Modelling Diffusion and Reaction for Inert-Core Catalyst in Batch

and Fixed Bed Reactors

Ping Li,1* Guohua Xiu2 and Alirio E. Rodrigues3

1. State Key Laboratory of Chemical Engineering, College of Chemical Engineering, East China University of Science and Technology,
Shanghai 200237, China
2. Shanghai Monodomain Chemical Technology Co., Ltd., Shanghai 201206, China
3. Laboratory of Separation and Reaction Engineering, Dept. of Chemical Engineering, Faculty of Engineering, University of Porto, Rua Dr.
Roberto Frias, s/n 4200-465 Porto, Portugal

New analytical solutions of concentration time curves are derived for an isothermal inert-core spherical catalyst based on the mathematical models
by taking into account first-order irreversible reaction and mass transfer resistances in batch and fixed bed reactors.[1,2] The effects of mass transfer
resistances and Thiele modulus on catalytic efficiency and conversion of reactant are examined over a wide range of parameters for an inert-core
catalyst. The results show that the inert-core catalyst can significantly improve the efficiency for fast catalytic reactions, where mass transfer
limitations occur; however, the reactant conversion will decrease due to a lower loading ratio of active species of catalyst for inert-core catalyst.

Keywords: inert-core catalyst, first-order reaction, mass transfer resistances, reactors, analytical solutions

INTRODUCTION but also has a great significance in investigating the reaction and
mass transfer in the catalyst pellet.[1,33] Catalytic efficiency of the

T
he catalyst performance indexes include effectiveness, catalyst particle may be restrained by both intraparticle diffusion
selectivity, yield, and deactivation rate. The key parameters resistance and external mass transfer resistance, especially for fast
which affect the choice of the catalyst are kinetic behaviour catalytic reactions.[1,2,23–29] The application of the inert-core
of the reactions, the mass transfer limitations, and the production catalyst can significantly improve the catalyst efficiency in this
cost of the catalyst. case.
Increased attention has been given to the catalyst synthesis with Fixed bed reactors have traditionally been used for most large-
a non-uniform distribution of catalyst activity within the pellet, in scale catalytic reactors because of their efficiency, low cost, and
order to improve the selectivity of catalyst and increase the catalyst ease of construction, operation, and maintenance. Efforts have
efficiency.[3–7] Sometimes uniform distribution of catalyst activity been made in studying the reaction kinetics, the mass transfer
is undesirable, especially when the cost of the catalyst is the main mechanism, and the mathematical model of fixed bed reactors.
consideration, such as in the case of noble metal catalysts,[3,8–12] or Lots of researchers have focused on both steady-state
when diffusion limitation occurs.[13–22] More specific design of operations[24–26] and unsteady-state operations.[2,27–29] Therefore,
catalyst with a non-uniform distribution is a catalyst comprising a modelling of mass transfer and reaction by an inert-core catalyst in
thin annular shell of catalytically active material and an inert, non- a fixed bed reactor is of interest.
catalytically active, nonporous core, called a inert-core catalyst. Our objective is to derive new analytical solutions for transient
The inert-core particles minimize the diffusion distance inside the and volume-averaged concentration profiles of reactant in the
particle to just the porous layer of the shell, resulting in faster catalytic shell of an inert-core catalyst particle for a batch reactor,
movement in and out of the particle leading to less diffusion and to predict the effluent concentration profiles with the existence
limitation. of mass transfer resistances for a fixed bed reactor for an
A central issue in design, operation, optimization, and control of isothermal, first-order irreversible reaction, to reveal the relation-
batch or fixed bed reactors is the prediction of the concentration ship between the concentration time expressions for batch and
time relationship or breakthrough curve of the effluent stream. fixed bed reactors, furthermore, to evaluate the effects of mass
Mathematical modelling of dynamics of batch and fixed bed transfer resistances and Thiele modulus on catalytic efficiency and
reactors is extensive and it possesses a rich history.[1,2,23–29] the conversion of reactant over a wide range of parameters for
However, newly developed batch and fixed bed reactors with inert-core catalysts.
applications use inert-core spherical particles, and those solutions
are no longer valid to predict the concentration time relationship
or the elution curves for inert-core catalysts. Some investigations
have been done for reactive chromatography for inert-core * Author to whom correspondence may be addressed.
adsorbents.[30–32] It is therefore most desirable to develop a E-mail address: liping_2007@ecust.edu.cn
mathematical model and present the analytical solutions for batch Can. J. Chem. Eng. 97:217–225, 2019
and fixed bed reactors with inert-core catalysts. © 2018 Canadian Society for Chemical Engineering
DOI 10.1002/cjce.23189
The concentration distribution of reactant in a porous catalyst Published online 12 March 2018 in Wiley Online Library
pellet not only is the basis of calculating the effectiveness factor, (wileyonlinelibrary.com).

VOLUME 97, JANUARY 2019 THE CANADIAN JOURNAL OF CHEMICAL ENGINEERING 217
MATHEMATICAL MODEL Fixed Bed Reactor
Figure 1 shows the schematic configuration of an inert-core @C @ 2 C u0 @C ð1  eb Þ 3 h   i
¼ DL 2   kf C  cp r¼R ð6Þ
catalyst particle, a porous catalytic shell with a non-porous and @t @Z eb @Z eb R
non-catalytic inert core. Uniform distribution of catalyst activity
inside the porous shell and isothermal conditions are assumed; the CðZ ¼ 0; tÞ ¼ C0 ð7Þ
reaction rate law is assumed to be first-order irreversible, with
constant effective diffusivity of the reactant species inside the CðZ ! 1; tÞ ¼ 0 ð8Þ
porous support, and homogeneous distributed catalyst activity in
the shell. For the first-order reaction coupled with mass transfer CðZ; t ¼ 0Þ ¼ 0 ð9Þ
process, the material balance equation of reactant in the inert-core
porous shell support matrix is then given by: in which Z is the axial distance from the column entrance; u0

 denotes superficial velocity; DL is the axial dispersion coefficient;
@cp 1 @ 2 @cp
ep ¼ De 2 r  kcp ðRC  r  RÞ ð1Þ and eb is the fraction void volume in the column.
@t r @r @r
The number of parameters used in this mathematical model can
in which cp is the concentration in the catalyst; r is the radial be reduced by non-dimensionalization. For our work, the
distance of the catalyst particle; t is the time; ep is the shell porosity; following dimensionless variables and groups are introduced:
De is effective pore diffusivity of the reactant; R is the radius of the
catalyst; RC is the radius of the inert core; and k is the first-order cp C
x¼ ;y ¼ dimensionless concentrations
reaction rate constant. C0 C0
The pore diffusion equation, Equation (1), is coupled with the
external bulk concentrations through the following boundary and Z
z¼ dimensionless axial distance
initial conditions: L

 h r
@cp   i j¼ dimensionless radial distance
De ¼ kf C  cp r¼R ð2Þ R
@r r¼R

 u0 t
@cp t¼ dimensionless contact time
¼0 ð3Þ Leb
@r r¼RC
kf R
Bi ¼ Biot number
cp ðr; t ¼ 0Þ ¼ 0 ð4Þ De

in which C is the reactant concentration in the fluid phase; kf is the Pe ¼

u0 L
Peclet number based on reactor length
external mass transfer coefficient. DL eb
For batch and fixed bed reactors, the mass balance equations for sffiffiffiffiffiffi
the reactant in the fluid phase with boundary conditions and initial R k
conditions respectively are as follows. f¼ Thiele modulus for spherical catalyst
3 De

Batch Reactor eb De L
Dr ¼ bed length parameter
For simplification, we assumed that the batch volume is big ep R2 u0
enough to maintain the bulk reactant concentration constant with
time, i.e.: ep ð1  eb Þ
g¼ porosity ratio
eb
Cðt ¼ 0Þ ¼ CðtÞ ¼ C0 ð5Þ
Then by writing in dimensionless form, Equations (1)–(9) become:


1 @x 1 @ @x
¼ 2 j2  9f2 x ðjC  j  1Þ ð10Þ
Dr @t j @j @j

 h i
@x
¼ Bi y  ðxÞj¼1 ð11Þ
@j j¼1


@x
¼0 ð12Þ
@j j¼jC

xðj; t ¼ 0Þ ¼ 0 ð13Þ

Batch Reactor
Figure 1. Catalyst comprising a thin porous shell of catalytically active
yðDrtÞ ¼ 1 ð14Þ
material and a non-porous inert core.

218 THE CANADIAN JOURNAL OF CHEMICAL ENGINEERING VOLUME 97, JANUARY 2019
Fixed Bed Reactor in which:
@y 1 @ y @y 2 h i
pffiffiffiffiffi
¼   3gDrBi y  ðxÞj¼1 ð15Þ An ¼ ½Bi  jC ln ð1  jC Þ  BijC ð1  jC Þ þ 1 þ j2C ln tan
@t Pe @z 2 @z pffiffiffiffiffi

ð1  jC Þ ln ð26Þ
yðz ¼ 0; tÞ ¼ 1 ð16Þ
and the dimensionless time parameter for batch reactor in
yðz ! 1; t Þ ¼ 0 ð17Þ Equation (25) is:

De t
yðz; t ¼ 0Þ ¼ 0 ð18Þ t b ¼ Drt ¼ ð27Þ
ep R2

In the following sections, we will first derive the new analytical

solutions by taking mass transfer limitations and first-order The time domain solution of Equation (25) is expressed in terms
irreversible reaction into account; the general formula and its of ln , which is determined from the transcendental equation:
derived equations are helpful in studying the transient mass  
transfer and reaction in batch and fixed bed reactors. pffiffiffiffiffi
pffiffiffiffiffi ðBi  1ÞjC þ 1
tan ð1  jC Þ ln ¼ ln ðln 6¼ 0Þ ð28Þ
jC ln þ 1  Bi
Solutions in Laplace Domain for Batch and Fixed Bed Reactors

Batch Reactor For conventional catalyst particle (i.e., jC ! 0), the solutions of
  Equation (25) has been previously obtained.[1]
Bi 1 The volume-averaged concentration of reactant over the catalyst

¼
x ð19Þ
jp YD ðpÞ þ Bi shell is defined as:

lcoth½lð1  jC Þ þ l2 jC ZjC
YD ðpÞ ¼ 1 ð20Þ 3
ljC coth½lð1  jC Þ þ 1 xav ¼  xj2 dj ð29Þ
1  j3C
rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi 1
p
l ¼ 9f2 þ ð21Þ which yields:
Dr
 
in which the over bar indicates the Laplace transform, p is Laplace Bi 3fð1  jC Þ þ 9f2 jC  1 tanh½3ð1  jC Þf
xav ¼ 2     
operator, and YD ðpÞ is the particle transfer function. 3f 1  j3C 3fðBijC þ 1  jC Þ þ 9f2 jC þ Bi  1 tanh½3ð1  jC Þf
pffiffiffiffiffi pffiffiffiffiffi
6Bi X1 ð1  jC Þ ln  ðln jC þ 1Þtan½ð1  jC Þ ln 
Fixed Bed Reactor     p ffiffiffiffiffi
1  j3C n¼1 9f2 þ ln An ln
 2 

1

y ¼ expðr 1 zÞ exp  9f þ ln tb ð30Þ
ð22Þ
p
rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi Effectiveness factor is frequently used to evaluate the catalyst
Pe 1 2 utilization with the existence of mass transfer resistances. The
r1 ¼  Pe þ Pe½p þ 3gDrGðpÞ ð23Þ
2 4 effectiveness factor is simply the average concentration divided by
the bulk concentration, or in dimensionless variables:
BiYD ðpÞ
GðpÞ ¼ ð24Þ
Bi þ YD ðpÞ hb ¼ limtb !1 xav ð31Þ

It is clear that the concentration expressions in the Laplace It is straightforward to derive the effectiveness factor hb for
domain for batch and fixed-bed reactors are linked by the particle inert-core catalyst from the first item on the right hand of
transfer function YD ðpÞ. Equation (30):
Analytical Solution in Time Domain for Batch and Fixed Bed
1
Reactors hb ¼  h ð32Þ
1  j3C
Batch Reactor
The transient concentration profiles inside a inert-core catalyst Equation (32) was first derived by Wang and Varma.[13]
particle are:

Bi sinh½3ðj  jC Þf þ 3jC fcosh½3ðj  jC Þf

x¼   
j 3fðBijC þ 1  jC Þcosh½3ð1  jC Þf þ Bi þ 9jC f2  1 sinh½3ðj  jC Þf
pffiffiffiffiffi pffiffiffiffiffi pffiffiffiffiffi pffiffiffiffiffi  ð25Þ
2Bi X1 ln sin½ðj  jC Þ ln  þ jC ln cos½ðj  jC Þ ln   

þ   pffiffiffiffiffi exp  ln þ 9f2 tb

j n¼1 An ln þ 9f2 cos½ð1  jC Þ ln 

VOLUME 97, JANUARY 2019 THE CANADIAN JOURNAL OF CHEMICAL ENGINEERING 219
 vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
uqffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
in which h is defined by: u  2 2  b  2
t 9f þ Dr  9f2
n o f2 ¼ ð46Þ
1 1þ3jC ftanh½3ð1jC Þf
 3f
1 2
f 3jC fþtanh½3ð1jC Þf
h¼ n o ð33Þ
1þ3jC ftanh½3ð1jC Þf
1 þ 3f
Bi 3jC fþtanh½3ð1jC Þf  3f
1

Derived Analytical Solutions for Fixed Bed Reactor

For high values of Pe (i:e:; DL ! 0), we can disregard all terms of
Equation (33) will be used to evaluate the reactant exit
order 1=Pe, Equation (34) becomes:
concentrations at the steady-state for the fixed bed reactor.
Z
Fixed Bed Reactor y1 1 1
y¼ þ expð3gDrBiI 1 zÞ  sin½bt  zðb þ 3gDrBiI 2 Þ
The breakthrough curve is expressed as: 2 p 0
db
0 sp ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
ffi1 ðPe ! 1Þ ð47Þ
Z ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi b
1 2 þ b2 þ a
y 1 Pez a
y¼ 1þ exp@ z A  sin
2 p 0 2 2 in which:
0 sp
ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ffi1
2 þ b2  a db
 
@bt  z a A y1 ¼ exp 9f2 gDrhz ðPe ! 1Þ ð48Þ
ð34Þ
2 b
Furthermore, when Bi ! 1, Equation (47) is:
with: Z
y1 1 1

  y¼ þ exp½3gDrð1  HD1 Þz  sin½bt  zðb þ 3gDrHD2 Þ
Pe 2 p 0
y1 ¼ exp d z ð35Þ db
2 ðPe ! 1; Bi ! 1 Þ ð49Þ
b
rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
Pe2
d¼ þ 9f2 PegDrh ð36Þ where y1 is the same as defined by Equation (48) in which h is
4
replaced by hBi!1 :


Pe 
a ¼ Pe þ 3gDrBiI 1 ð37Þ 1 1 þ 3jC ftanh½3ð1  jC Þf 1
4 hBi!1 ¼  ðBi ! 1 Þ ð50Þ
f 3jC f þ tanh½3ð1  jC Þf 3f
b ¼ Peðb þ 3gDrBiI 2 Þ ð38Þ
It is evidence that Equations (34), (47), and (49) will be reduced
BiðHD1 þ BiÞ to the case with a conventional catalyst particle (i.e., jC ! 0).[2]
I1 ¼ 1  ð39Þ When jC ! 1, the reaction occurs only on the outer surface for
ðHD1 þ BiÞ2 þ H2D2
non-porous catalyst. The analytical solution Equation (34) is
BiHD2 reduced to:
I2 ¼ ð40Þ
ðHD1 þ BiÞ2 þ H2D2

      

j2C ðf1 H1  f2 H2 Þ f21 þ f22 þ jC f21 þ f22 H21 þ H 22 þ f21  f22 þ ðf1 H1 þ f2 H2 Þ
HD1 ¼ 1 ð41Þ
½jC ðf1 H1 þ f2 H2 Þ þ 12 þ ½jC ðf2 H1  f1 H2 Þ2

  Z 1

j2C ðf2 H1 þ f1 H2 Þ f21 þ f22 þ 2jC f1 f2 þ ðf2 H1  f1 H2 Þ y1 1 Pez db
HD2 ¼ y¼ þ exp  az  sinðbt  bzÞ ðjC ! 1Þ ð51Þ
½jC ðf1 H1 þ f2 H2 Þ þ 12 þ ½jC ðf2 H1  f1 H2 Þ2 2 p 0 2 b
ð42Þ
in which:


sinh½2f1 ð1  jC Þ Pe pffiffiffiffiffi
H1 ¼ ð43Þ y1 ¼ exp z  z f1 ðjC ! 1Þ ð52Þ
cosh½2f1 ð1  jC Þ  cos½2f2 ð1  jC Þ 2
vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
ffi
uqffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
sin½2f2 ð1  jC Þ u 2
H2 ¼ ð44Þ t f1 þ ðPebÞ þ f1
2
cosh½2f1 ð1  jC Þ  cos½2f2 ð1  jC Þ a¼ ð53Þ
2
vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
vq
uqffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi u ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
u  2 2  b  2 u 2
t 9f þ Dr þ 9f2 t f21 þ ðPebÞ  f1
f1 ¼ ð45Þ b¼ ð54Þ
2 2

220 THE CANADIAN JOURNAL OF CHEMICAL ENGINEERING VOLUME 97, JANUARY 2019
 Pe2 jC ) in an inert-core catalyst particle as shown in Figure 2b, the
f1 ¼ þ Pef2S ð55Þ
4 transient concentration profiles of reactant inside the porous shell
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi become flat, while the reactant concentration increases at the same
Lk time compared to those in Figure 2a as results of the decrease of the
fS ¼ ð1  eb Þ ð56Þ diffusion path and less loading amount of active catalytic material.
u0
Furthermore, Figure 3 presents the averaged concentration
profiles as a function of the reactive time over the porous shell of
Furthermore, when Pe ! 1, based on Equation (51) we have: an inert-core catalyst, calculated by Equation (30) for various
( values of the shell thickness in an inert-core catalyst particle. With
0 for t < z the decrease of the shell thickness in an inert-core catalyst particle,
y¼   ðjC ! 1; Pe ! 1Þ ð57Þ the volume-averaged concentration profiles of reactant inside the
exp zf2S for t > z
porous shell shift to high values. As mentioned in Equation (31),
the volume-averaged concentration approaches the effectiveness
factor; this means that the effectiveness factor will increase with
decreasing the porous shell thickness of an inert-core particle.
RESULTS AND DISCUSSION The effects of the intraparticle diffusion resistance and the
Figures 2a and 2b show the typically transient concentration external film mass transfer resistance on the effectiveness factor of
profiles inside a conventional catalyst particle (jC ¼ 0) and the an inert-core catalyst are presented in Figures 4a–4c, where
porous shell of an inert-core catalyst (jC ¼ 0:8), respectively, effectiveness factor hb as a function of the shell thickness jC for
calculated by Equation (25) where a first-order irreversible various values of f2 at Bi ¼ 100 (a), Bi ¼ 10 (b), and Bi ¼ 1 (c), is
reaction is coupled with both the intraparticle diffusion resistance calculated by Equation (32). The Thiele modulus f stands for the
and the external film mass transfer resistance. At the beginning, ratio of the reaction rate to the intraparticle diffusion rate which is
the chemical reaction only takes place on the outer surface of the a measure of whether the process is reaction rate controlled (low f
catalyst particle; the reactant front diffuses into the catalyst value) or diffusion rate controlled (high f value); the Biot number
particle with time. Decreasing the shell thickness (i.e., increasing Bi is the ratio of the intraparticle diffusion resistance to the
external film mass transfer resistance, which is characterized by
the mass transfer limitations that include the external film mass
transfer resistance and the intraparticle diffusion resistance. In
Figure 4a where Bi ¼ 100 meaning the external film mass transfer
resistance is negligible, the effectiveness factor is decreased with
increasing f. For example f2 ¼ 0:01, the chemical reaction rate is
low resulting in the effectiveness factor hb being nearly unity for
different values of jC , i.e., the catalyst is in high utilization.
However, when f2 ¼ 100, the chemical reaction rate is fast
resulting in hb decreasing greatly with decreasing the shell
thickness jC , i.e., the catalyst is in high utilization only for high
values of jC .
When mass transfer limitations occur, the catalyst utilization
decreases significantly, for example in Figure 4c for the case at
f2 ¼ 100 and Bi ¼ 1. The effectiveness factor hb approaches zero if
jC < 0:9. Therefore, the design of an inert-core catalyst with a thin

Figure 2. Transient concentration profiles of reactant inside a Figure 3. Typical volume-averaged concentration profiles of reactant with
conventional catalyst (jC ¼ 0) and the shell of an inert-core catalyst reaction time over the shell of an inert-core catalyst, calculated by
(jC ¼ 0:8), calculated by Equation (25) for batch reactor. Equation (30) for batch reactor.

VOLUME 97, JANUARY 2019 THE CANADIAN JOURNAL OF CHEMICAL ENGINEERING 221
Figure 4. Effectiveness factor hb as a function of the shell thickness jC of an inert-core catalyst for various values of f2 at Bi ¼ 100 (a), Bi ¼ 10 (b), and Bi ¼ 1
(c), calculated by Equation (32) for batch reactor.

shell will significantly improve the catalyst utilization when both t > 2:5, the time courses of the exit concentrations approach the
intraparticle diffusion resistance and external film mass transfer steady-state.
resistance exist. The effect of f on y  t curves is shown in Figure 6 at Bi ¼ 10;
Figure 5 presents typical time courses of exit concentrations y in Pe ¼ 102 ; Dr ¼ 1:0; and jC ¼ 0:5 for f2 ¼ 0:01, f2 ¼ 0:25, and
a fixed bed reactor for first-order irreversible reaction calculated by f2 ¼ 1. For higher values of f2 , the catalytic reaction rate is fast
Equation (34) with Equations (35)–(46) at Bi ¼ 10; Pe ¼ 102 ; Dr ¼ resulting in more consumption of reactant. As expected in
1; and f2 ¼ 0:01 for jC ¼ 0; jC ¼ 0:5; and jC ¼ 0:9. We assumed Figure 6, the exit concentrations decrease with increase of f2 ,
that z ¼ 1; eb ¼ 0:5, and ep ¼ 0:5 at the following discussion. The for example, y1  0:1 for f2 ¼ 1, while y1  0:95 for f2 ¼ 0:01.
integrand of Equation (34) is the product of an exponential The exit reactant conversion X at the steady-state is defined as:
decaying function and a periodic sine function, the integration is
performed over each half-period of the sine wave.[34] When jC is X ¼ 1  ðy1 Þz¼1 ð58Þ
small, for example jC ¼ 0 as shown in Figure 5, there is
considerable broadening of mass transfer zone compared to those
Figure 7 shows the effect of jC on h, y1 , and X at the steady-state
for jC ¼ 0:5 and jC ¼ 0:9 mainly due to a greater loading amount of
for Bi ¼ 10; Pe ¼ 102 ; Dr ¼ 1 for f2 ¼ 0:01 (a), f2 ¼ 0:25 (b), and
active catalytic material and increase of the diffusion path. When
f2 ¼ 1 (c).

Figure 5. Effect of jC on y  t curves at Bi ¼ 10, Pe ¼ 102 , Dr ¼ 1, g ¼ 0:5, Figure 6. Effect of f2 on y  t curves at Bi ¼ 10, Pe ¼ 102 , Dr ¼ 1, g ¼ 0:5,
and f2 ¼ 0:01 for fixed bed reactor. and jC ¼ 0:5 for fixed bed reactor.

222 THE CANADIAN JOURNAL OF CHEMICAL ENGINEERING VOLUME 97, JANUARY 2019
 Figure 8 demonstrates the relationship among Dr, Pe, h, and f2 time. From a practical point of view, it is not possible to make both
at the steady-state for the first-order reaction for jC ¼ 0 (a), jC ¼ of them small. Therefore, the solution Equation (35) is valid to
0:5 (b), and jC ¼ 0:9 (c). The reactant conversion X increases with calculate the exit reactant concentration when t ! 1.
increasing Dr. The bed length parameter Dr is the ratio of the To sum up Figures (7) and (8), it is evident that jC is also an
residence time of the reactant in the reactor to the intraparticle adjustable design parameter, especially for the cases of mass
diffusion time of reactant into the porous support. It is a major transfer limitations, as shown in Figures (7b) and (7c); the
adjustable parameter to decide the reactant conversion X, while reactant conversions change rapidly when jC > 0:8. As expected,
the other operating conditions should be specified. there is a good example for application of inert-core catalyst in the
The exit concentration at steady-state ðy1 Þz¼1 can be derived immobilized enzyme-catalyzed kinetic resolution of racemate
from the zeroth moment based on the Danckwerts boundary where mass transfer limitations reduce the enzyme effective
conditions.[2,35] The final result is: enantioselectivity, leading to an overall decrease in optical
purity.[2,37–39] Based on previous work,[1,2,37] higher values of


1 Pe effectiveness factor favour increasing the enzyme effective
ðy1 Þz¼1 ¼ Pe  exp ð59Þ
4d þ Pe
d
sinhd þ coshd 2 enantioselectivity. However, a longer bed length is needed for
fixed bed reactor for jC ¼ 0:9 to reach the same desired conversion
as for jC ¼ 0 due to less loading amount of enzyme in the catalyst,
The difference of y1 calculated by Equations (35) and (59) have as shown in Figure 8.
a significant difference only when both Pe and Dr are small.[2,36]
Peclet number, Pe, is the ratio of backmixing time to residence

Figure 7. Effects of jC on X, h, and y 1 at Bi ¼ 10, Pe ¼ 102 , Dr ¼ 1, Figure 8. Effect of Dr on X, h, and y 1 at Bi ¼ 102 , Pe ¼ 102 , g ¼ 0:5, and
g ¼ 0:5 for f2 ¼ 0:01 (a), f2 ¼ 0:25 (b), and f2 ¼ 1 (c) for fixed bed f2 ¼ 0:01 for jC ¼ 0 (a), jC ¼ 0:5 (b), and jC ¼ 0:9 (c) for fixed bed
reactor. reactor.

VOLUME 97, JANUARY 2019 THE CANADIAN JOURNAL OF CHEMICAL ENGINEERING 223
CONCLUSIONS catalyst (¼ RC =R)
eb bed porosity (m3/m3)
Mathematical models have been developed for isothermal inert- ep support porosity (m3/m3)
core catalyst in batch and fixed bed reactors. For first-order z dimensionless axial distance (¼ Z=L)
irreversible reaction, new analytical solutions are derived from the t dimensionless contact time (¼ u0 t=Leb )
mathematical model for spherical catalyst by taking into account tb dimensionless time for batch reactor (¼ Drt ¼ De t=ep R2 ),
the intraparticle diffusion resistance and the external mass defined by Equation (27)
transfer resistance. The effects of mass transfer resistances and pffiffiffiffiffiffiffiffiffiffiffi
f Thiele modulus for spherical catalyst (¼ðR=3 Þ k=De )
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
Thiele modulus on catalytic efficiency and conversion of reactant fS dimensionless reaction parameter (¼ Lkð1  eb Þ=u0 ),
are examined over a wide range of parameters. defined by Equation (56)
The results show that the inert-core catalyst can significantly h defined by Equation (33)
improve the catalyst efficiency for fast catalytic reactions, where hb effectiveness factor, defined by Equation (32)
the catalytic efficiency is restrained by both intraparticle diffusion
resistance and external mass transfer resistance; however, the
reactant conversion will decrease due to less loading ratio of active Subscripts
species of catalyst for inert-core catalyst.
1 steady-state corresponding to t ! 1
The new general formula and its derived equations for inert-core
b batch reactor
catalyst particle are helpful to catalyst and reactor design,
operation, optimization, and control of batch and fixed bed
reactors by prediction of the effectiveness factor and the
concentration time curves. REFERENCES
[1] G. H. Xiu, L. Jiang, P. Li, Ind. Eng. Chem. Res. 2000, 39, 4054.
ACKNOWLEDGEMENTS [2] G. H. Xiu, L. Jiang, P. Li, Biotech. Bioeng. 2001, 74, 29.
This study is financially supported by the National Natural Science [3] R. G. Chaudhuri, S. Paria, Chem. Rev. 2012, 112, 2373.
Foundation of China (Grant No. 21276080, U1610102) and [4] J. M. Yan, X. B. Zhang, T. Akita, M. Haruta, Q. Xu, J. Am.
Intergovernmental Cooperation Project on Science and Technol- Chem. Soc. 2010, 132, 5326.
ogy Innovation (Grant No. 2016YFE0132500). [5] F. Wei, W. Q. Zhang, G. W. Wei, Y. Wang, J. Z. Zhang, M. C.
Zhang, L. Q. Shi, Chem. Mater. 2008, 20, 2144.
NOMENCLATURE [6] W. J. Zhou, J. Y. Lee, Electrochem. Commun. 2007, 9, 1725.
Bi Biot number (¼ kf R=De ) [7] A. J. Majewski, J. Wood, W. Bujalski, Int. J. Hydrogen Energy
cp concentration in catalyst (kg/m3) 2013, 38, 14531.
C concentration in fluid (kg/m3) [8] K. J. J. Mayrhofer, V. Juhart, K. Hartl, M. Hanzlik, M. Arenz,
DL longitudinal dispersion coefficient (m2/s) Angew. Chem. Int. Edit. 2009, 48, 3529.
2
De effective pore diffusivity
 (m /s)  [9] Y. W. Ma, H. M. Zhang, H. X. Zhong, T. Xu, H. Jin, X. Y.
Dr bed length parameter ¼ eb De L=ep R2 u0 Geng, Catal. Commun. 2010, 11, 434.
k first-order reaction rate constant (1/s) [10] B. Z. Fang, M. Kim, J. S. Yu, Appl. Catal. B-Environ. 2008, 84,
kf external mass transfer coefficient (m/s) 100.
L length of column (m) [11] Q. Zhang, I. Lee, J. B. Joo, F. Zaera, Y. D. Yin, Accounts Chem.
p Laplace operator Res. 2013, 46, 1816.
Pe Peclet number based on reactor length ð¼ u0 L=DL eb Þ
[12] X. H. Wang, B. B. He, Z. Y. Hu, Z. G. Zeng, S. Han, Sci.
r radial distance of the catalyst particle (m)
Technol. Adv. Mater. 2014, 15, 043502.
R radius of the catalyst particle (m)
RC radius of the inert core of the inert-core catalyst (m) [13] J. B. Wang, A. Varma, Chem. Eng. Sci. 1978, 33, 1549.
t time (s) [14] L. K. Doraiswamy, M. M. Sharma, Heterogeneous Reactions:
u0 the interstitial fluid velocity in the column (m/s) Analysis, Examples, and Reactor Design, Vol. 1: Gas-Solid
x dimensionless concentrations (¼ cp =C0 ) and Solid-Solid Reactions, 1st edition, John Wiley & Sons,
xav volume-averaged dimensionless concentration, defined by New York 1984.
Equation (29) [15] J. A. Schwarz, M. S. Heise, J. Colloid Interf. Sci. 1990, 135,
X reactant conversion, defined by Equation (58) 461.
y dimensionless concentrations (¼ C=C0 ) [16] A. Gavriilidis, A. Varma, M. Morbidelli, Catal. Rev.-Sci. Eng.
Z axial distance from column entrance (m) 1993, 35, 399.
[17] US. 5516740 (1996), Imperial Oil Canada, Invs.: I. A. Cody,
Greek Letters M. M. Hafez, D. N. Zinkie.
[18] M. Morbidelli, A. Gavriilidis, A. Varma, Catalyst Design:
b variable of integration
Optimal Distribution of Catalyst in Pellets, Reactors, and
g porosity ratio (¼ ep ð1  eb Þ=eb )
Membranes, 1st edition, Cambridge University Press,
d defined by Equation (36)
Cambridge 2001.
ln eigenvalue of Equation (28)
j dimensionless radial distance of the catalyst particle [19] J. M. Haugh, I. C. Schneider, Chem. Eng. Sci. 2006, 61, 5603.
(¼ r=R) [20] A. J. Majewski, J. Wood, W. Bujalski, Int. J. Hydrogen Energ.
jC dimensionless diameter of the inert-core of the inert-core 2010, 35, 11829.

224 THE CANADIAN JOURNAL OF CHEMICAL ENGINEERING VOLUME 97, JANUARY 2019
[21] K. Shams, S. Najari, Chem. Eng. J. 2011, 172, 500.
[22] M. Morbidelli, A. Varma, Chem. Eng. Sci. 1983, 38, 297.
[23] D. J. Duan, Int. J. Heat Mass Tran. 2004, 47, 2861.
[24] W. N. Marrazzo, R. L. Merson, B. J. McCoy, Biotechnol.
Bioeng. 1974, 17, 1515.
[25] J. M. Smith, Chemical Engineering Kinetics, 3rd edition,
McGraw-Hill, New York 1981.
[26] G. F. Froment, K. B. Bischoff, Chemical Reactor Analysis and
Design, 2nd edition, John Wiley & Sons, New York 1990.
[27] M. Ishida, C. Y. Wen, AIChE J. 1968, 14, 311.
[28] A. Rasmuson, Chem. Eng. Sci. 1982, 37, 411.
[29] Y. Z. Zheng, T. Y. Gu, Chem. Eng. Sci. 1996, 51, 3773.
[30] K. Shams, A. Fayazbakhsh, J. Chromatogr. A 2014, 1370, 93.
[31] S. Qamar, S. Perveen, A. Seidel-Morgenstern, Ind. Eng.
Chem. Res. 2016, 55, 2471.
[32] S. Qamar, S. Bashir, S. Perveen, A. Seidel-Morgenstern, Ind.
Eng. Chem. Res. 2017, 56, 9657.
[33] M. Goto, J. M. Smith, B. J. McCoy, Chem. Eng. Sci. 1990, 45,
443.
[34] A. Rasmuson, AIChE J. 1985, 31, 518.
[35] M. Suzuki, J. M. Smith, Chem. Eng. Sci. 1971, 26, 221.
[36] N. S. Raghavan, D. M. Ruthven, AIChE J. 1983, 29, 922.
[37] S. L. Matson, J. L. Lopez, “Multiphase Membrane Reactors
for Enzymatic Resolution: Diffusional Effects on Stereo-
selectivity,” in Frontiers in Bioprocessing, 1st edition, S. K.
Sikdar, M. Bier, P. Todd, Eds., CRC Press, Florida 1990,
pp. 391–403.
[38] M. Indlekofer, F. Brotz, A. Bauer, M. Reuss, Biotechnol.
Bioeng. 1996, 52, 459.
[39] W. R. Berendsen, A. Lapin, M. Reuss, Chem. Eng. Sci. 2007,
62, 2375.

Manuscript received November 23, 2017; revised manuscript

received February 6, 2018; accepted for publication February 7,
2018.

VOLUME 97, JANUARY 2019 THE CANADIAN JOURNAL OF CHEMICAL ENGINEERING 225