W·L−1) more elevated temperatures (240−260 °C) are
required to liberate the CH2OH groups.
Figure 4. Comparison of CD microwave and conventional hydrolysis:
(A) cellulose mass loss; (B) individual sugar yields (220 °C)
(logarithmic Y-axis).
Based on HPLC measurement a glucose yield of 21% was
found using the CM6 microwave. This is higher than CD
microwave operation at 220 °C (11%) and conventional
heating at 250 °C (16%).27
The maximal glucose yield at 220 °C under microwave
treatment was found to be nearly 50 times higher than that
under similar conventional hydrolysis conditions. Interestingly,
levoglucosan is only obtained under microwave conditions
(Figure 4B); its presence suggests that the microwave active
center within cellulose is not accessible to the hydrolyzing
media.24
From 13C NMR (relaxation time 30 s) the ratio of HMF to
glucose in the samples could be accurately determined (Figures
6A and S3a−c), and thus it was possible to calculate an HMF
According to the proposed model, the yield of glucose
obtained below 220 °C is only related to the depolymerization
of amorphous cellulose and not the crystalline content that
becomes active above 220 °C.15,25 While the CD system is
restricted to 230 °C, the efficiency of microwave hydrothermal
depolymerization of cellulose at temperatures up to 270 °C was
evaluated using a CEM MARS6 (CM6) microwave. With both
systems the experiments were performed at the same scale and
mass ratio (2 g of cellulose to 20 mL of water). Also, the same
heating rate of ∼15 K·min−1 and holding time (1 min) at the
final temperature were applied. In addition, the reaction
temperatures for the CM6 system were verified by measuring
its pressure (e.g., 252 °C = 600 psi).26 As expected the glucose
yield at 250 °C is higher than the one obtained at 220 °C,
reaching 21% (Figure 5A). This increase of yield is
accompanied by a decrease in selectivity from 75% to 36%
(Figure 5B). Surprisingly, using the CM6 no activation/
depolymerization of amorphous cellulose at 220 °C is observed.
This can be explained by the different power densities of the
CD and CM6 microwave systems, respectively being 800 and
Figure 6. (A) The anomeric carbon regions in the 13C NMR spectra
of the freeze-dried cellulose hydrolysis products as a function of the
holding time (CM6, 250 °C); (B) the yields of glucose and HMF as a
function of the holding time (CM6, 250 °C) determined from both
the 13C NMR and HPLC.
yield of 7.5% using the determined glucose value as a standard
(Figure 6B). The ratio of glucose to HMF obtained using the
CM6 microwave is substantially higher than the one obtained
under conventional conditions (2.8 vs 1).27 The high glucose to
HMF ratio at the given hydrolysis temperature of 250 °C
further supports the above proposed mechanism. Indeed, while
microwaves are able to activate the CH2OH groups enclosed in
cellulose cavities, they are unable to activate these groups in
individual glucose monomers as their energy transfers very
efficiently to the surrounding water molecules. The remaining
35% (conversion ∼60%) is likely to be composed of water and
a range of organic acids and aldehydes, most of which were
removed during the freeze-drying process necessary to prepare
the NMR samples.28
Figure 5. Comparison of CD and CM6 experiments: (A) glucose
yield; (B) glucose selectivity for CD microwave. The contributions
from the amorphous and crystalline regions are indicated.
C
dx.doi.org/10.1021/ja4056273 | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Communication
(4) (a) Ladisch, M. R.; Ladisch, C. M.; Tsao, G. T. Science 1978, 201,
743. (b) Geboers, J. A.; Van de Vyver, S.; Ooms, R.; de Beeck, B. O.;
Jacobs, P. A.; Sels, B. F. Catal. Sci. Technol. 2011, 1, 714. (c) Sasaki,
M.; Kabyemela, B.; Malaluan, R.; Hirose, S.; Takeda, N.; Adschiri, T.;
Arai, K. J. Supercrit. Fluids 1998, 13, 261. (d) Kobayashi, H.;
Komanoya, T.; Hara, K.; Fukuoka, A. ChemSusChem 2010, 3, 440.
(e) Mascal, M.; Nikitin, E. B. Angew. Chem., Int. Ed. 2008, 47, 7924.
(f) Eyal, A.; Jansen, R.; Ltd, H. C., Ed. US, 2012. (g) Fan, L. T.;
Gharpuray, M. M.; Lee, Y. H. Cellulose Hydrolysis; Springer-Verlag:
Berlin, 1987. (h) Saeman, J. F. Ind. Eng. Chem. 1945, 37, 43.
(5) Klyosov, A. A. Biochemistry 1990, 29, 10577.
(6) Krieger-Brockett, B. Res. Chem. Intermed. 1994, 20, 39.
(7) Lin, Y. C.; Cho, J.; Tompsett, G. A.; Westmoreland, P. R.; Huber,
G. W. J. Phys. Chem. C 2009, 113, 20097.
(8) (a) Orozco, A.; Ahmad, M.; Rooney, D.; Walker, G. Process Safety
and Environmental Protection 2007, 85, 446. (b) Ross, A. B.; Biller, P.;
Kubacki, M. L.; Li, H.; Lea-Langton, A.; Jones, J. M. Fuel 2010, 89,
2234.
(9) (a) Wan, Y.; Chen, P.; Zhang, B.; Yang, C.; Liu, Y.; Lin, X.; Ruan,
R. J. Anal. Appl. Pyrolysis 2009, 86, 161. (b) Zhao, X.; Song, Z.; Liu, H.;
Li, Z.; Li, L.; Ma, C. J. Anal. Appl. Pyrolysis 2010, 89, 87. (c) Luque, R.;
Menendez, J. A.; Arenillas, A.; Cot, J. Energy Environ. Sci. 2012, 5,
5481. (d) Kappe, C. O. Angew. Chem., Int. Ed. 2004, 43, 6250.
(10) Wu, Y. Y.; Fu, Z. H.; Yin, D. L.; Xu, Q.; Liu, F. L.; Lu, C. L.;
Mao, L. Q. Green Chem. 2010, 12, 696.
(11) Budarin, V. L.; Zhao, Y. Z.; Gronnow, M. J.; Shuttleworth, P. S.;
Breeden, S. W.; Macquarrie, D. J.; Clark, J. H. Green Chem. 2011, 13,
2330.
Over time the concentrations of both glucose and HMF are
reduced, therewith showing their intermediate nature in the
cellulose hydrolysis process (Figure 6B). It was found that
organic acids were produced, due to the degradation of HMF,
progressively reducing the pH of the reaction mixture to 3.5,
which is in good agreement with the literature.29 These acids
could further accelerate the whole cellulose depolymerization
process (see Figure 3B). As to the possible formation of
humins, the mass balance combined with the 13C NMR and
elemental analysis data rules out this possibility (SI, section X).
In conclusion, microwave-mediated cellulose hydrolysis re-
sulted in high glucose yields and selectivities (respectively up to
21% and 75%) compared to conventionally heated noncatalytic
processes. Furthermore the amorphous region was activated at
significantly lower temperatures than possible under conven-
tional conditions. Moreover, it was demonstrated that the
glucose yield and selectivity can be controlled by altering the
microwave power density/distribution. Additionally, it was
found that the weakening of the hydrogen bond network within
the molecular cellulose matrix at temperatures >180 °C allows
the polar CH2OH groups to act similarly to “molecular
radiators”, initiating the cleavage of the polysaccharide chain
and selective formation of glucose. Currently, studies are
underway to further improve these yields with the aim of
equaling or surpassing the presently highest reported glucose
yield (∼30%), which is based on the use of a strong acid
catalyst.30 In this respect we are aiming at increasing the yield/
selectivity through recycling operations and possible continu-
ous processing. This would make microwave-assisted hydrolysis
of cellulose a strong contender for industrial transformation of
biomass into sugars.
(12) Saka, S.; Ueno, T. Cellulose 1999, 6, 349.
(13) (a) Chamblee, T. S.; Weikel, R. R.; Nolen, S. A.; Liotta, C. L.;
Eckert, C. A. Green Chem. 2004, 6, 382. (b) Luo, C.; Wang, S.; Liu, H.
Angew. Chem., Int. Ed. 2007, 46, 7636. (c) Kraus, G. A.; Lee, J. J. J.
Surfactants Detergents 2013, 16, 317.
(14) (a) Picker, K. M.; Hoag, S. W. J. Pharm. Sci. 2002, 91, 342.
(b) Budarin, V. L.; Clark, J. H.; Lanigan, B. A.; Shuttleworth, P.;
Macquarrie, D. J. Bioresour. Technol. 2010, 101, 3776.
ASSOCIATED CONTENT
■
(15) Wada, M. J. Polym. Sci., Part B: Polym. Phys. 2002, 40, 1095.
(16) Kovalenko, V. I. Russ. Chem. Rev. 2010, 79, 231.
S
* Supporting Information
Experimental details of conventional and microwave experi-
ments, setups, T-, P-, and power profiles; ESI, GC, 13C NMR
spectra and CHN analysis of cellulose hydrolysis products after
microwave treatment; and FTIR spectra of cellulose at different
temperatures. This material is available free of charge via the
Internet at http://pubs.acs.org.
(17) Matthews, J. F.; Bergenstrahle, M.; Beckham, G. T.; Himmel, M.
E.; Nimlos, M. R.; Brady, J. W.; Crowley, M. F. J. Phys. Chem. B 2011,
115, 2155.
(18) Fan, M.; Dai, D.; Huang, B. In Fourier Transform - Materials
Analysis; Salih, M., Ed.; InTech: 2012; p 45.
(19) Kondo, T.; Sawatari, C. Polymer 1996, 37, 393.
(20) Gabriel, C.; Gabriel, S.; Grant, E. H.; Halstead, B. S. J.; Mingos,
D. M. P. Chem. Soc. Rev. 1998, 27, 213.
(21) (a) Kappe, C. O.; Pieber, B.; Dallinger, D. Angew. Chem., Int. Ed.
2013, 52, 1088. (b) de la Hoz, A.; Diaz-Ortiz, A.; Moreno, A. Chem.
Soc. Rev. 2005, 34, 164. (c) Kaiser, N. F. K.; Bremberg, U.; Larhed, M.;
Moberg, C.; Hallberg, A. Angew. Chem., Int. Ed. 2000, 39, 3596.
(22) (a) Agarwal, V.; Dauenhauer, P. J.; Huber, G. W.; Auerbach, S.
M. J. Am. Chem. Soc. 2012, 134, 14958. (b) Kawamoto, H.; Hatanaka,
W.; Saka, S. J. Anal. Appl. Pyrolysis 2003, 70, 303. (c) Mamleev, V.;
Bourbigot, S.; Le Bras, M.; Yvon, J. J. Anal. Appl. Pyrolysis 2009, 84, 1.
(23) Bennett, N. M.; Helle, S. S.; Duff, S. J. B. Bioresour. Technol.
2009, 100, 6059.
AUTHOR INFORMATION
■
Corresponding Author
james.clark@york.ac.uk
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
The authors would like to thank members of the Green
Chemistry Centre of Excellence for their input and useful
discussions. P.S. gratefully acknowledges the Ministerio de
(24) Lakshmanan, C. M.; Gal-Or, B.; Hoelscher, H. E. Starke 1970,
̈
22, 221.
́
Ciencia e Innovacion for the concession of a Juan de la Cierva
(JCI-2011-10836) contract.
(25) Horikawa, Y.; Sugiyama, J. Cellulose 2008, 15, 419.
(26) In The engineering toolbox: http://www.engineering-toolbox.
com/saturated-steam-properties-d_101.html.
(27) Weingarten, R.; Conner, W. C.; Huber, G. W. Energy Environ.
Sci. 2012, 5, 7559.
(28) Moeller, M.; Harnisch, F.; Schroeder, U. Biomass Bioenergy
2012, 39, 389.
(29) Kimura, H.; Nakahara, M.; Matubayasi, N. J. Phys. Chem. A
2011, 115, 14013.
REFERENCES
■
(1) (a) Alonso, D. M.; Wettstein, S. G.; Dumesic, J. A. Chem. Soc.
Rev. 2012, 41, 8075. (b) Hallock, J. L.; Tharakan, P. J.; Hall, C. A. S.;
Jefferson, M.; Wu, W. Energy 2004, 29, 1673.
(2) Lin, Y.; Tanaka, S. Appl. Microbiol. Biotechnol. 2006, 69, 627.
(3) (a) Carroll, A.; Somerville, C. Ann. Rev. Plant Biol. 2009, 60, 165.
(b) Naik, S. N.; Goud, V. V.; Rout, P. K.; Dalai, A. K. Renewable
Sustainable Energy Rev. 2010, 14, 578.
(30) Zhang, Z.; Zhao, Z. K. Carbohydr. Res. 2009, 344, 2069.
D
dx.doi.org/10.1021/ja4056273 | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX