J. Anal. Appl.

Pyrolysis 71 (2004) 187–199

Rapid pyrolysis of wood block by

microwave heating
Masakatsu Miura a , Harumi Kaga a,∗ , Akihiko Sakurai b ,
Toyoji Kakuchi c , Kenji Takahashi c,1
a National Institute of Advanced Industrial Science and Technology (AIST), Sapporo 062-8517, Japan
b Department of Applied Chemistry and Biotechnology, Faculty of Engineering,

Fukui University, Fukui 910-8507, Japan

c Graduate School of Engineering, Hokkaido University, Sapporo 060-8628, Japan

Received 20 November 2002; accepted 5 May 2003

Abstract

Pyrolysis of cylindrical wood blocks (diameter: 60–300 mm; weight: 80–12 000 g) by microwave
irradiation has been studied using two microwave ovens (1.5 and 3 kW). The pyrolysis of wood can
be performed within ≈15 min for wood of 100 mm in diameter. It was demonstrated that temperature
distribution, heat transfer and mass transfer are quite different from a conventional heating method. The
char prepared by the microwave heating has a large specific surface area (≈450 m2 g−1 ). Scanning
electron microscopy shows that there are few pyrolysis carbon-particles in the micropores of the
char. The char yields from cylindrical wood blocks with different diameters can be correlated by an
equation, in which the electric power consumption per weight is inversely proportional to the square
of the wood block diameter. In other words, a larger wood block can be pyrolyzed with less electric
power consumption per unit weight than a smaller wood block. The yields of tar were ≈15–30 wt.%.
The tar contained levoglucosan. The concentration of levoglucosan in the tar was 5–9 wt.%.
© 2003 Elsevier B.V. All rights reserved.

Keywords: Microwave; Wood; Dielectric loss; Levoglucosan; Char; Scanning electron microscopy; Cellulose

1. Introduction

Rapid pyrolysis of wood can produce levoglucosan (1,6-anhydro-␤-d-glucopyranose),

which is a product from cellulose and chars with a large specific surface area [1,2]. In
∗ Corresponding author. Fax: +81-11-857-8988.
E-mail address: h.kaga@aist.go.jp (H. Kaga).
1 Present address: Faculty of Engineering, Kanazawa University, Kanazawa 920-8667, Japan.

0165-2370/$ – see front matter © 2003 Elsevier B.V. All rights reserved.
doi:10.1016/S0165-2370(03)00087-1
188 M. Miura et al. / J. Anal. Appl. Pyrolysis 71 (2004) 187–199

a conventional pyrolysis method, however, it is generally required to reduce the size of

the wood to perform rapid pyrolysis because the heat conduction rate of wood is quite
slow [3]. For example, in our previous study using an infrared-ray furnace [4], the wood
sample must be pulverized to increase the yield of levoglucosan. When a large wood block
is treated with a conventional pyrolysis method, we could not obtain levoglucosan with
high yields because the escape rate of volatile products from the wood block is slow. The
slow escape of volatile prevents obtaining the desired reaction products before undesired
secondary reactions occur. However, size reduction of the wood increases total costs in
wood pyrolysis and decreases process treatment rates. Therefore, an alternative technique
of wood pyrolysis is desired [3].
Microwave heating has been used for organic reactions [5,6], drying wood [7] and ex-
traction [8]; however, there have been few studies of pyrolysis of wood using microwave
[9]. We have previously demonstrated a rapid pyrolysis of a large cylindrical wood block
using microwave heating [10]. Microwaves of 2450 MHz from a domestic oven were used
to irradiate a cylindrical wood block (diameter: 60 mm; height: 60 mm). Levoglucosan, a
primary product from cellulose, was successfully produced from the large wood block. An
overall yield of levoglucosan was ≈3% based on the initial weight of the wood. The specific
surface areas of char obtained were 600–660 m2 g−1 at the central region of wood. These
values of surface area are much greater than those prepared by the conventional charcoal
furnaces. One of the advantages of microwave heating is that a larger size piece of wood can
be used and thus it is possible to reduce pretreatment processing costs, such as pulverization
of the wood. To complete the wood pyrolysis using microwaves required only 7 or 8 min.
This rapid reaction process is another advantage.
Key points in the pyrolysis of wood are to prevent undesired secondary reactions of the
volatiles and to increase the yields of some useful volatile products. Another important
feature is to produce chars that have large specific surface areas. A slow heating process
favors increasing the char yield. On the other hand, slow heating leads to decreasing the yield
of volatiles. Therefore, we have to overcome the conflicting features in wood pyrolysis [11].
As described below, microwave pyrolysis has the potential to satisfy both requirements.
A principal of microwave heating is based on ‘molecular friction’ or dielectric loss [12].
Energy transfer is not by conduction or convection as in conventional heating. When high
frequency voltages are applied to a material, the response of molecules with a permanent
dipole moment or induced dipole to the applied potential field is to change orientation in
a direction opposite that of the applied field. For example, at 2450 MHz of microwave
irradiation, the field oscillates 4.9 × 109 times per second and synchronized agitation of
molecules generates heat. The maximum energy absorption occurs at the frequency of
f=1/τ, where τ is the relaxation time of the molecule. It is difficult to define the relaxation
time of wood because wood is composed of several different materials. Therefore, the
relaxation times of wood generally have a wide distribution. For wood at room temperature,
only a small segment of the entire molecule will be responsive to the microwave fields.
Thus, wood does not absorb the microwaves strongly and the penetration depth is deep [9].
This results in uniform heating. With increasing the temperature of wood, molecules or
segment of polymers can move easily because viscosity or rigidity of the wood decreases.
In this article, we will report on our study of wood pyrolysis using microwave heating. The
wood sample was a large piece of larch, which was chosen because, in the Hokkaido area of
 M. Miura et al. / J. Anal. Appl. Pyrolysis 71 (2004) 187–199 189

Japan, there is a huge amount of larch that is not utilized as a biomass resource. Microwave
heating ovens are newly designed to perform rapid pyrolysis of a large wood block. The
characteristic of microwave pyrolysis through analysis of tar and char will be discussed.

2. Experimental

Different sizes of cylindrical blocks of larch (Lalix leptolepis GORDON) were used as
samples to examine the effect of wood size on microwave pyrolysis and to demonstrate
the practical application of microwave heating to wood pyrolysis. The diameters of wood
samples were 60, 80, 100 and about 300 mm, the heights were the same as the diameter,
and the weights were about 80, 190, 370, 12 000 g, respectively. Moisture contents in the
samples were adjusted to 8–12%.
Two microwave ovens, type A and B, were used as heating apparatuses. Type A, which was
modified from former heating apparatus [13], is a relatively small-scale oven and designed
for pyrolysis of small wood samples. Type B is a pilot scale oven that can handle larger size
wood samples. Fig. 1 outlines the microwave oven of type A. A magnetron made by Toshiba
(variable power generator (0.1–1.5 kW), oscillation frequency of 2450 MHz) was used to

Fig. 1. Microwave oven of type A. 1: Microwave power generator; 2: Isolator; 3: Output monitor; 4: Adjustment
device; 5: Reactor; 6: Condenser; 7: Collection bottle; 8: Aspirator; 9: Quartz frame.
190 M. Miura et al. / J. Anal. Appl. Pyrolysis 71 (2004) 187–199

Photo 1. Microwave oven of type B.

provide microwave heating. In order to reduce load fluctuations, an isolator to absorb reflec-
tive electric power was installed. The inside diameter and height of the reaction chamber
were 370 and ≈460 mm, respectively. The wood sample was hung in the chamber using a
frame made with quartz. The pressure in the chamber was controlled using a vacuum pump.
Photo 1 and Fig. 2 show the microwave oven of type B. The reaction chamber size
was 600 × 600 × 700 mm. Two magnetrons (oscillation frequency of 2450 MHz, variable
(0.1–1.5 kW) and fixed power generator (1.5 kW), respectively) were installed. The wood
sample was set on a rotating table. Safety shutoff equipment was installed for emergencies.
The type B reactor was used to produce quantities of char and tar.
In both microwave ovens, the reactors were purged with a slow stream of nitrogen and
evacuated with an aspirator so that volatile products were extracted from the irradiated
chamber. The extracted volatile gas from the chamber was cooled and condensed to collect
pyrolysis liquids and tar. After the pyrolysis, tar adhering to the interior wall of the reactor
was washed down into a collection bottle or tank with methanol. The methanol solution
was concentrated below 40 ◦ C by a vacuum evaporator. The viscous concentrated pyrolysis
liquid, from which moisture was removed, was regarded as tar.
The solid residue was weighed as a charred residue. The char was pulverized into particles
using a hand roller to measure specific surface area. The volatile portions and ash contents
of char particles were measured. Specific surface area and mean pore radius of the char
were measured by the BET method (QUANTASORB) in which nitrogen is absorbed at the
temperature of liquid nitrogen. Complete desorption and adsorption required several hours
to 1 day or more, depending on the samples.
 M. Miura et al. / J. Anal. Appl. Pyrolysis 71 (2004) 187–199 191

Fig. 2. Microwave oven of type B. 1: Sample; 2: Reactor (width: 600 mm; depth: 600 mm; height: 700 mm); 3:
Microwave power generator; 4: Power monitor; 5: Isolator; 6: Computer; 7: Collection tanks; 8: Pressure gauge;
9: Afterburner; 10: Aspirator pump; 11: Control panel; 12: Blower; 13: Kerosene; 14: Condenser; 15: Nitrogen
gas; 16: Exhaust gas; 17: Rotating table; 18: Safety shutoff.

Carboxylic acids passing through Millipore filters were analyzed by gas chromatography
(GC: Yanagimoto G 180; column: Unisole F 200, 2 m × 3 mm ID; column temperature:
135 ◦ C; N2 flow rate: 42.5 ml min−1 ; FID-detector). Levoglucosan was analyzed as a
trimethylsilyl derivative using gas chromatography (GC: Shimadzu GC9A; column: Shi-
madzu CBP10, 25 m × 0.32 mm ID × 0.5 ␮m; column temperature: 110–260 ◦ C; heating
rate: 7.5 ◦ C min−1 ; glass capillary column: 25 m; He; FID-detector).

3. Results and discussion

3.1. A characteristic of dielectric heating of wood

In order to understand how dielectric heating proceeds in wood and to demonstrate a

rough temperature distribution in wood during microwave irradiation, the microwave irra-
diation was stopped in the middle of pyrolysis and the wood was taken from the chamber.
Photo 2 shows a cross section of a wood block (80 mm in diameter) after 3 min of microwave
irradiation. No significant changes on the surface area were observed after the microwave
irradiation. The center of the wood, however, changed to black, indicating that the temper-
ature at the wood core was higher than that for the surface. It is clear that carbonization
developed from the center to the surface layer in the wood block. This demonstrates that
early primary volatiles of wood will transfer from the center to the surface layer. There-
fore, it will be possible to reduce undesired secondary reactions of the volatiles because the
volatiles produced in the wood core will diffuse towards the lower temperature region.
192 M. Miura et al. / J. Anal. Appl. Pyrolysis 71 (2004) 187–199

Photo 2. Cross-section of wood after 3 minutes of microwave irradiation.

From the above result, we can discuss a difference between conventional and microwave
pyrolysis. Fig. 3 illustrates a schematic diagram of temperature distribution, heat transfer
and mass transfer in the conventional and microwave heating of wood. In the conventional
method, wood is heated by convective heat transfer from high-temperature gas and by

Fig. 3. Schematic diagram of temperature distribution, heat transfer and mass transfer in the conventional and
microwave heating of wood.
 M. Miura et al. / J. Anal. Appl. Pyrolysis 71 (2004) 187–199 193

Photo 3. Scanning electron microscopy (SEM) of char: (a) prepared by the conventional heating method; (b)
prepared by the microwave heating method using the type A reactor.

conduction from the surface to the wood core. The temperature at the wood surface is higher
than that at the wood core because of the poor thermal conductivity of wood. As the pyrolysis
develops toward a direction from the surface to the center, the primary volatiles produced
near the center must transfer through the high temperature surface region. Therefore, it is
difficult to prevent undesired secondary reactions, which will produce low molecular weight
species. Secondary reactions will also produce polymerization of molecules.
However, in the dielectric heating method, first the wood core attains a higher temperature
as shown in Photo 2. Recent studies for microwave heating also demonstrated this unique
temperature distribution [14,15]. The radial distribution of temperature in the wood is reverse
to that of the conventional heating methods; i.e., higher temperature in the center and lower
temperature in the outer region. As the volatile gas produced at the wood core transfers
through the lower-temperature section, undesired secondary reactions are reduced.

3.2. Char characteristics

Photo 3 shows scanning electron microscopy (SEM) of the charred residue. The char
in photo 3(a) was produced by conventional heating. Photo 3(b) is the char resulting from
194 M. Miura et al. / J. Anal. Appl. Pyrolysis 71 (2004) 187–199

dielectric heating. Noticeable differences between them are their structure and carbon-like
adhesives (small pyrolysis carbon-particles deposited in the micropore). For the char pre-
pared by conventional heating, the micropores are filled with large amounts of the carbon-like
adhesives. As discussed above, in conventional heating, the volatile material produced at
the central region of wood sample will transfer toward the outer layer. The escape rate of
volatile products from the wood block is not fast enough to prevent thermal decomposition
of the volatile products because of the slow heat-transfer rate. Therefore, during pyrolysis,
pyrolysis carbon, as a result of cracking, will deposit on the pores.
On the other hand, for the char prepared by microwave heating, the micropores are quite
clean and exhibit a more open structure as a result of the release of the volatile matter.
There are few carbon-like adhesives in the micropore. As an adsorbent, a large specific
surface area is one of the important characteristics of char. We already reported that the
specific surface areas of char produced by microwave heating are larger than that produced
by conventional heating [10].
Fig. 4 shows the relationships of the specific surface area and the mean pore radius of char
as a function of microwave irradiation time. The results were obtained using the microwave
oven of type A. The sample size was 100 mm in diameter and height. As irradiation time
increased, the specific surface area increased and the mean pore radius decreased. After ≈12
min of irradiation, no appreciable changes were observed in both the specific surface area
and mean pore radius. These results demonstrate that, after sufficient irradiation, microwave
has no significant effect on the char products. When the yields of char were ≈20% or lower,
the specific surface area of the char was ≈450 m2 g−1 . The surface area was ≈2.5 times that
of the char produced by our conventional charcoal furnace (184 m2 g−1 ) [10]. The mean
pore radius was ≈10 Å after 12 min irradiation.
Fig. 5 demonstrates the char yield as a function of irradiation time using the type A
and B reactors. The carbonization rate is slow until the evaporation of water in the wood
has been completed. After evaporation, the carbonization proceeds rapidly. For example,

Fig. 4. Relationships of specific surface area and mean pore radius of char as a function of irradiation time. The
results were obtained using the type A reactor. 䊉, Specific surface area; , mean pore radius.
 M. Miura et al. / J. Anal. Appl. Pyrolysis 71 (2004) 187–199 195

Fig. 5. Changes of char yield as a function of irradiation time.

in the 80 mm wood sample by the type A reactor, the carbonization proceeds within 10
min of irradiation and there are no significant changes in the yields despite continued
microwave irradiation. This indicates that, when the char yield reaches ≈20%, the forma-
tion of volatile substances has stopped and carbonization has ceased. While the pyrolysis
can be carried out within ≈15 min for the small wood blocks (60 and 80 mm) using the
type B reactor, ≈80 min were required for the large wood block (300 mm) to complete
the pyrolysis. For the smallest wood sample examined (60 mm), the char yields for the
type B microwave oven was greater than that for the type A reactor. Also, the rate of car-
bonization for the type B reactor was slower than that for the type A reactor. This is due
to a difference in the sample holding method in the reaction chamber. In the type A re-
actor, the wood was hung by a quartz frame. Therefore, there was no heat loss by heat
conduction. On the other hand, in the type B reactor, the wood was set on a rotating ta-
ble made with stainless steel, which allowed heat to escape from the wood to the table
by heat conduction. Therefore, the pyrolysis was not completed near the bottom of the
wood.
Fig. 6 shows the relationships between the yield of char and electric power consumption
for four different diameters of samples using the type A and B reactors. Regardless of the
diameter, the yield decreased linearly until ≈20%. Fig. 6 also demonstrates that larger wood
samples require smaller electric power consumption per weight for carbonization. This is
a significant advantage for pyrolysis of a large size piece of wood. We found that, for the
type A reactor, the char yields Yc can be correlated by the following equation within ±10%
errors by considering the wood surface area and power consumption.

ln Y c = 5 − 200d 2 P/W

where P is the power consumption (kWh), W, the weight of wood sample (kg), Yc is the
yield of char (%) and d, the diameter of wood (m). This correlation means that a larger
196 M. Miura et al. / J. Anal. Appl. Pyrolysis 71 (2004) 187–199

Fig. 6. Relationships between the yields of char and electric power consumption.

wood block can be pyrolyzed with less electric power consumption per unit weight than a
smaller wood block. This is a significant advantage for pyrolysis of a large size wood block.

3.3. Tar characteristics

When the microwave irradiation subsequently continued, white smoke from the wood
was observed. The volatile gas was condensed and accumulated in the cooling pipe receiver.
The yields of tar were examined as a function of irradiation time. Table 1 shows the yield of

Table 1
Yields of tar, char and levoglucosan using the type A reactor
d W t V A Pc Yt Yc LG YLG

80 0.186 3 140 10.4 0.39 0 90.2 0 0

80 0.186 6 140 10.4 0.78 10.8 49.7 5.3 0.6
80 0.189 9 140 10.5 1.17 19.6 32.8 5.1 1.0
80 0.187 12 140 10.5 1.57 24.0 22.1 5.0 1.2
80 0.190 15 140 10.5 1.93 22.6 20.9 7.4 1.6
80 0.191 18 140 10.5 2.31 18.1 21.1 5.3 1.0
100 0.351 7 190 12.0 0.76 26.4 33.6 8.1 2.1
100 0.368 9 190 11.9 0.92 16.5 25.8 8.2 1.4
100 0.388 10 190 11.7 0.96 24.3 22.8 8.2 2.0
100 0.348 11 150 12.5 0.99 25.9 20.6 8.8 2.3
100 0.371 11 190 11.8 1.11 31.5 19.6 5.2 1.9
100 0.377 12.5 190 11.7 1.23 29.7 18.2 6.7 2.0
d, Diameter of wood (mm); W, weight of wood (kg); t, microwave irradiation time (min); V, voltage (V); A, current
(A); Pc, power consumption (kWh kg−1 ); Yt, tar yield (wt.%); Yc, charred residue yield (wt.%); LG, levoglucosan
yield based on tar weight (wt.%); YLG, levoglucosan yield based on wood weight (wt.%).
 M. Miura et al. / J. Anal. Appl. Pyrolysis 71 (2004) 187–199 197

tar and levoglucosan using the microwave oven of type A. Also included are experimental
conditions and the yield of char. With increasing the microwave irradiation time, the charred
residue yields decreased and the yield of volatiles (or tars) increased. For example, when a
wood sample of 80 mm diameter were irradiated with 140 V, ≈10% of the weight is lost
in 3 min, which almost corresponds to the weight of water in the sample. After irradiation
for 15 min, char with a yield of 20% was obtained. The char yield decreased monotonically
with the irradiation time; however, the time profile of the tar shows a different behavior. A
maximum yield of tar was obtained after 12 min of irradiation. Further irradiation decreased
the yield of tar.
The tar composition was analyzed using gas chromatography. The tar includes car-
boxylic acids (acetic acid, propinic acid, n-butyric acid, iso-butyric acid, n-valeric acid,
iso-valeric acid, n-caproic acid), furfural, o-cresol, guaiacol, eugenol, l-(+)-arabinose,
d-(−)-arabinose, ribose, xylitol, levoglucosan, mannnosan (1,6-anhydro-␤-d-mannopyra-
nose), galactosan (1,6-anhydro-␤-d-galactopyranose), d-(+)-glucose. We also examined
the relationship between the yield of acetic acids and electric power consumption. Acetic
acid at 1.2–2.4 wt.% per initial weight of wood are generated.
The tar sample contained levoglucosan. Levoglucosan is an important starting material
for the synthesis of stereoregular polysaccharides possessing biological activities, such as
anti-human immunodeficiency virus and blood anticoagulant activities [16]. The deriva-
tives of stereoregular polysaccharides can be used for the chiral stationary phase in high-
performance liquid chromatography [17–19]. The concentration of levoglucosan in the tar
was ≈5–9 wt.% and the yields based on the initial weight of wood was 0.5–2.3 wt.%. In our
previous study, in which the pyrolysis of pure cellulose by dielectric heating was performed
using the prior type A reactor [13], levoglucosan with higher yields (12 wt.% from filter
papers (200 g), 6 wt.% from used papers (200 g)) was obtained than that from the present
study.
The main components of the larch are ␣-cellulose (40%), lignin (31%) and pentosan
(12%) [20]. If all of the ␣-cellulose in the larch were converted to levoglucosan and as-
suming 12% yield (from our previous work [13]) of levoglucosan from the cellulose, the
levoglucosan yield of 4.8% based on the wood weight can be expected. In our present study,
the maximum yield of levoglucosan is 2.3 wt.%. This value is close to the ideal yield in
our apparatus. The difference between the ideal value and the present yield is due to sev-
eral reasons, such as imperfect collection of the volatile gas. Levoglucosan is an anhydride
of glucose, which can be considered to result from the loss of water from glucose with
the simultaneous formation of the 1,6-oxygen bridge. It is also known that levoglucosan
polymerizes into dimers, tetramers and hexamers under the influence of heat and forms
crystalline ethers, as well as crystalline triacetates and tribenzoates [21]. Therefore, lev-
oglucosan in the volatile gas that is not collected quickly from the reaction chamber could
be lost by the above fates.
Lakshmanan et al. [22] obtained ≈40% levoglucosan from starch by conventional heating.
They also tried the dielectric heating method and found that the yields of levoglucosan were
low compared to the yield obtained by a conventional heating method. They attributed the
reason for low yield to decomposition and polymerization of levoglucosan due to a glow
discharge in the reactor. However, in an experiment of cellulose pyrolysis by Allan et al.
[23], it was possible to obtain levoglucosan with high yields (≈40%) using microwave
198 M. Miura et al. / J. Anal. Appl. Pyrolysis 71 (2004) 187–199

heating. In their experiment, the cellulose sample was set in a quartz reaction tube (2.54
cm in inside diameter) and flowing helium gas swept out reaction products. Although their
yield of levoglucosan is higher than our previous work [13], the amount of levoglucosan
in their study is significantly less because they studied the pyrolysis on an analytical-scale
using a small amount of sample (cellulose pellet, 1.27 cm in diameter and 2.6 g). These
results suggest that, to increase the yield of levoglucosan, an effective recovery/transfer
method of volatile gas from the reaction zone is an important factor. Our low yields of
levoglucosan in the cellulose pyrolysis [13] compared to the Allan study are presumably
due to a large sample size and long residence time of the volatiles in the reaction chamber.
In the present experimental conditions (N2 flow rate of 5 l min−1 ), the residence time of
volatile gases could be >10 min for the type A reactor. The long residence time will increase
the chance for adhesive of the volatile products on the reactor wall. After optimization of
process conditions, it could be possible to increase the yield of levoglucosan.

4. Conclusion

The pyrolysis of wood blocks by microwave irradiation using two microwave ovens was
studied. It was possible to perform pyrolysis of the wood block in a few minutes. The
dielectric heating method has a different heat transfer mechanism from the conventional
methods and can prevent undesired pyrolysis of products. The results from scanning electron
microscopy of char demonstrated that the amounts of carbon-like adhesives on the inside
of pores are less than those in the conventional methods. The specific surface area of char
is larger than that produced by conventional heating. The char yield from cylindrical wood
blocks with different diameters can be correlated by an equation, in which the electric
power consumption per weight is inversely proportional to the square of the diameter. In
other words, a larger wood block can be pyrolyzed with less electric power consumption
per unit weight than a smaller wood block. The tar contained levoglucosan with a yield of
5–8.8 wt.%.

References

[1] A.V. Bridgwater, J. Anal. Appl. Pyro. 51 (1999) 3.

[2] A.V. Bridgwater, G.V.C. Peacoke, Renew Sustain Ene. Rev. 4 (2000) 1.
[3] D. Meier, O. Faix, Biores. Tech. 68 (1999) 71.
[4] M. Miura, S. Tanaka, K. Ando, Kagaku Kogaku Ronbunshu 21 (1995) 843.
[5] S. Caddick, Tetrahedron 51 (1995) 10403.
[6] M. Larhed, C. Moberg, A. Hallberg, Acc. Chem. Res. 35 (2002) 717.
[7] A. Oloyede, P. Groombridge, J. Mate, Proc. Tech. 100 (2000) 67.
[8] C.S. Eskilsson, E. Björklund, J. Chromatogr. A 902 (2000) 227.
[9] B.B. Krieger, Res. Chem. Intermed. 20 (1994) 39.
[10] M. Miura, H. Kaga, S. Tanaka, K. Takahashi, K. Ando, J. Chem. Eng. Jpn. 33 (2000) 299.
[11] E. Sjöström, Wood Chemistry 2nd ed., Academic Press, San Diego, 1993, p. 234.
[12] D.A. Copson, Microwave Heating 2nd ed., The AVI Pub, Connecticut, 1975, p. 12.
[13] M. Miura, H. Kaga, T. Yoshida, K. Ando, J. Wood Sci. 47 (2001) 502.
[14] H. Zhao, I.W. Turner, Appl. Math. Model 24 (2000) 183.
 M. Miura et al. / J. Anal. Appl. Pyrolysis 71 (2004) 187–199 199

[15] P. Ratanadecho, K. Aoki, M. Akahori, Appl. Math. Model 26 (2002) 449.

[16] K. Hattori, T. Yoshida, H. Nakashima, M. Premanathan, R. Aragaki, T. Mimura, Y. Kaneko, N. Yamamoto,
T. Uryu, Carbohydr. Res. 312 (1998) 1.
[17] A. Kusuno, M. Mori, T. Satoh, M. Miura, H. Kaga, T. Kakuchi, CHIRALITY 14 (2002) 498.
[18] A. Kusuno, M. Mori, M. Miura, H. Kaga, T. Kakuchi, Enantiomer 6 (2001) 27.
[19] T. Kakuchi, A. Kusuno, M. Miura, H. Kaga, Macromol. Rapid Commun. 21 (2000) 1003.
[20] R.C. Petterson, in: R. Rowell (Ed.), The Chemistry of Solid Wood, Advances in Chemistry Series 207,
American Chemical Society, Washington, DC, 1984, p. 57.
[21] J.D.S. Carvalho, W. Prins, C. Schuerch, J. Am. Chem. Soc. 81 (1966) 16122.
[22] C.M. Lakshmanan, B. Gal-Or, H.E. Hoelscher, I & EC Pro. Res. Dev. 8 (1969) 261.
[23] G.G. Allan, B.B. Krieger, D.W. Work, J. App. Poly. Sci. 25 (1980) 1839.