Chemical Engineering and Processing 70 (2013) 294300

Contents lists available at SciVerse ScienceDirect

Chemical Engineering and Processing:

Process Intensication
journal homepage: www.elsevier.com/locate/cep

Short communication

The shrinking core model applied on anaerobic digestion

Dominik da Rocha a, , Eckhard Paetzold b , Norbert Kanswohl c
a

Interdisciplinary Faculty, University of Rostock, Germany

Leibniz Institute for Catalysis, Rostock, Germany
c
Faculty of Agricultural and Environmental Sciences, Agricultural Technology, University of Rostock, Germany
b

a r t i c l e

i n f o

Article history:
Received 7 January 2011
Received in revised form 23 January 2013
Accepted 3 May 2013
Available online 6 June 2013
Keywords:
Shrinking core model
Mass transport
Anaerobic digestion
Lignocellulosic bers
Hydrothermal treatment

a b s t r a c t
In this study, the gas formation of anaerobic digestion was analyzed by the shrinking core model. This
model is based on the mass transport equations. The experiments were carried out with hydrothermal
treated wheat straw. Additionally a control group of untreated wheat straw was examined.
With untreated straw the beginning of microbiological growth was limited by convection through the
surrounding uid lm. With further incubation time the bacteria formed a biolm. Diffusion through this
layer limited the degradation.
A short hydrothermal treatment decreased the convection-limited phase.
The gas yield of the straw was 0.54 dm3 (0 C, 1 atm) per gram volatile solid. The pretreated straw yielded
in 0.51 dm3 (0 C, 1 atm) per gram volatile solids with the same mean content of methane (49 vol%) and
carbon dioxide (51 vol%).
2013 Elsevier B.V. All rights reserved.

1. Introduction
The shortage of resources is a challenge for the energy production. It will be necessary to use renewable energy sources. The
energetic utilization of organic residues is an interesting possibility, because they are left over in high amounts in agriculture.
The residues mostly consist of lignocellulosic bers. These are a
composite of carbohydrates (cellulose, hemicelluloses) and lignin,
a phenolic macromolecule. The lignocelluloses provide the plants
framework. It also protects the plant against physical and biological
inuence from the environment. These properties deteriorate the
utilization of residues for energy production, [1].
Compared to energy crops the methane yield and the degradation kinetic of lignocelluloses are lower. The bers also impede the
pump and stirring properties of the ferment, Chen et al. [2], Karla
and Panwar [3,4]. Pretreatments similar to the ethanol production
from lignocellulosic matter could solve these problems, [57].
Batstone et al. [8] published the Anaerobic Digestion Model
No. 1 (ADM1) and gave a good standard for modeling the biogas
process. In case of straw and other crops rich of ber the rate
limiting step is the disintegration and hydrolysis of the carbohydrates. They assumed in a rst approximation a rst order kinetic
for both processes and suggested a more accurate surface related
model if necessary. This surface related model was developed by
Vavilin et al. [9]. They assumed shrinking spherical particles for the

Corresponding author. Tel.: +49 381 498 3340; fax: +49 381 498 3346.
E-mail address: dominik.da.rocha@gmail.com (D. da Rocha).
0255-2701/$ see front matter 2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.cep.2013.05.003

hydrolysis of cellulose and sewage sludge. They divided the kinetic

in two phases. The First one is the colonization of the particle surface by hydrolytic bacteria. In a second phase the bacteria grow on
the surface and the particle shrinks.
In the case of lignocellulosic bers the shape of the particles is
not spherical but rather cylindrical, as suggested by [10]. He also
mentioned that the macroscopic shape of the particles was not as
important for the hydrolysis kinetics, because in the scale of the
organisms the particles shape appears nearly planar. However, the
microstructure of the plant material consists mostly of lignin, celluloses and hemicelluloses and these are organized in a structure
of cylindrical bundles.
Furthermore the particles do not shrink because the framework of the lignin is not biodegradable. Therefore, in this paper
the shrinking core model was used to describe the hydrolysis of
lignocellulosic framework.
2. Materials and methods
2.1. Materials
Wheat straw was used to study the biological degradation of
lignocellulosic bers. To get bers with nearly same size, the straw
was comminuted in a knife mill and it was sieved through 1 mm
and 0.25 mm mesh. For investigations the sieve residual of the
0.25 mm mesh was used. The water content of the straw amounted
to 8.7 wt%. It was dried at 105 C for 17.5 h. The dry matter was
combusted at 600 C for 8 h and yielded an ash content of 3.0 wt%
of the dry matter.

D. da Rocha et al. / Chemical Engineering and Processing 70 (2013) 294300

295

Nomenclature
Symbols
a
b
B
Bo
c
D
Da
g
G
k
m
n
r
R
t
V
X
Y




Indices
0
1st
B
C
CH4
CO2
eff
end
ex
F
G
P
R
S

stoichiometric coefcient hydrogen

stoichiometric coefcient oxygen
bacteria
Bodenstein number
concentration
diffusion coefcient
Damkhler number
stoichiometric coefcient of gas
gas
reaction rate constant, convective transfer coefcient
mass
stoichiometric coefcient carbon
variable radius
constant radius
time
volume
conversion of substrate
yield
relative time (time over incubation time)
relative radius (core radius over ber radius)
incubation time
concentration ratio of bacteria in uid per bacteria
in the solid

at starting time
rst order kinetic
bacteria
core, convective
methane
carbon dioxide
effective
at end of incubation time
exchange
uid
gas
particle
reaction
substrate, solid

The inoculum arose from an agricultural anaerobic digester

(Dummerstorf, Germany), mainly fed with corn silage and cow
manure. Water and ash contents were determined. They amounted
to 91.9 wt% of water and 20.8 wt% ash referring to the dry matter.
After taking from the digester, the inoculum was stored two weeks
at 20 C to decrease own gas formation. The light microscopic studies were carried out with a Carl Zeiss Laboval 4 at a magnication
of 400.
2.2. Anaerobic digestion
The gas formation was observed according to the Hohenheimer
Yield Test, [11]. Therefore 100 ml glass syringes were used as
digesters. They are sorted in a carousel for mixing the ferment.
The whole apparatus was placed in an incubator (Memmert INE600) at 38 C. The syringes had a scale of 1 ml steps to measure
the produced gas. They were sealed with highly viscous parafn.
The digester was lled with 30 g of inoculum and 1 g of wet wheat
straw. Three syringes were just lled with inoculum to subtract its
own gas formation from the inoculum.

Fig. 1. Wheat straw before (a) and after (b) 40 days of incubation, magnied 400
times with a light microscope. The microstructure of the straw is preserved after
the incubation.

The carbon dioxide and methane contents were measured with

a Brel & Kjr gas monitor type 1302. This device was tightly connected to a 500 ml Erlenmeyer ask by the gas inlet and outlet. For
determination of the gas, 2 ml of the formatted biogas were injected
into the ask and well mixed by the internal pump of the gas monitor. When the maximum capacity of the gas monitor (4000 ppm)
was reached, the ask was ushed with silica gel dried air. With
this assembly the ratio of carbon dioxide and methane could be
determined with a precision of 2%. The mass of the produced gas
was calculated under assumption of ideal gas behavior.
2.3. Hydrothermal treatment
The hydrothermal treatment was done in a 100 ml autoclave
(Parr reactor No. 4593, temperature controller 4842). The reactor
was equipped with a magnetically coupled blade stirrer, which was
running at 1000 rpm. It was lled with 6 g straw (wet) and 54 g
deionized water. The heat up time to reach 200 C amounted to
40 min. By reaching the predened temperature the reaction was
stopped by cooling the reactor in an ice-water bath. The time for
cooling to 40 C required less than 5 min. The pressure could be
observed by a gauge (050 bar).
2.4. Characterization
The gas yield of the untreated straw amounted to 0.70 g of gas
per gram dry matter or 0.54 dm3 (0 C, 1 atm) per gram volatile
solids. The mean chemical composition of the gas was 51 vol% CO2
and 49 vol% CH4 .
The straw, which was treated at 200 C for a short time, has
nearly the same biogas yield, 0.67 g of gas per gram dry matter
or 0.51 dm3 (0 C, 1 atm) per gram volatile solids. The mean carbon dioxide content was 49 vol% and the mean methane content
amounted to 51 vol%.
The degradation of straw bers was examined under a light
microscope. Fig. 1a and b shows the straw bers before and after 40
days of incubation. The micro structure of the bers is unscathed
over the time.

296

D. da Rocha et al. / Chemical Engineering and Processing 70 (2013) 294300

3.2. The reaction mechanism

Fig. 2. Scheme of mass ux steps in the shrinking core model. The section of a ber
shows the shrinking unreacted core over the time, Levenspiel [14].

3. The shrinking core model

3.1. Model basics
The shrinking core model was rst published by [12,13], and was
well explained by [14]. It describes the mass ux between particles
and uids. The basic idea is that the uid ushes into the particle and reacts to uid and solid products respectively. The uid
products can leave the particle, while the solids form a layer. An
unreacted shrunken core remains in the middle of the particle.
The mass ux proceeds in ve steps (Fig. 2):
- Convection of reactants from the uid main body through the
surrounding uid lm.
- Diffusion through the solid products layer.
- Reaction at the cores surface.
- Diffusion of uid products through the solid layer.
- Convection of uid products through the uid lm back in the
uid main body.
Primarily the model was developed to describe the combustion
of particles. In the course of time it has been applied to many other
uid-particle-reactions.
Subramanian et al. [25] published the application of the shrinking core model on the discharge of metal hydride electrodes. In
addition, the shrinking core model could describe the hydrolysis of
PET (polyethylene terephthalate) in sulfuric and nitric acid, [15,16].
Fowler and Crundwell [17] showed that a formed sulfur layer
limited the leaching of zinc sulde by ferric sulfate. In this investigation the formation of the sulfur layer was prevented by adding
sulfur-consuming bacteria (Thiobacillus ferrooxidans). A higher
yield and a faster leaching were the results.
In the case of the anaerobic digestion of lignocellulosic bers,
the particles shape is a cylinder. According to [14] the conversion
as a function of the shrinking core radius is: ( = rc /R, core radius
over ber radius):
X = 1 2

(2)

n
2

b
a

4
2

a
b

8
4

H2 O

n

a
b
+
8
4

CO2

CH4

(5)

The formation of ammonia and hydrogen sulde is neglected

because the straw consists mainly of carbohydrates.
3.3. The conversion
The conversion is calculated by:
X=

mS,0 mS
mS,0

(6)

In this equation mS,0 is the mass of fermentable substance in the

straw. If a complete conversion is assumed, it is given by:
mS = mCO2 + mCH4 + mB

(7)

The mass of formed carbon dioxide and methane depends on

growth of bacteria:
mB = (mCO2 + mCH4 )YB

(8)

So that the mass of substrate is:

mS = (mCO + mCH4 )(1 + YB )

(9)

The bacteria yield (YB ) could be canceled from the equation of

the conversion:
X=

mCO2 ,end + mCH4 ,end (mCO2 + mCH4 )

mCO2 ,end + mCH4 ,end

(10)

3.4. Reaction kinetics

3.4.1. Fluid lm controls
According to [14] the mass ux equation for convection is:
(11)

(3)

The mass exchange surface is the bers lateral surface. It could

be assumed that the ber is much longer than thick. Hence the end
face is negligible.

(4)

Sex = 2LR

If the reaction rate controls the mass ux:

 = 1 (1 X)1/2

dnB
= kC (cB,F cB )
Sex dt

If diffusion through the solid layer controls:

 = X + (1 X) ln(1 X)

Cn Ha Ob +

(1)

If the convection through the surrounding uid lm controls

then the relative reaction time as function of the conversion is:
=X

The reaction mechanism for anaerobic digestion is very complex. Sahm divided the degradation in a chain reaction system
of four phases. The degradation starts with the hydrolytic and
acidogenic phase. Macromolecules like proteins, fats and carbohydrates degrade to higher organic acids and alcohols. The acetogenic
phase follows. Higher organic acids and alcohols are reduced to
acetic acid, carbon dioxide and hydrogen. The last phase is the
methanogenic, where methane is formed, [18].
Noike et al. determined the rate-limiting step of anaerobic
digestion. In experiments with glucose, starch and cellulose, they
pointed out, that the degradation of cellulose characterizes the
overall reaction rate. The maximum rate constant for the degradation of cellulose was 1.25 per day and for the methanogenesis
the rate constant amounted to 10.9 per day, Noike et al. [19].
Therefore the rate of the biogas formation is nearly equal to the
degradation rate of the straw bers. The reaction system is assumed
to: Digestible carbohydrates (Cn Hm Ok ) react in presence of bacteria
to carbon dioxide and methane. According to the simplied Buswell
equation [20]:

(12)

D. da Rocha et al. / Chemical Engineering and Processing 70 (2013) 294300

Additionally the concentration at the ber surface is much

smaller than in the uid, while convection is controlling the mass
ux.
cB,F cB cB,F

(13)

The bacteria amount is the cell density in the solid multiplied

by the biolm volume. The biolm volume is tube shaped.dnB =
cB,S dV = cB,S d(L(R2 rC2 )) = cB,S 2LrC drC
dnB = cB,S 2LrC drC

A dimensionless equation follows. The parameters could be

reduced to a 1st order Damkhler number and a concentration ratio
of bacteria in uid per bacteria in the solid [26].

(16)

kC 
DaC =
R

(17)

(18)

(19)

3.4.2. Reaction rate controls

The mass ux equation if the reaction rate controls the degradation is:
(20)

(25)

The exchange surface is the variable core surface:

(26)

The concentration at the core will always be zero and the concentration at the bers surface is equal to the uid concentration:

dnB
2Ldt

rC

dnB
ln
2Ldt

dr
= Deff
r

r 
C

dcB
cB,F

= Deff cB,F

cB,S 2LrC drC

ln
2Ldt

r 
C

= Deff cB,F

Multiplied by incubation time and divided by the ber radius

the equation is:
drC 
 cB,F
Deff
=
dt R
rC ln(rC /R) R cB,S
d
R
D  cB,F
= eff2
d
R cB,S rC ln(rC /R)
A dimensionless equation follows. The parameters can be
reduced to the Bodenstein number and a concentration ratio of
bacteria in uid per bacteria in the solid [26]
Bo =

Deff 
R2

d
Bo
=
 ln()
d

Now the exchange surface is the cores surface:

Sex = 2LrC

(24)

Including Eq. (10):

If the convection through the surrounding uid lm controls the

mass ux, the change of the relative radius by time is:

dnB
= kR cB
Sex dt

d
= DaR
d

(15)

t
=


d
DaC
=

d

(23)

Sex = 2Lr

drC 
kC  cB,F R
=
R cB,S rC
dt R

cB,F
cB,S

kR 
R

dnB
dcB
= Deff
Sex dt
dr

The equation multiplied by incubation time and divided by the

ber radius:

DaR =

3.4.3. Diffusion controls

Ficks law gives the mass ux for diffusion:

cB,0 2LrC drC

= kC cB,F
2LRdt

rc
R

A dimensionless equation follows. The parameters could be

reduced to a 1st order Damkhler number and a concentration ratio
of bacteria in uid per bacteria in the solid [26].

(14)

All inserted in the mass ux equation:

=

297

(27)

(28)

(21)

If the reaction rate controls the degradation, the biolm will be

small or the diffusion is so fast the bacteria concentration at the
core is nearly the uid concentration.

3.4.4. Combining mass transport mechanisms

If not a single transport mechanism controls the mass ux, the
change of relative radius can be added.

cB cB,F

dtotal = dConv + dDiff + dReac

(22)

The change of the core radius is described as:

cB,S 2LrC drC
= kR cB
2LrC dt
Multiplied by incubation time and divided by the ber radius
the equation is:
drC 
kR  cB,F
=
R cB,S
dt R

Therefore, the developed equations can be combined by adding

the reciprocal:
d

=
d
 ln()/Bo /DaC 1/DaR

(29)

To integrate the ordinary differential equation a RungaKotta

algorithm included in the software MATLAB was used, Dormand
and Prince [21]. The differential equation is solved in a MATLAB
example script in the appendix.

298

D. da Rocha et al. / Chemical Engineering and Processing 70 (2013) 294300

4. Results and discussion

4.1. Applicability of the shrinking core model
For the equation suggested by [14], the bacteria concentration
and the ber radius had to be known. These parameters were hard
to measure. The different species of bacteria were not dened.
Counting the cells under the microscope did not give the number
of active organisms. There were also microorganisms on and in the
particles. They were not countable under a microscope because the
particles were opaque.
That is why the parameter for the bacteria concentration
was assumed to a value of 1. This parameter is a constant
factor in the Eqs. (19), (24), (28), (29) so that experiments
with the same inoculum were comparable. Maybe the shrinking core model could be used to compare experiments with
different inoculums, if a standard substrate is used as reference. The parameter of the bacteria could be calculated
indirectly.
The reduction to dimensionless parameters had the disadvantage that the absolute value got lost. However, it was a helpful
instrument to compare the degradation kinetics.
To describe the geometry of the ber a cylinder was assumed
(Eq. (1)). This relationship is valid for any other bodies that
are longer then thick. The equation described the degradation
in two spatial dimensions. This two-dimensional degradation
was also observed in the enzymatic degradation of cellulose
[22].
The degradation of the cellulose ber started in the amorphous
regions. Here the endoglucanases broke the -glycosidic bonds. It
exposed glucose molecules from the chain. This is the degradation
in the radial direction (rst spatial dimension). The cellobiohydrolases can degrade the ber to cellobiose units. This degradation
proceeds in axial direction (second spatial dimension). Cellobiose
is soluble in water and can leave the ber. Therefore the ber radius
(R) in the equations is not equal to the geometric size of the particles. The Radius is rather the size of the cellulose bundles in the
straw-micro-structure. So hollow spaces and swelling of the straw
ber only have an effect on the untreated straw at the beginning of
the fermentation.
The degradation of the untreated straw started with a
convection-limiting step. In this step the bacteria had to reach the
cellulose bundles. The main effect of the hydrothermal pretreatment was to expose the bundles by solving the hemicelluloses from
the straw.

4.2. Biogasformation of the untreated straw

The microorganisms started the degradation with a delay. Usually this delay is interpreted as the LAG-phase of bacterial growth.
In this phase the microorganisms adapt their metabolism to the
substrate [23].
The gas formation course was calculate with a 1st order kinetic.
The reaction rate constant (k1st ) for the untreated straw was 0.1 s1 .
Compared to a 1st order kinetic the shrinking core model with a
diffusion limited rate (with Bo = 0.25 and = 1) is more accurate
(Fig. 3a).
The equation for convection limited mass ux tted the
LAG-phase well with a Damkhler number of 0.6. Therefore
it seems that the microorganisms had to settle on the ber.
If the surface of the ber is completely covered, the equation
for diffusion-limited degradation tted the measured degradation with a Bodenstein number of 0.263. The parameter for
the bacteria concentration was assumed to value of 1
(Fig. 4a).

Fig. 3. (a and b) Gas formation of the untreated (a) and pretreated straw (b), the
solid curve was calculated with the shrinking core model (diffusion controlled,
Eq. (28) with Bo = 0.25 and = 1 for (a) and (b)). It tted the course of the measured points more accurate compared to a 1st order Kinetic (k1st = 0.1 s1 for (a) and
k1st = 0.12 s1 ).

4.3. Biogasformation of the pretreated straw

The pretreatment of the straw increases the degradation at
beginning of the biogas formation. The hemicellulose of the straw
was extracted by the hydrothermal treatment. It was more available for the bacteria. The smaller gas yield of the treated straw
was an indicator for the formation of indigestible reaction products
during the hydrothermal treatment (Fig. 3b).
In order to reduce this effect the heat up time for the hydrothermal treatment will have to be reduced. A more complex
apparatus will be necessary. For example a Percolator with preheated water can reduce the heat up time, according to [24].
Also for the pretreated straw the gas formation course was tted
by a 1st order kinetic. The rate constant (k1st ) amounted to 0.12 s1 .
The shrinking core model with a duffusion limited rate equation is
more accurate compared to the 1st order kinetic (Fig. 4b).
4.4. Implementation in the ADM1
The ADM1 and other anaerobic digestion models were usually
based on insteady mass balance. The kinetic equations were solved
to the derivation of the concentration by time. So the diffusion
limited rate Eq. (28) was formed to such an expression.
The conversion was dened as:
X =1

cs
cS,0

D. da Rocha et al. / Chemical Engineering and Processing 70 (2013) 294300

299

the dimension mole per time and volume. The following expression
could implemented to the ADM1.
dcS
=
dt

kHyd

ln

cs
cS,0

(30)

The practical use of the shrinking core model is similar to the

two phase model suggested in the ADM1 Batstone et al. [8]. The
default kinetic should be the rst order kinetic. Only if a very low
residence time will be modeled, the difference between 1st order
kinetic and the shrinking core model will be recognizable.
Another possible application is the modeling of batch or solid
bed fermenters if it is necessary to calculate the concentration of
the reactants over the residence time.
5. Conclusions
The analysis of gas formation with the shrinking core model
pointed out that the model was applicable. It could be shown that
the diffusion limited transport mechanism dominates the degradation of the straw. Eqs. (29) and (30) could be implemented in the
Anaerobic Digestion Model No. 1 to calculate the the concentration
of the reactants over the residence time. The hydrothermal treatment increases the biogas formation at the beginning. The solved
pentoses are faster available for the microorganism. On the other
hand the absolute gas yield decreased because of the formation of
unfermentable reaction products.
Acknowledgment
Thanks to Stefanie May for spell and grammar checking.
Fig. 4. (a and b) Shrinking-core-plot of the untreated (a) straw, at the beginning the
convection limited Eq. (19) tted the course of the measured points. After this initial
phase the course was calculated for the diffusion limited Eq. (28). The pretreated
straw (b) degrades according to the diffusion limited kinetic.

X =1

 r 2
C

Therefore the relative radius was:

rC
=
R

cs
cS,0

1/2

Ficks law of diffusion in the partially integrated form for a cylinder:

dnB
ln
2Ldt

r 
C

= Deff cB,F

The change of bacteria was proportional to the change of substrate. So the biomass yield was inserted. The exponent in the
logarithm was extracted to a factor and the equation was divided
by the fermenter volume (VR ).
4LDeff cB,F
dcS
 
=
dt
VR ln ccs
S,0

According to [22] the hydrolysis rate did not depend on the

bacteria concentration. Because the microorganisms began to produce their exoenzymes if other easier carbon sources were running
out. Therefore the change of substrate was independent from the
concentration of the bacteria in the uid. Furthermore are ber
length, diffusion coefcient and fermenter volume are constants.
These parameters were combined to a single constant (kHyd ) with

References
[1] S.R. Decker, et al., Biomass Conversion, vol. 2, Kent and Riegels Handbook of
Industrial Chemistry and Biotechnology, 2007, pp. 14491548.
[2] Y. Chen, J.J. Cheng, K.S. Creamer, Inhibitition of anaerobic digestion process: a
review, Bioresource Technology 99 (10) (2008) 40444064.
[3] M.S. Karla, J.S. Panwar, Anaerobic digestion of rice crop residues, Agricultural
Waste 17 (1986) 263269.
[4] Sharma, K. Sudhir, et al., Effect of particle size on biogas generation from
biomass residues, Biomass 17 (1988) 251263.
[5] I. Ballesteros, et al., Ethanol production from steam-explosion preatreated
wheat straw, Applied Biochemistry and Biotechnology 129132 (2006)
496508.
[6] O. Bobleter, Hydrothermal degradation and fractionation of saccharides and
polysaccharides. [book auth.] Severian Dumitriu. Polysaccharides: structural
diversity and functional versatility, 2nd edition, Marcel Dekker, New York,
2005, pp. 839935.
[7] Y. Sun, J. Cheng, Hydrolysis of lignocellulosic materials for ethanol production:
a review, Bioresource Technology 83 (1) (2001) 111.
[8] C.J. Batstone, et al., Anaerobic Digestion Model No. 1, IWA Publishing, Padstow
Cornwall, 2002, ISBN 1-900222-78-7.
[9] V.A.V. Vavilin, S. Rytov, L.Y. Lokshina, A Description of hydrolysis kinetics in
anaerobic degradation of particulate organic matter, Bioresource Technology
56 (1996) 229237.
[10] P.N. Hobson, A model of some aspects of microbial degradation of particulate
substrates, Journal of Fermentation Technology 4 (1987) 431439.
[11] D. Hellfrich, H. Oechsner, Hohenheimer Biogasertragstest, Argrartechnische
Forschung 9 (3) (2003) 2730.
[12] S. Yagi, D. Kunii, Chemical Engineering (Japan) 19 (500) (1955) 231.
[13] S. Yagi, D. Kunii, Chemical Engineering Science 16 (1961) 364.
[14] O. Levenspiel, Chemical Reaction Engineering, 3rd edition, Wiley, New York,
1999, ISBN 0-471-25424-X.
[15] T. Yoshioka, N. Okayama, A. Okuwaki, Kinetics of hydrolysis of PET powder
in nitric acid by a modied, Industrial & Engineering Chemistry Research 37
(1998) 336340.
[16] T. Yoshioka, T. Motoki, A. Okuwaki, Kinetics of hydrolysis of polyethylene
terephtelate powder in sulfuric acid by a modied shrinking-core model, Industrial & Engineering Chemistry Research 40 (2001) 7579.
[17] T.A. Fowler, F.K. Crundwell, Leaching of zinc sulde by thiobacillus ferrooxidans, Applied and Enviromental Microbiology 65 (1999) 12.
[18] Sahm, Hermann, Biologie der Methan-Bildung, Chemical Engineering Technology 53 (11) (1981) 854863.

300

D. da Rocha et al. / Chemical Engineering and Processing 70 (2013) 294300

[19] T. Noike, et al., Characteristics of carbohydrate degradation and the ratelimiting step in anaerobic digestion, Biotechnology and Bioengineering
(1985).
[20] G.E. Symons, A.M. Buswell, The methane fermentation of carbohydrates, Journal of the American Chemical Society 55 (5) (1933) 20282036.
[21] J.R. Dormand, P.J. Prince, A family of embedded RungeKutta formulae, Journal
of Computational and Applied Mathematics 6 (1980) 1926.
[22] P. Bguin, J.-P. Aubert, The biological degradation of cellulose, Microbiology
Reviews 13 (1994) 2558.

[23] M.T. Madigan, J.M. Martinko, J. Parker, Brock Biology of Microorganisms, 9th
edition, 2000, ISBN 978-0130819222.
[24] O. Bobleter, G. Bonn, The hydrothermolysis of cellobiose and its reactionproduct
d-cellulose, Carbohydrate Research 124 (1983) 185193.
[25] V.R. Subramanian, et al., Shrinking core model for the discharge of a
metal hydride electrode, Journal of Elektrochemical Society 147 (8) (2000)
28682873.
[26] M. Zlokarnik, Scale-up in Chemical Engineering, Wiley, New York, 2002, pp.
219, ISBN 3527302662.