You are on page 1of 5

Bioresource Technology 107 (2012) 451455

Contents lists available at SciVerse ScienceDirect

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

Determination of effective moisture diffusivity and drying kinetics for poplar


sawdust by thermogravimetric analysis under isothermal condition
Dengyu Chen, Yan Zheng, Xifeng Zhu
Key Laboratory for Biomass Clean Energy of Anhui Province, University of Science and Technology of China, Hefei 230026, China

a r t i c l e

i n f o

Article history:
Received 31 August 2011
Received in revised form 7 December 2011
Accepted 8 December 2011
Available online 14 December 2011
Keywords:
Biomass
Poplar sawdust
Effective moisture diffusivity
Drying kinetics
Thermogravimetric analysis

a b s t r a c t
The current study presents a thermogravimetric method to determine the effective moisture diffusivity
and drying kinetics of biomass. Drying experiments on poplar sawdust were performed at four temperatures (60, 70, 80, and 90 C) by a thermogravimetric analyzer (TGA). The major assumption in experimentally determining effective diffusivity by Ficks diffusion equation is that drying is mass transfer
limited and temperature remains isothermal during drying. The results indicated that TGA could well
achieve these determining conditions. The drying process of sawdust mostly took place in the falling rate
period. MidilliKucuk model showed the best t for all experimental data. The effective diffusivity values
changed from 9.38  1010 m2/s to 1.38  109 m2/s within the given temperature range, and the activation energy was calculated to be 12.3 kJ/mol.
2011 Elsevier Ltd. All rights reserved.

1. Introduction
Biomass utilization is gaining increasing attention as a potential
source of renewable energy (Sims et al., 2010). Thermochemical
conversions (pyrolysis, gasication, and combustion) are the major
biomass utilization techniques for gas, liquid, and char production.
Unfortunately, biomass materials often contain considerable
amount of water (Chen et al., 2009). The presence of water has
many negative effects on the performance of biomass and the
development of conversion technology (Cai and Chen, 2008).
Therefore, drying pretreatment of materials is essential to improve
the efciency of biomass utilization.
Drying is one of the most common methods used to improve
raw biomass (Fagerns et al., 2010). Ficks second law of diffusion
has been widely used to describe the drying process for most biological materials (Doymaz, 2010). Effective moisture diffusivity
(Deff) is the key drying parameter that represents the conductive
term of all moisture transfer mechanisms (Srikiatden and Roberts,
2006). This parameter is usually determined from experimental
drying curves. The activation energy is another important drying
parameter that represents the energy level of water molecules
for moisture diffusion and evaporation. The temperature dependence of the effective moisture diffusivity can be represented by
an Arrhenius relationship. Thus, the activation energy can be determined from the slope of the Arrhenius plot, ln(Deff) vs. 1/T (Shi

Corresponding author. Tel./fax: +86 551 3600040.


E-mail address: xfzhu@ustc.edu.cn (X. Zhu).
0960-8524/$ - see front matter 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.biortech.2011.12.032

et al., 2008). In particular, the major assumption in experimentally


determining Deff is that drying is mass transfer limited, that is, the
temperature should remain isothermally throughout the whole
sample during the entire course of drying (Roberts and Tong,
2003). In addition, the temperature used in the Arrhenius analysis
to determine the activation energy is the ambient temperature of
drying, that is, the temperature of the sample should be similar
to that of the surrounding drying environment (Vega-Glvez
et al., 2010). Therefore, the isothermal assumption has also been
applied in determining the activation energy.
However, achieving isothermal drying conditions is very difcult and sometimes impossible, because the control of this variable
is a challenging problem that demands the evaluation of many
interconnected non-linear mass and heat transport phenomena
(Di Scala and Crapiste, 2008). Many researchers have pointed out
that temperature gradients in the sample during drying have a signicant effect on proper determination of the diffusion coefcient.
When temperature gradients are present, the prediction of moisture loss fails to follow experimental drying curves (Vaccarezza
et al., 1974). Srikiatden and Roberts (2006) found that the discrepancies between predictions and experimental data resulted from
the temperature gradients in the sample (Srikiatden and Roberts,
2006). Nevertheless, when they used an isothermal drying apparatus, the temperature-dependent diffusion coefcient was correctly
determined, and Ficks diffusion model accurately predicted moisture loss during isothermal drying. In addition, many researchers
have reported another phenomenon in which no thermal gradients
between the surface and the center of biomass are found during
drying; however, the sample temperature varied with drying time.

452

D.Y. Chen et al. / Bioresource Technology 107 (2012) 451455

Alzamora et al. found that only 10% of the total moisture remains
when the sample temperature is within 5 C of the dry-bulb
temperature (Alzamora et al., 1979). A similar behavior was also
observed by Rovedo et al. (1995). The temperature of the potato
samples in their study did not reach the drying temperature until
95% of the potatoes moisture was removed.
Therefore, without a specic isothermal condition established,
there is evident difculty in determining effective diffusivity and
activation energy. However, the application of the thermogravimetric method in the drying process may prevent this problem. Thermogravimetric analysis, which rapidly measures a materials weight
loss over a wide temperature range or over a long period at a specic
temperature, has proven to be one of the best techniques for the
experimental research and kinetic analysis of biomass (Damartzis
et al., 2011). The application of thermogravimetric analyzer (TGA)
for biomass drying has many advantages, such as the immediate
heating of biomass to the set temperature, online recording of the
drying process, and minimal material requirement (Gardelle et al.,
2011). Most importantly, TGA could effectively restrict the temperature gradients between inside and outside the biomass particles.
Moreover, drying kinetics plays an importance role in the design
of a dryer. Some researchers have applied TGA to the mathematical
modeling of mass transfer, which is a very useful tool in investigating the intrinsic kinetics of a drying process (Madhava et al., 2001;
Cai and Liu, 2007). Thus, with the immediate establishment of
isothermal conditions and precise temperature control, as well as
online sample weight measurements, the thermogravimetric method is probably suitable for isothermal drying studies. However, no
reports have been found regarding the use of thermogravimetric
method to investigate moisture diffusivity mechanism and to determine the drying kinetics of biomass.
The objectives of the current work are to present the thermogravimetric method for calculation of Deff and activation energy
of poplar sawdust, and to determine a mathematical model that
suitably describes the drying characteristics.

2. Methods
2.1. Raw material
The biomass material used in the present study was poplar sawdust, which was selected from a local furniture factory. The particles chosen for the drying experiments were 0.1250.3 mm in size.
The initial moisture content of poplar sawdust was determined by
means of AOAC method No. 934.06 (AOAC, 1990). Sawdust at high
temperatures (>200 C) undergoes pyrolysis reactions with signicantly morphological changes in color and volume. Whereas, in the
drying temperature range (<150 C), these changes are negligible.

2.2. Drying experiments


Drying experiments were performed at four temperatures (60,
70, 80, and 90 C) using a thermogravimetric analyzer (TGA,
Q5000IR, TA Instruments, New Castle, USA). The TGA, with its new
infrared furnace, provided a high heating rate of 500 C/min and
precise temperature control. Poplar sawdust in the sample pan
(platinum) was maintained at 5 mg for every experiment, and the
ow rate of carrier gas (nitrogen) was kept at 100 mL/min. At the
selected conditions, the limitation of heat transfer inside the particles and in the external gas phase was negligible (Yang et al.,
2006). A computer connected to the thermogravimetric analyzer
automatically recorded the changes in weight, and then processed
the data. All experiments were replicated three times at each
temperature and averages of weight loss were used.

2.3. Mathematical modeling of drying curves


The moisture ratio (MR) of poplar sawdust was calculated using
the following equation (Vega-Glvez et al., 2011):

MR

M  Me
M0  Me

where M is the moisture content at any time, M0 is the initial moisture content, and Me is the equilibrium moisture content of the
sample. Me values were obtained from the drying curves and were
set to be equal to moisture content, at which sample weight became
constant with drying time. The drying curves were tted with ve
common mathematical models listed in Table 1, namely, Page,
Newton, MidilliKucuk, Logarithmic, HendersonPabis (Midilli
et al., 2002; Midilli and Kucuk, 2003).
2.4. Statistical evaluation of drying data
The model parameters determined using non-linear regression
analysis. The goodness of t of the tested mathematical models
to the experimental data was evaluated from the coefcient of
determination (R2) and the chi-square (v2) between the predicted
and experimental values. The higher R2 values and the lower v2
values, the better is the goodness of t (Miranda et al., 2010). v2
can be calculated as follows:

v2

PN

 MRexp;i 2
Nn

i1 MRpre;i

where MRexp,i and MRpre,i are experimental and predicted moisture


ratios, respectively, N is number of observations, and n is number of
drying constants.
2.5. Determination of the effective diffusivity
The drying process of biological materials mostly occurs in the
falling rate period. Ficks second law of diffusion, shown in Eq.
(3), has been widely used to describe the drying process and interpret the experimental drying data during the falling rate period
since internal mass transfer controls the drying process (Crank,
1975). The mathematical solution of Ficks second law for diffusion
is shown in Eq. (4):

@MR
rDeff rMR
@t
MR

1
8 X

p2

n0

1
2n 12

exp 

2n 12 p2 Deff t

!
4

4L2

where n is a positive integer, t is the drying time (s) and L is the half
thickness of the sample (m).
When sample shrinkage is negligible, initial moisture distribution is uniform, and constant moisture diffusivity is assumed, Eq.
(4) is suitable for determining Deff. This equation could be further
simplied into Eq. (5) by taking the rst term of a series solution
as follows (Vega-Glvez et al., 2010):

Table 1
Five commonly used models for isothermal drying.
No.

Model

Equation

Parameters

1
2
3
4
5

Page
Newton
MidilliKucuk
Logarithmic
HendersonPabis

MR = exp(ktn)
MR = exp(kt)
MR = a exp(ktn) + bt
MR = a + b exp(kt)
MR = a exp(kt)

k,
k
a,
a,
a,

n
k, n, b
k, b
k

D.Y. Chen et al. / Bioresource Technology 107 (2012) 451455

lnMR ln

p2 Deff
4L2

453


5

Effective diffusivities are typically determined by plotting


experimental drying data in terms of ln(MR) vs. time. From Eq.
(5), a plot of ln(MR) vs. time gives a straight line with a slope
(k0) of:

k0

p2 Deff

4L2

In addition, the effect of temperature on effective diffusivity can


be described by an Arrhenius-type equation (Vega-Glvez et al.,
2009):


Deff D0 exp 


Ea
RT 273:15

7
Fig. 2. Drying curves of poplar sawdust at different drying temperatures.
2

where D0 is the pre-exponential factor of the Arrhenius equation (m /


s), Ea is the activation energy for the moisture diffusion (kJ/mol), R is
the ideal gas constant (J/mol K), and T is the drying temperature (C).
The activation energy was calculated by plotting ln(Deff) vs. the
reciprocal of the temperature [1/(T + 273.15)].

3. Results and discussion


3.1. Temperature proles
TGA records temperature data of the sample pan (platinum) online during drying. The results of the temperature proles are
shown in Fig. 1. It can be observed that the temperature of sample
pan reached the desired temperature within a very short time and
was maintained at this temperature throughout drying. Owing to
the precise temperature control capability of TGA and the excellent
thermal conductivity of the sample pan, the temperature gradient
between sawdust and sample pan was negligible. Moreover, owing
to the low heat capacity of sawdust and the small amount of sawdust in the sample pan, as well as the relatively low drying temperature, sawdust moisture evaporation had a very little effect on
ambient temperature. Hence, the temperature gradient between
the surface and the center of sawdust was also negligible. Therefore, the temperature prole of the sample pan shown in Fig. 1
was also that of sawdust in each experiment. Proper determination
of Deff is conducted as the isothermal condition is established.

Fig. 1. Temperature proles of the sample pan in TGA at different drying


conditions.

3.2. Moisture transfer characteristics


The results of the drying experiments are shown in Figs. 2 and 3,
which show the drying curves and the drying rate curves, respectively. The drying temperature signicantly affected the moisture
change in sawdust. Increasing the drying temperature in the tested
range increased the drying rate of sawdust and reduced the time
required to achieve specic moisture content. The initial moisture
content of poplar sawdust was determined as 12.5% on wet basis.
Temperature is the main driving force of drying. Therefore, hightemperature drying can remove more moisture during drying.
The nal moisture content was 5.8% for drying at 60 C, whereas
it was only 0.9% for drying at 90 C.
From Figs. 2 and 3, there was no constant rate drying period in
all drying conditions, probably because the thin layer of sawdust
could not provide a constant supply of moisture during the drying
period. This phenomenon was in agreement with the drying characteristics of many bioproducts (Doymaz, 2007). However, a short
rising rate period was evident for all drying conditions. In particular, this rising rate period precisely corresponded to the increasing
temperature period of sawdust in TGA. Therefore, the short period
of the rising rate could be attributed to the increasing temperature
of sawdust, which directly improved moisture evaporation on its
surface. As the drying process continued, free water escaped in
the form of vapor. With insufcient free moisture on the sawdust

Fig. 3. Drying rate curves of poplar sawdust at different drying temperatures.

454

D.Y. Chen et al. / Bioresource Technology 107 (2012) 451455

surface, the drying rate was then dominated by the moisture diffusion from the inside to the surface of sawdust. Therefore, the drying rate started to decrease and the falling rate period begun,
similar to the drying characteristics of many agricultural products
(Kaya and Aydin, 2009). With less water in the sawdust, the drying
rate gradually decreased to zero. The time needed for the two drying periods depended on the energy supplied. In the present study,
the time for the rising rate period was signicantly shorter than
that for the falling rate period. Most of the water was removed in
the falling rate period, which made drying a time-consuming
process.
Thus, most of the drying process clearly took place during the
falling rate period, and internal diffusion was the dominant physical mechanism governing moisture transfer in the samples. Other
researchers have reported similar results (Shi et al., 2008).
3.3. Calculation of effective diffusivity and activation energy
Fig. 4. Correlation between Deff and drying temperatures.

The results obtained have shown that the isothermal condition


was established immediately and maintained throughout drying.
Therefore, with negligible sample shrinkage and uniform initial
moisture distribution as well as constant moisture diffusivity
assumptions, the mathematical solution of Ficks second law for diffusion (Eq. (4)) was suitable for determining the Deff. The calculated
values of Deff for sawdust drying were 9.38  1010, 1.09  109,
1.17  109, and 1.38  109 m2/s at 60, 70, 80, and 90 C, respectively. The values of Deff increased greatly with increasing temperature and were in the suitable general range of 1011109 m2/s for
biomass. The activation energy was calculated using Eq. (7). The
results are presented in Fig. 4. The gure shows a linear relationship
duo to the Arrhenius-type dependence (R2 = 0.9651). From the slope
of this line, an activation energy value of 12.3 kJ/mol was determined for sawdust drying.
3.4. Modeling of drying kinetics

Fig. 5. Experimental and MidilliKucuk model predicted values of moisture ratio


for poplar sawdust at different drying temperatures.

The drying data obtained by TGA were tted using the ve drying models listed in Table 1. The best model describing the drying
kinetics of sawdust was chosen as the one with the highest R2 and
the least v2. Table 2 lists the statistical analysis results for the

Table 2
Statistical results obtained from all drying models.
Model

Temperature (C)

MidilliKucuk

60
70
80
90

Page

Model parameters

R2

v2

1.03
1.04
1.04
1.00

1.40  106
2.38  105
2.74  105
2.23  105

1.63  103
3.44  103
4.90  103
4.32  103

1.28
1.21
1.18
1.29

0.9992
0.9984
0.9972
0.9962

7.62  105
1.27  104
1.89  104
2.15  104

60
70
80
90

1.09  103
2.54  103
3.75  103
4.45  103

1.35
1.26
1.23
1.28

0.9985
0.9970
0.9956
0.9955

1.33  104
2.31  104
2.97  104
2.52  104

Newton

60
70
80
90

6.53  103
9.01  103
1.08  102
1.48  102

0.9733
0.9841
0.9866
0.9847

2.43  103
1.23  103
9.13  104
8.53  104

Logarithmic

60
70
80
90

6.32  102
1.05  102
2.12  104
5.21  105

6.35  103
9.79  103
1.20  102
1.64  102

0.9962
0.9960
0.9953
0.9920

3.45  104
3.11  104
3.21  104
4.46  104

HendersonPabis

60
70
80
90

1.15
1.13
1.12
1.11

7.47  103
1.01  102
1.21  102
1.64  102

0.9917
0.9956
0.9953
0.9920

7.61  104
3.40  104
3.21  104
4.46  104

1.18
1.13
1.12
1.11

D.Y. Chen et al. / Bioresource Technology 107 (2012) 451455

models. It can be seen that Page, Logarithmic and MidilliKucuk


models presented very good ts. However, MidilliKucuk model
contributed the best experimental value of tting with the highest
R2 and least v2. These results were in agreement with some other
similar works (Doymaz et al., 2006; Raee et al., 2009). Fig. 5
shows the tting results of MidilliKucuk model for all drying
experiments. These results clearly demonstrate that the Midilli
Kucuk model ts the drying kinetics of sawdust perfectly.
4. Conclusions
The drying characteristics of poplar sawdust were studied in a
thermogravimetric analyzer. Internal diffusion is the dominant
physical mechanism governing moisture transfer. With the immediate establishment of isothermal conditions and precise temperature control capability, as well as online sample weight
measurements, the thermogravimetric method is suitable for
investigating moisture diffusivity mechanism and for determining
drying kinetics. The values of effective moisture diffusivity ranged
from 9.38  1010 to 1.38  109 m2/s. The activation energy was
calculated to be 12.3 kJ/mol. The experimental data were tted to
ve drying kinetics models. MidilliKucuk model best described
the drying characteristics of sawdust.
Acknowledgements
The authors thank National Natural Science Foundation of China (50930006, 50876099), National Basic Research Program of China (2007CB210203), National Key Technology R&D Program
(2011BAD22B07) and The CAS Special Grant for Postgraduate Research, Innovation and Practice for nancial support.
References
Alzamora, S.M., Chirife, J., Voillaz, P., 1979. A simplied model for predicting the
temperatures of foods during air dehydration. J. Food Technol. 14, 369380.
AOAC, 1990. Ofcial Method of Anlisis, Association of Ofcial analytical Chemists
No. 934.06, 15th ed. Arlington, MA, Washington.
Cai, J.M., Chen, S.Y., 2008. Determination of drying kinetics for biomass by
thermogravimetric analysis under nonisothermal condition. Drying Technol.
26, 14641468.
Cai, J.M., Liu, R.H., 2007. Research on water evaporation in the process of biomass
pyrolysis. Energy Fuels 21, 36953697.
Chen, L.J., Li, X., Han, L.J., 2009. Renewable energy from agro-residues in China: solid
biofuels and biomass briquetting technology. Renewable Sustainable Energy
Rev. 13, 26892695.
Crank, J., 1975. The Mathematics of Diffusion, second ed. Oxford University Press,
London, UK.

455

Damartzis, T., Vamvuka, D., Sfakiotakis, S., Zabaniotou, A., 2011. Thermal
degradation studies and kinetic modeling of cardoon (Cynara cardunculus)
pyrolysis using thermogravimetric analysis (TGA). Bioresour. Technol. 102,
62306238.
Di Scala, K., Crapiste, G., 2008. Drying kinetics and quality changes during drying of
red pepper. LWT Food Sci. Technol. 41, 789795.
_ 2007. Air-drying characteristics of tomatoes. J. Food Eng. 78, 12911297.
Doymaz, I.,
_ 2010. Thin-layer drying characteristics of sweet potato slices and
Doymaz, I.,
mathematical modelling. Heat Mass Transfer 47, 277285.
_ Tugrul, N., Pala, M., 2006. Drying characteristics of dill and parsley
Doymaz, I.,
leaves. J. Food Eng. 77, 559565.
Fagerns, L., Brammer, J., Wiln, C., Lauer, M., Verhoeff, F., 2010. Drying of biomass
for second generation synfuel production. Biomass Bioenergy 34, 12671277.
Gardelle, B., Duquesne, S., Vu, C., Bourbigot, S., 2011. Thermal degradation and re
performance of polysilazane-based coatings. Thermochim. Acta 519, 2837.
Kaya, A., Aydin, O., 2009. An experimental study on drying kinetics of some herbal
leaves. Energy Convers. Manage. 50, 118124.
Madhava, M., Rao, P.S., Goswami, T.K., 2001. Drying kinetics of paddy using
thermogravimetric analysis. Drying Technol. 19, 12011210.
Midilli, A., Kucuk, H., 2003. Mathematical modeling of thin layer drying of pistachio
by using solar energy. Energy Convers. Manage. 44, 11111122.
Midilli, A., Kucuk, H., Yapar, Z., 2002. A new model for single-layer drying. Drying
Technol. 20, 15031513.
Miranda, M., Vega-Glvez, A., Garca, P., Di Scala, K., Shi, J., Xue, S., Uribe, E., 2010.
Effect of temperature on structural properties of Aloe vera (Aloe barbadensis
Miller) gel and Weibull distribution for modelling drying process. Food Bioprod.
Process. 88, 138144.
Raee, S., Keyhani, A., Shari, M., Jafari, A., Mobli, H., Tabatabaeefar, A., 2009. Thin
layer drying properties of soybean (Viliamz cultivar). J. Agric. Sci. Technol. 11,
289300.
Roberts, J.S., Tong, C.H., 2003. The development of an isothermal drying apparatus
and the evaluation of the diffusion model on hygroscopic porous material. Int. J.
Food Prop. 6, 165180.
Rovedo, C.O., Suarez, C., Viollaz, P.E., 1995. Drying of foods: evaluation of a drying
model. J. Food Technol. 26, 112.
Shi, J.L., Pan, Z.L., McHugh, T.H., Wood, D., Hirschberg, E., Olson, D., 2008. Drying and
quality characteristics of fresh and sugar-infused blueberries dried with
infrared radiation heating. LWT Food Sci. Technol. 41, 19621972.
Sims, R.E.H., Mabee, W., Saddler, J.N., Taylor, M., 2010. An overview of second
generation biofuel technologies. Bioresour. Technol. 101, 15701580.
Srikiatden, J., Roberts, J., 2006. Measuring moisture diffusivity of potato and carrot
(core and cortex) during convective hot air and isothermal drying. J. Food Eng.
74, 143152.
Vaccarezza, L.M., Lombardi, J.L., Chirife, J., 1974. Kinetics of moisture movement
during air drying of sugar beet root. J. Food Technol. 6, 317327.
Vega-Glvez, A., Miranda, M., Daz, L.P., Lopez, L., Rodriguez, K., Di Scala, K., 2010.
Effective moisture diffusivity determination and mathematical modelling of the
drying curves of the olive-waste cake. Bioresour. Technol. 101, 72657270.
Vega-Glvez, A., Notte-Cuello, E., Lemus-Mondaca, R., Zura, L., Miranda, M., 2009.
Mathematical modelling of mass transfer during rehydration process of Aloe
vera (Aloe barbadensis Miller). Food Bioprod. Process. 87, 254260.
Vega-Glvez, A., Uribe, E., Perez, M., Tabilo-Munizaga, G., Vergara, J., Garcia-Segovia,
P., Lara, E., Di Scala, K., 2011. Effect of high hydrostatic pressure pretreatment on
drying kinetics, antioxidant activity, rmness and microstructure of Aloe vera
(Aloe barbadensis Miller) gel. LWT Food Sci. Technol. 44, 384391.
Yang, H.P., Yan, R., Chen, H.P., Zheng, C.G., Lee, D.H., Liang, D.T., 2006. In-depth
investigation of biomass pyrolysis based on three major. Energy Fuels 20, 388
393.

You might also like