Bioresource Technology 148 (2013) 255260

Contents lists available at ScienceDirect

Bioresource Technology
journal homepage: www.elsevier.com/locate/biortech

Thermal decomposition of lignocellulosic biomass in the presence of acid

catalysts
Cherif Larabi a,b, Walid al Maksoud a,b, Kai C. Szeto a, Anne Roubaud c, Pierre Castelli c,
Catherine C. Santini a,, Jean J. Walter b
a
b
c

Universit de Lyon, ICL, C2P2, UMR 5265 CNRS-ESCPE Lyon-UCBL, 43 bd du 11 Novembre 1918, 69616 Villeurbanne Cedex, France
Synthopetrol, 37 Rue des Mathurins, 75008 Paris 8, France
CEA, LITEN, LTB, F 38054 Grenoble, France

h i g h l i g h t s
 The decomposition of lignocellulose is studied by in situ and ex situ techniques.
 Three fractions, solid, liquid and gas are quantied and characterized.
 The degradation temperature is lowered by 100 C in the presence of H3PW12O40.
 At 300 C with H3PW12O40, the amount of liquid collected reaches 30 wt.%.
 Traces of furfural was observed from pure wood, with H3PW12O40 it reaches 2 wt.%.

a r t i c l e

i n f o

Article history:
Received 17 July 2013
Received in revised form 8 August 2013
Accepted 10 August 2013
Available online 23 August 2013
Keywords:
Thermolysis
Pine wood
Acidic support
Heteropolyacid
Thermal pretreatment

a b s t r a c t
Transformation of lignocellulosic biomass to biofuels involves multiple processes, in which thermal
decomposition, hydrotreatment are the most central steps. Current work focuses on the impact of several
solid acids and Keggin-type heteropolyacids on the decomposition temperature (Td) of pine wood and the
characterization of the resulted products. It has been observed that a mechanical mixture of solid acids
with pine wood has no inuence on Td, while the use of heteropolyacids lower the Td by 100 C. Moreover,
the treatment of biomass with a catalytic amount of H3PW12O40 leads to formation of three fractions:
solid, liquid and gas, which have been investigated by elemental analysis, TGA, FTIR, GCMS and NMR.
The use of heteropolyacid leads, at 300 C, to a selective transformation of more than 50 wt.% of the holocellulose part of the lignocellulosic biomass. Moreover, 60 wt.% of the catalyst H3PW12O40 are recovered.
2013 Elsevier Ltd. All rights reserved.

1. Introduction
The steady depletion of fossil fuels resources throughout the
world as well as ecological and sustainable development issues
have turned current research interest for alternative fuel sources
(Menon and Rao, 2012). Biomass is considered as renewable energy source. Its contribution to the global warming is negligible
compared to fossil fuels, because of its ability to x CO2 from
the atmosphere. Moreover, employing biofuels is reported to give
a positive effect with respect to current greenhouse gas emission
(Menetrez, 2012; Pires et al., 2012). Biofuels can be divided into
rst, second and third generation. First generation comprises
liquid biofuels originated from corn, sugarcane, soybean, oil palm,
etc. The most common rst generation biofuels is bio-ethanol.
Corresponding author. Tel.: +33 (0) 4 72 43 18 10; fax: +33 (0) 4 72 43 17 95.
E-mail address: catherine.santini@univ-lyon1.fr (C.C. Santini).
0960-8524/$ - see front matter 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.biortech.2013.08.070

However these sources are considered as unsustainable, mainly

due to their use of food feed (Poganietz, 2012). Second-generation
biofuels originates from non-edible biomass sources, such as lignocellulosic materials, forest residues and wastes. The synthesis
of second generation biofuels is indeed independent on the use
of crops for food consumption (Poganietz, 2012). However, extensive research efforts are required to make this process economical
(Macrelli et al., 2012). The third generation biofuel is known also
as Oilgae. It is believed that the cost of their production is low.
Moreover, the energy produced as biofuel per acre is higher
compared to the land required by other conventional feedstock
(Nigam and Singh, 2011). Nevertheless, the production of fuels
from microalgae is still at its early stage of development (Fiorese
et al., 2013). On the other hand, bioreneries for second generation biofuel production can easily be implemented and size
adapted just near the lignocellulosic material, this is not the case
for the algae resources.

256

C. Larabi et al. / Bioresource Technology 148 (2013) 255260

The transformation of lignocellulosic materials to biofuel as

well as chemical feedstocks, requires thermal or thermocatalytic
treatment. The challenge lies in depolymerizing the complex structure of the cellulose, hemicellulose and lignin and ideally further
upgrade without decomposition into CO, CO2 and H2O. A possibility to circumvent this is to perform the thermolysis at lower temperature. It is reported that during the thermal decomposition
process of lignocellulosic biomass, the holocellulose is decomposed
rst followed by lignin at higher temperatures (Gronli et al., 2002).
In the case of the pine wood, it was found that the decomposition
of the holocellulose starts at 275 C and the lignin decomposes at
ca. 350 C (Larabi et al., 2013). Furanic compounds including furfural were released during the decomposition process (Larabi et al.,
2013; Sagehashi et al., 2006). Upon hydrotreatment, furfural can
be converted to methylfuran (Sitthisa et al., 2011), which is reported to have a high octane number and can be used as additive
for conventional fuels (Roman-Leshkov et al., 2007). Some of the
thermal reactions such as dehydration, depolymerization and even
decarboxylation can be catalyzed by solid acidic catalyst (Chheda
and Dumesic, 2007; Wan et al., 2009). Solid acids under oxidative
conditions are largely used in the transformation of the wood into
paper pulp (Gaspar et al., 2007), and cellulose into chemicals (Deng
et al., 2012; Guo et al., 2012; Tian et al., 2010). Herein, we focus on
the transformation of pine wood, in the presence of acid catalysts
under reductive condition in order to lower the temperature of
the thermolysis, to produce valuable chemicals, reduce the formation of CO and CO2 and the potential cost.
For this purpose, the thermal decomposition of pine wood was
performed in batch and dynamic reactors, under hydrogen, at temperature lower than 300 C in the presence of Keggin heteropolyacid
(H3PW12O40). Keggin type heteropolyacid was chosen for its stability
and strong acidity, which is more pronounced that of typically used
strong mineral acids (Cui et al., 2011). Moreover, heteropolyacid
maintains its property in the liquid phase as well as a solid by precipitation or grafting on supports (Marme et al., 1998; Rao et al., 2005;
Shimizu et al., 2009). During this investigation, three main fractions
were collected, quantied and characterized, by various techniques
including FTIR, GCMS, elemental analysis and solid state NMR.
2. Experimental
2.1. Materials
The lignocellulosic biomass used is originated from pine wood.
It was grounded in a Retcsh type RM100 mortar mill, and then
sieved to give particle size less than 0.5 mm. In order to distinguish
between the water formed during the decomposition process and
the physisorbed, the wood dust was pretreated and dried at
110 C, then stored in the glove box.
c-Al2O3 (Puralox TM 50/150, Sasol), SiO2Al2O3 (Azko Nobel),
ZSM-5 (Sasol), SiO2 (Evonik aerosil-200), H3PW12O40
24H2O
(HPW, Phosphotungstic acid hydrate), H3PMo12O40
24H2O (HPMo,
Phosphomolybdic acid hydrate), H4SiW12O40
24H2O (HSiW,
Silicotungstic acid hydrate), H4SiMo12O40
24H2O (HSiMo,
Silicomolybdic acid hydrate), (Aldrich) were used as received. 11Molybdo-1-vanadophos-phoric, H4SiMo11VO40
24H2O (HSiMoV),
and
10-Molybdo-2-vanadophosphoric
H4SiMo10V2O40
24H2O
(HSiMoV2), acids were synthesized according to the literature
(Tsigdino and Hallada, 1968; Berndt et al., 1998). Their purity
was checked by 31P and 29Si liquid-NMR.
2.2. Characterization of product
2.2.1. In-operando DRIFTS-GC/MS
The experiments were carried out in an integrated system
comprising mass ow controllers (Brooks), FT-IR adapted high

temperature reaction chamber (Harrick Scientic) and online

GC/MS (Agilent GC 6850 MS 5975C). The reaction chamber was
equipped with ZnSe windows and tted into the Praying Mantis
optical unit also provided by Harrick. In the glove box, about
15 mg of pine wood were placed onto a porous stainless steel frit
in the reaction chamber. A selected gas (argon or hydrogen) was
continuously owing through the pine wood bed. FT-IR spectra
were recorded in a Nicolet 6700 spectrophotometer with a MCT
detector and 4 cm 1 resolution. The formed compounds were
separated by a MS compatible 25 m PORA BOND Q column and
analyzed in the online GCMS.
2.2.2. Nuclear magnetic resonance (NMR) characterization
Solution NMR spectra were recorded on BRUKER AVANCE 300
spectrometer (1H: 300.1 MHz, 13C: 75.4 MHz; 31P: 121.5 MHz;
29
Si: 59.6 MHz). Chemical shifts were measured relative to 85%
H3PO4 aqueous solution for 31P. 29Si chemical shifts are referenced
to Me4Si in DMSO-d6 using the substitution method.
Solid state NMR spectra were collected on BRUKER AVANCE III
500 spectrometer operating at 202.5 MHz for 31P, 125 MHz for
13
C. The zirconia impeller of 4 mm is lled with the desired product
and sealed with a kel-f stopper. It was then transferred into the
probe Bruker CP 4 mm spectrometer allowing rotation of the rotor
at a speed of 10 kHz. The time between two acquisitions was
always optimized to allow complete relaxation of the protons.
2.2.3. Elemental analyses (C, H, and N)
Microanalyses were performed at the Welience Ple Chimie
Molculaire Facult des Sciences Mirande (Dijon, France), using
CHNS/O thermo electron ash 1112 Series elemental analyzer.
2.2.4. Ex situ IR analysis
Diffuse reectance Fourier-transformed infrared (DRIFT) spectra were recorded on a Nicolet 6700-FT spectrometer using a cell
equipped with CaF2 window. Typically, 64 scans were accumulated
for each spectrum (resolution 4 cm 1).
2.2.5. IR quantication of the CO and CO2 analysis
The quantication of evolved CO and CO2 was performed by
transmission FT-IR spectroscopy using a Nicolet 5700-FT spectrometer and infrared cell equipped with CaF2 windows. Typically,
64 scans were accumulated for each spectrum (resolution 1 cm 1).
To estimate the amount of gases (CO, CO2) in the reactor during the
pyrolysis by FTIR technique, the IR detector was calibrated toward
CO and CO2. A calibration curve was obtained by reporting for a given gas pressure the corresponding area related to their asymmetric stretching frequency integrated between 22202140 cm 1 for
CO and 24002250 cm 1 for CO2. The calibration curve is provided
in Fig. S.1.
2.3. Design of experiments
A mixture of grounded wood and a desired amount of heteropolyacids (HPA) was stirred in water at 25 C for 1 h. After evaporation of water at 60 C, the impregnated wood was dried 2 h at
80 C under vacuum (10 5 mbar), then stored in glove box. A mixture of grounded wood and 10 wt.% of solid acid (SA) was mechanically mixed, dried 2 h at 80 C under vacuum (10 5 mbar), then
stored in glove box.
2.3.1. Thermogravimetric analysis
Approximately 10 mg of the sample was placed in an Al2O3 crucible and heated under 30 mL min 1 of nitrogen. Three independent decomposition studies were performed, (i) the weight loss
from room temperature to 700 C in the presence of 10% in weight
of solid acid (SA) or heteropolyacids (HPA) with a heating rate of

C. Larabi et al. / Bioresource Technology 148 (2013) 255260

257

5 C min 1. (ii) The weight loss from room temperature to 700 C in

the presence of different amount of H3PW12O40
24 H2O (2, 10, 20,
30, 40, 50 wt.%). (iii) The weight loss proles with time under isothermal conditions (150, 200, 250 C) with a heating rate of
5 C min 1 with or without H3PW12O40
24H2O.

275, 300 C) in a reactor under hydrogen ux. The solid obtain in

each experience was analyzed by ex situ DRIFT.

2.3.2. In-operando DRIFT/GCMS

The crucible in the DRIFT cell was lled with dried wood
(15 mg) or with the dried wood impregnated with 10 wt.% of
H3PW12O40
24H2O, HPW, in the glove box under argon atmosphere. Once the cell was inserted into the spectrophotometer, a
controlled mass ow of argon or hydrogen (6 ml min 1) was introduced into the reaction chamber at 1 bar. Heating was programmed from 20 C to a desired temperature, using heating rate
of 5 C min 1. The nal temperature was then kept for 5 h. A DRIFT
spectrum of 64 scans was recorded every minute. A continuous
analysis by GCMS of the released volatiles was also carried out.

The thermal treatment of the pine wood in the presence of acid

catalysts has been performed mainly under hydrogen atmosphere.
First, the impact of the solid acids (SA = Al2O3, SiO2Al2O3, ZSM-5
and SiO 2 ) compared to the water soluble heteropolyacids
(HPA = H 3 PW 12 O 40 , H 3 PMo 12 O 40 , H 4 SiW 12 O 40 , H 4 SiMo 12 O 40 ,
H4PVMo11O40, H5PV2Mo10O40) has been investigated.
The DRIFT spectra of dried wood (Larabi et al., 2013), and of the
wood sample impregnated with 10 wt.% H3PW12O40 are represented in (Fig. S.2). In Fig. S.2, besides the absorption bands of pine
wood, supplementary peaks at 1089, 989, 896 and 818 cm 1 are
observed, attributed to m(POa), m(W = Ot), m(WOcW) and m
(WOeW), respectively where a, t, c and e correspond to different
oxygen position atoms in Keggin structure (internal, terminal,
corner and edge-shared) (Caliman et al., 2010).
The 31P solid state NMR of a wood sample impregnated with
10 wt.% H3PW12O40 (Fig. S.3b), shows one peak at 15.5 ppm in
agreement with the chemical shift observed in 31P liquid NMR
for H3PW12O40 (Fig. S.3a), suggesting that the Kegging structure
of the HPA is preserved after impregnation.

2.3.3. Scale up wood thermolysis

In this part, experiments were performed either in a continuous
ow reactor or batch stainless autoclave under 1 bar of hydrogen.
2.3.3.1. Dynamic reactor. In a glove box, 1 g of dried wood, or 1.1 g
of wood impregnated with 10 wt.% of H3PW12O40
24H2O were
introduced in the reactor, then placed in the oven and connected
to a gas line of hydrogen. The wood thermolysis was carried out
in a temperature range between 150 and 300 C under pressure
of 1 bar and a ow of 5 mL min 1. This allowed the analysis of light
hydrocarbons by an online GC (Varian CP 3800).
2.3.3.2. Batch reactor. 1 g of dried wood or 1.1 g in the case of the
impregnated wood with 10 wt.% of HPA, were introduced into
the batch reactor and heated at different temperatures. The reactor
was connected and ashed under vacuum to remove the water that
could be adsorbed while transferring the biomass sample and during the xation of the autoclave, in order to distinguish between
the adsorbed water and the one produced from the degradation
process. Afterwards 1 bar of hydrogen was introduced and the
autoclave was heated at different temperature ranging from 150
to 300 C and kept for two hours with a heating rate of 5 C min 1.
The liquid phase collected after distillation under vacuum (10 2
mbar) was characterized by GCMS, and the amount of furfural
produced was evaluated. The residual solid was weighed, and analyzed by elemental analyses (C, H), and 13C CP MAS solid state
NMR.
2.3.4. FTIR characterization and evaluation of gas phase
Under argon, 250 mg of wood or 275 mg in the case of the
impregnated wood with H3PW12O40
24H2O were placed in the
reactor. The cell was closed and the argon was removed under vacuum (10 5 mbar). The sample was then heated with a rate of
5 C min 1 until the target temperature (150, 200, 250, 275,
300 C) and kept at that temperature for 2 h. The gases obtained
were analyzed by mean of transmission FT-IR spectroscopy using
a NICOLET 5700 FT-IR spectrometer.
CO calibration: The peaks were integrated between 2220 and
2140 cm 1 and the variation of the amount (mmol) was proportional to the surface with a coefcient of 2.684 (Fig. S.1a and b).
CO2 calibration: The peaks were integrated between 2400 and
2250 cm 1 and the variation of the quantity (mmol) was proportional to the surface with a coefcient of 0.022 (Fig. S.1c and d).
2.3.5. Ex situ DRIFT
The samples of wood and wood impregnated with 10% of HPA
are thermally treated at different temperatures (150, 200, 250,

3. Results and discussion

3.1. Thermo-gravimetric studies of wood and wood/solid acid or wood/

heteropolyacid under argon
Mixtures of wood/SA and wood/HPA from 40 C to 700 C with
heating rate of 5 C min 1 have been investigated by TGA. The evolution of the weight loss with the temperature in the presence of
various SA (SA = Al2O3, SiO2Al2O3, ZSM-5 and SiO2) is depicted
in Fig. 1a. The decomposition temperature (Td) of the wood in
the presence of the selected SA shows no change (275 C). The evolution of the weight loss with the temperature for wood/HPA,
(HPA = H3PW12O40, H3PMo12O40, H4SiW12O40, H4SiMo12O40,
H4PVMo11O40, H5PV2Mo10O40), (Fig. 1b) shows that the (Td) is lowered with ca. 100 C. However, (Td) is not affected by the acidity
strength of HPA (H 3 PW 12 O 40 > H 3 PMO 12 O 40 > H 4 SiW 12 O 40 >
H4SiMo12O40 > H4PVMo11O40 > H5PV2Mo10O40).
The inuence of the amount of H3PW12O40 on (Td) is reported in Fig. 1c. The (Td) decreases from 280 C to 250 C when
2 wt.% of HPW are added, and in the presence of quantities
higher than 10 wt.% the decomposition occurs at 180 C. The increase of the nal weight with the loading of HPW can be explained by the fact that the residue comprises charcoal
originated from the biomass itself and tungsten oxide formed
after thermal treatment at high temperature (higher than
400 C) of HPW.
These experiments show that (Td) is independent on the presence of solid acids, while HPA have a signicant impact on (Td)
with an optimum for the weight ratio wood/HPA equal to 90/10.
Therefore, further studies are dedicated to the thermolysis of the
pine wood in the presence of mainly phosphotungstic heteropolyacid H3PW12O40, hereafter noted as HPW.
The isothermal TGA experiments, at 150, 200 and 250 C carried out on wood and wood/HPW (90/10) show a rapid weight
loss of ca. 6%, related to the removal of physisorbed water
(Fig. S.4). At 150 C, only the weight of wood/HPW sample decreases continuously (Fig. S.4). At 200 and 250 C, the weight of
both samples decreases but more deeply and earlier for the
wood/HPW sample. Clearly, the kinetic of the decomposition of
the wood is faster and occurs at lower temperature in the presence of HPW.

258

C. Larabi et al. / Bioresource Technology 148 (2013) 255260

Fig. 2. The amount of the CO (a) and CO2 (b) produced during the thermal
treatment of the biomass black and biomass in the presence of 10 wt.% of HPW in
gray.

during the thermal treatment of each sample (wood/HPW (90/10)

and wood) at a selected temperature (200, 250, 275 and 300 C),
the gas phase, the liquids and the solid residue have been isolated,
quantied and characterized by elemental analyses, GC, GCMS, IR,
solid state and solution NMR spectroscopies. Lastly, the possibility
to recover the HPW is essential in order to make the system more
sustainable. For this, the HPW structure has been determined at
different steps of the treatment and its recovering has been
realized.

Fig. 1. Thermogravimetric analyses of pine wood from 40 C to 700 C with heating

rate of 5 C min 1 under argon. (a) In the presence of different solid catalyst (Al2O3,
SiO2, SiO2Al2O3, ZSM-5); (b) in the presence of different HPA (H3PW12O40,
H3PMo12O40, H4SiW12O40, H4SiMo12O40, H4PVMo11O40, H5PV2Mo10- O40), (c) in the
presence of different loading of HPW (H3PW12O40) 2 wt.%, 10 wt.%, 20 wt.%, 30 wt.%,
40 wt.% and 50 wt.%.

3.2. Thermolysis of the wood and of the wood/HPW under hydrogen

Determination of the formed products after the decomposition
of pine wood in the presence of HPW is of high interest. Therefore,

3.2.1. In situ monitoring by DRIFT-GCMS coupled system

The samples wood and wood/HPW have been treated, in three
independent experiments, at 200, 250, and 300 C in the high temperature IR cell. The volatiles produced were analyzed by online
GCMS, Fig. S.5. In the absence of HPW, no products were detectable at 300 C or below. With wood/HPW, furfural was observed
at 200 C and, at 250 C, other compounds including acetic acid
and methanol were also detected, mainly originated from the
decomposition of holocellulose part of the lignocellulosic biomass
as already described (Larabi et al., 2013; Tian et al., 2010). It is reported as well that the acid hydrolysis of cellulose leads to the formation of glucose (de Vasconcelos et al., 2013; Ni et al., 2013; Tian
et al., 2010), further degradation affords furfural derivates (Hu
et al., 2013; Li et al., 2012).
To allow the identication of the chemicals obtained during the
thermolysis, 250 mg of wood/HPW were treated in the same
experimental conditions. The volatiles analyzed by GCMS

C. Larabi et al. / Bioresource Technology 148 (2013) 255260

Fig. 3. Evolution with the temperature of the weight (in % from starting wood) of
the isolated liquid, gas and solid phases during the thermal treatment of the
samples wood (a) and wood/HPW (90/10) (b).

Fig. 4. Evolution of the furfural amount produced during the thermal treatment of
the samples wood (black) and wood/HPW (gray).

(Fig. S.6) conrms that only the holocellulose moieties have been
transformed.
3.2.2. Scale up wood thermolysis in a continuous ow and batch
reactors
The quantication of the different fractions was reproduced in
higher quantities in a continuous ow reactor. The thermolysis of
wood and wood/HPW samples, at 200, 250, 275 and 300 C under
1 bar of hydrogen were run in a batch stainless autoclave reactor.
For each experiment, three fractions (gas, liquid and solid) have
been obtained and analyzed by GC, GCMS, FTIR, elemental analyses and solid state NMR. Their proportions (wt.%) are mainly
dependent on the reaction temperature.
Gas phase: With both samples the gas phase contained alkanes
and mainly CO2 and CO. With the pure wood, the light alkanes
were identied and quantied by GC. The methane was the major
component followed by ethane and traces of other light alkanes
(Cn (n = 15) 1 wt.% of the initial mass), (Fig. S.7). With the
wood/HPW sample, the amount of the same light alkanes reaches
about 2 wt.% of the initial wood mass (Fig. S.7).
The evolution of the quantities of CO and of CO2 with the temperature has been followed by FTIR, (Fig. S.8) using their characteristic stretching bands. Their integration and the comparison with
the calibration curves reported in Fig. S.1 gave the amount of
evolved CO2 and CO at different temperature, and shown in
Fig. 2. The amount s of CO and CO2 increase with the temperature
as already observed by Sagehashi et al. during the pyrolysis of Sugi,
Japanese cedar from 150 C to 400 C (Sagehashi et al., 2006). In the
presence of HPW, the amounts of CO and CO2 for the temperature
up to 260 C are higher compared to the wood sample, while,
above 270 C, it is lower, suggesting that HPW catalyst yields other
oxygenated volatiles than CO and CO2. At 300 C, the mixture
wood/HPW produced 45 mg of CO and 45 mg of CO2 per gram of
wood while the wood affords 70 mg of CO and 60 mg of CO2. In

259

the presence of HPW, the quantity of light alkanes is doubled

(Fig. S.7). The amount of CO and CO2 is higher for temperatures
lower than 260 C indicating that the reaction takes place faster
due to presence of HPW. The tendency is inversed for temperatures
higher than 260 C, suggesting that the presence of HPW leads to
the formation of other oxygenated chemical including furan
derivatives.
Liquid phase: The liquid phase collected for each experiment run
at 200, 250, 275 and 300 C for both samples was weighed and analyzed by GCMS. The weights of the liquid phase obtained for each
experiment are represented in Fig. 3. The formation of liquids occurred at 275 C for the wood and at 200 C for the sample
wood/HPW. Its amount increases with the temperature. It is observed as well that the yield of the liquids for each temperature
was superior in the presence of HPW.
In the liquid phase, furfural has been observed. Its amount, still
higher in the presence of HPW, increases with temperature and
reaches a peak at 275 C (30% of the liquid phase, i.e. 2% of the initial wood mass), Fig. 4. As already reported, the formation of furfural from the holocelluloses is favored by acid catalyst (Deng et al.,
2012). It is believed that the furfural is mainly obtained from the
decomposition of hemicelluloses (Sahu and Dhepe, 2012). The
amount of the hemicelluloses in the pine wood is ca. 15 wt.% as
determined by Van Soest method (Vansoest et al., 1991), which
also limits the maximum yield of furfural. The observed amount
of furfural is decreasing with temperature due to the polymerization of the furfural favored at higher temperature (Guerbuez
et al., 2013).
Solid residue: The solid residues collected at the end of the thermal treatments were weighed (Fig. 3) and characterized by elemental analysis, 13C and 31P solid state NMR. The amount of the
solid residue decreases with increasing temperature. Moreover,
the mass of the residue is always lower in the presence of HPW under the given conditions, indicating that HPW catalyzes the decomposition reactions.
The 13C CPMAS solid state NMR spectra of the wood treated for
2 h at different temperatures (200, 250, 275 and 300 C) are shown
in Fig. 5. The spectra of the wood treated for 2 h at 200 or 250 C
are unchanged compared to initial material (Bardet et al., 1997,
2002, 2004). The well resolved peaks at 64 and 74 ppm attributed
to different carbon of carbohydrates of the holocelluloses are

Fig. 5. 13C CPMAS solid state NMR of the solid residue recovered after thermolysis
at 200, 250, 275, 300 C; (A) without HPA, (B) in the presence of 10 wt.% of HPW.

260

C. Larabi et al. / Bioresource Technology 148 (2013) 255260

slightly affected at 275 C and decrease sharply at 300 C. At this

temperature, three broad signals, between 10 and 60 ppm (saturated carbon bonds, OC@CH carbons composing the furanic
ring), 100165 ppm (C@C carbon of aromatic rings) and
200220 ppm (attributed to carbonyl functionalities) are observed.
Additionally, the intensity of the peaks at 55 and 147 ppm assigned
to OCH3 and C@C of aromatic rings of lignin increase indicating a
higher relative amount of lignin in the sample since the holocelluloses have been converted into liquids. The evolution the spectra
of the wood/HPW samples are similar except that as early as
200 C the holocelluloses are affected as evidenced by the
decrease in the intensity of the peaks at 64 and 74 ppm, and that
at 300 C, the amount of the holocellulose present in the residue
is lower (Bardet et al., 1997, 2002, 2004).
During the thermal treatment of wood/HPW (90/10) samples at
different temperature, the 31P solid state NMR were performed on
the residue. All spectra presente a well resolved peak at 15.5 ppm
characteristic H3PW12O40 (Fig. S.9), suggesting that the HPW structure is preserved after thermolysis even at 300 C.
After extraction of the residue by water the 31P solid state NMR
(Fig. S.10) the intensity of the peak at 15.5 ppm has drastically
decreased and c.a. 60% of the initial amount H3PW12O40 is recovered and reused.
This observation is supported by the comparison of the results
of elemental analyses of all solids residues, Fig. S.11. Elemental
analyses have been obtained after Soxhlet extraction of the HPW
with water. The carbon content in the residue of the wood varies
from 49.71 wt.% at 150 C to 53.72 wt.% at 300 C i.e. an increase
of 4% with a concomitant decrease of 3.5% of the oxygen content.
For the wood/HPW (90/10) sample, after HPW extraction, the carbon content in the residue varies from 49.57 wt.% at 150 C to
58.05 wt.% at 300 C i.e. an increase of 8.48% with a concomitant
decrease of 6.8 wt.% of the oxygen content.

4. Conclusion
This work shows that the decomposition temperature of pine
wood impregnated with 10 wt.% Keggin-type heteropolyacid is
lowered by 100 C compare to pure wood or wood mixed with solid acids. At 300 C, a catalytic amount of H3PW12O40 transforms
more than 50% of the holocellulose into solid, liquid and gas phases
which have been characterized by various techniques including
elemental analysis, FTIR, GCMS and NMR. The yield of liquid is
higher in the presence of H3PW12O40 (30 wt.%, 2 wt.% furfural
versus 17 wt.% and trace of furfural for the wood). Finally, 60% of
H3PW12O40 is recovered.
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.biortech.2013.
08.070.
References
Bardet, M., Emsley, L., Vincendon, M., 1997. Two-dimensional spin-exchange solidstate NMR studies of C-13-enriched wood. Solid State Nucl. Magn. Reson. 8, 25
32.
Bardet, M., Foray, M.F., Maron, S., Goncalves, P., Tran, Q.K., 2004. Characterization of
wood components of Portuguese medieval dugout canoes with high-resolution
solid-state NMR. Carbohydr. Polym. 57, 419424.
Bardet, M., Foray, M.F., Tran, Q.K., 2002. High-resolution solid-state CPMAS NMR
study of archaeological woods. Anal. Chem. 74, 43864390.
Berndt, S., Herein, D., Zemlin, F., Beckmann, E., Weinberg, G., Schutze, J., Mestl, G.,
Schlogl, R., 1998. A new unifying structural model of heteropolymolybdate
salts: microstructure and thermal stability of a series of molecular oxides. Ber.
Bunsen-Ges. Phys. Chem. Chem. Phys. 102, 763774.

Caliman, E., Dias, J.A., Dias, S.C.L., Garcia, F.A.C., de Macedo, J.L., Almeida, L.S., 2010.
Preparation and characterization of H3PW12O40 supported on niobia.
Microporous Mesoporous Mater. 132, 103111.
Chheda, J.N., Dumesic, J.A., 2007. An overview of dehydration, aldol-condensation
and hydrogenation processes for production of liquid alkanes from biomassderived carbohydrates. Catal. Today 123, 5970.
Cui, Z.M., Li, C.M., Jiang, S.P., 2011. PtRu catalysts supported on heteropolyacid and
chitosan functionalized carbon nanotubes for methanol oxidation reaction of
fuel cells. Phys. Chem. Chem. Phys. 13, 1634916357.
de Vasconcelos, S.M., Santos, A.M.P., Rocha, G.J.M., Souto-Maior, A.M., 2013. Diluted
phosphoric acid pretreatment for production of fermentable sugars in a
sugarcane-based biorenery. Bioresour. Technol. 135, 4652.
Deng, W.P., Zhang, Q.H., Wang, Y., 2012. Polyoxometalates as efcient catalysts for
transformations of cellulose into platform chemicals. Dalton Trans. 41, 9817
9831.
Fiorese, G., Catenacci, M., Verdolini, E., Bosetti, V., 2013. Advanced biofuels: future
perspectives from an expert elicitation survey. Energy Policy 56, 293311.
Gaspar, A.R., Gamelas, J.A.F., Evtuguin, D.V., Neto, C.P., 2007. Alternatives for
lignocellulosic pulp delignication using polyoxometalates and oxygen: a
review. Green Chem. 9, 717730.
Gronli, M.G., Varhegyi, G., Di Blasi, C., 2002. Thermogravimetric analysis and
devolatilization kinetics of wood. Ind. Eng. Chem. Res. 41, 42014208.
Guerbuez, E.I., Gallo, J.M.R., Alonso, D.M., Wettstein, S.G., Lim, W.Y., Dumesic, J.A.,
2013. Conversion of hemicellulose into furfural using solid acid catalysts in
gamma-valerolactone. Angew. Chem. Int. Ed. 52, 12701274.
Guo, F., Fang, Z., Xu, C.C., Smith Jr., R.L., 2012. Solid acid mediated hydrolysis of
biomass for producing biofuels. Prog. Energy Combust. Sci. 38, 672690.
Hu, X., Wu, L.P., Wang, Y., Song, Y., Mourant, D., Gunawan, R., Gholizadeh, M., Li, C.Z.,
2013. Acid-catalyzed conversion of mono- and poly-sugars into platform
chemicals: effects of molecular structure of sugar substrate. Bioresour.
Technol. 133, 469474.
Larabi, C., al Maksoud, W., Szeto, K.C., Boyron, O., Roubaud, A., Castelli, P., Santini,
C.C., Walter, J.J., 2013. Monitoring pine wood thermolysis under hydrogen
atmosphere by in situ and ex situ techniques. J. Anal. Appl. Pyrolysis 100, 8187.
Li, G., Li, N., Wang, Z., Li, C., Wang, A., Wang, X., Cong, Y., Zhang, T., 2012. Synthesis of
high-quality diesel with furfural and 2-methylfuran from hemicellulose.
Chemsuschem 5, 19581966.
Macrelli, S., Mogensen, J., Zacchi, G., 2012. Techno-economic evaluation of 2nd
generation bioethanol production from sugar cane bagasse and leaves
integrated with the sugar-based ethanol process. Biotechnol. Biofuels 5, 18.
Marme, F., Coudurier, G., Vedrine, J.C., 1998. Acid-type catalytic properties of
heteropolyacid H3PW12O40 supported on various porous silica-based materials.
Microporous Mesoporous Mater. 22, 151163.
Menetrez, M.Y., 2012. An overview of algae biofuel production and potential
environmental impact. Environ. Sci. Technol. 46, 70737085.
Menon, V., Rao, M., 2012. Trends in bioconversion of lignocellulose: biofuels,
platform chemicals & biorenery concept. Prog. Energy Combust. Sci. 38, 522
550.
Ni, J., Wang, H., Chen, Y., She, Z., Na, H., Zhu, J., 2013. A novel facile two-step method
for producing glucose from cellulose. Bioresour. Technol. 137, 106110.
Nigam, P.S., Singh, A., 2011. Production of liquid biofuels from renewable resources.
Prog. Energy Combust. Sci. 37, 5268.
Pires, J.C.M., Alvim-Ferraz, M.C.M., Martins, F.G., Simoes, M., 2012. Carbon dioxide
capture from ue gases using microalgae: engineering aspects and biorenery
concept. Renew. Sust. Energy Rev. 16, 30433053.
Poganietz, W.R., 2012. Ligno-ethanol in competition with food-based ethanol in
Germany. Biomass Bioenergy 38, 4957.
Rao, P.M., Wolfson, A., Kababya, S., Vega, S., Landau, M.V., 2005. Immobilization of
molecular H3PW12O40 heteropolyacid catalyst in alumina-grafted silica-gel and
mesostructured SBA-15 silica matrices. J. Catal. 232, 210225.
Roman-Leshkov, Y., Barrett, C.J., Liu, Z.Y., Dumesic, J.A., 2007. Production of
dimethylfuran for liquid fuels from biomass-derived carbohydrates. Nature
447, 982985.
Sagehashi, M., Miyasaka, N., Shishido, H., Sakoda, A., 2006. Superheated steam
pyrolysis of biomass elemental components and Sugi (Japanese cedar) for fuels
and chemicals. Bioresour. Technol. 97, 12721283.
Sahu, R., Dhepe, P.L., 2012. A one-pot method for the selective conversion of
hemicellulose from crop waste into C5 sugars and furfural by using solid acid
catalysts. Chemsuschem 5, 751761.
Shimizu, K., Furukawa, H., Kobayashi, N., Itaya, Y., Satsuma, A., 2009. Effects of
Bronsted and Lewis acidities on activity and selectivity of heteropolyacid-based
catalysts for hydrolysis of cellobiose and cellulose. Green Chem. 11, 16271632.
Sitthisa, S., An, W., Resasco, D.E., 2011. Selective conversion of furfural to
methylfuran over silica-supported NiFe bimetallic catalysts. J. Catal. 284, 90
101.
Tian, J., Wang, J.H., Zhao, S., Jiang, C.Y., Zhang, X., Wang, X.H., 2010. Hydrolysis of
cellulose by the heteropoly acid H3PW12O40. Cellulose 17, 587594.
Tsigdino, Ga., Hallada, C.J., 1968. Molybdovanadophosphoric acids and their salts. I.
Investigation of methods of preparation and characterization. Inorg. Chem. 7,
437441.
Vansoest, P.J., Robertson, J.B., Lewis, B.A., 1991. Methods for dietary ber, neutral
detergent ber, and nonstrach polysaccharides in relation to animal nutrition. J.
Dairy Sci. 74, 35833597.
Wan, Y., Chen, P., Zhang, B., Yang, C., Liu, Y., Lin, X., Ruan, R., 2009. Microwaveassisted pyrolysis of biomass: catalysts to improve product selectivity. J. Anal.
Appl. Pyrolysis 86, 161167.