Heterogeneous Reaction: Shell and core model aka shrinking core model aka un-reacted core

model.

Problem description (follows the notation used in the book CRE by Octave Levenspiel)
Shell and core model: Consider the case of a spherical particle over which a fluid is flowing. The
fluid (A) and the solid (B) react to form products C and D, where C is a fluid and D is solid. D is
also called ash and it is porous. We assume that D is similar to B in terms of physical properties,
so that the overall particle size and physical structure remain the same.

Concentration of A in the
bulk gas phase = CAg
Fluid film (diffusion)

Ash (Shell). Porous

medium
Un-reacted solid
(Core)

Concentration of A on the
core surface = CAs

We use the following notation. The particle radius is R, and it is unchanging. Here particle
means the shell and core taken together. The density of the solid B (core) is ρB. We don’t need
the density of the shell in these calculations. The concentration of the fluid A in the bulk gas
phase is CAg. The concentration of A on the surface of the core is CAs. (Note, it is not the
concentration of A on the surface of the shell). The diffusivity of A through the gas film is given
by D and the effective diffusivity of A through the ash (shell, porous solid) is given by De.

We assume that the reaction is elementary (i.e. A + B
C+D is first order in A and that the
activity of the solid B is 1). The reaction rate constant is given by ks. Note that the units of ks are
length/time and that the units of the surface reaction rate are moles/area/time. Here, area refers
to the surface area of B that is available for the reaction.

The gas does NOT diffuse into the core part, i.e. B is either non-porous, or the reaction is so fast
that as soon as A sees the surface of the core, it reacts and forms the products. The gas can
diffusive through the ash-shell, because the shell is porous.

There are five steps involved in the reaction. 1. Diffusion of A from bulk phase to the shell
surface (i.e. diffusion through the gas film), 2. Diffusion of A from the shell surface to the core
surface (i.e. diffusion through the shell), 3. Reaction on the surface of the core to form C and ash
(shell), 4. Diffusion of the product C to from the core surface to the shell surface (i.e. diffusion
through the shell) and finally 5. diffusion of the product C from the shell surface to bulk gas (i.e.
diffusion through the gas film)

The 2nd and 4th steps are similar. Likewise, the 1st and 5th step are similar. Here, we want to know
if only one step is rate limiting, what will be the trend of conversion vs time? How will it vary
with the particle diameter? What can we do to speed up (or slow down) the reaction, if the first
step is rate limiting? Or if the second step is rate limiting?

Since 2nd and 4th step are similar and 1st and 5th step are similar, we analyze only the three cases.
The rate limiting step is (a) 1. Diffusion of A through the gas film, or (b) 2. Diffusion of A
through the shell and (c) 3. Surface reaction.

Determine the rate of change of core radius (rc) with time.

Solution:

(a) Diffusion through gas film is rate limiting.

We will use mass transfer coefficient kg instead of diffusivity D and boundary layer thickness(δ).
For a given fluid mixture (containing perhaps A, C and inert) at a given temperature, pressure,
velocity etc, the mass transfer coefficient is fixed.

At (pseudo) steady state conditions,

Diffusion of A through gas film = Diffusion of A through porous shell = reaction of A on the
core surface.
De
k g ( C Ag − C A− shell − surface ) 4π R 2 = ( C A− shell − surface − C As ) 4π = ks C As 4π rc2
1 1
 − 
 rc R 
* Note: The derivation of the equation for diffusion of A through the porous shell will be given
later. Right now, we will take it for granted.
* If rc is very close to R, we can use the following approximation.
De De
1 1
( C A− shell − surface − C As ) 4π ≃
( R − rc )
( C A− shell − surface − C As ) 4π R 2 . However, right now, it is
 − 
 rc R 
not necessary to employ this approximation

Now, we assume that De and ks are very large, but kg is small. Therefore, the concentrations of A
on the shell surface CA-shell-surface and on the core surface CAs are more or less the same and close
to zero.
The net reaction rate = net diffusion of A through gas film =
k g ( C Ag − C A− shell − surface ) 4π R 2 ≃ k g C Ag 4π R 2

Note that this is a constant. When diffusion through gas film is rate controlling, the overall
reaction rate is a constant.i.e. the quantity of B consumed per unit time is a constant.

This is also the rate of consumption of B. i.e. negative of the rate of formation of B.

The rate of formation of B is given by = =

4
dN B d ( ρ BVB ) ρ B d 3 π rc
3
( dr
= ρ B 4π rc2 c
)
dt dt dt dt

Therefore,
dr
− ρ B 4π rc2 c = k g C Ag 4π R 2
dt
− ρ B rc3
t= + constant
3k g C Ag R 2

ρB R  
3
ρB R r 
At t=0, rc = R. This shows that constant= . Therefore, t = 1 −  c  
3k g C Ag 3k g C Ag  R 

ρB R
The total time taken for complete reaction (i.e. for rc = 0), is denoted by τ. τ = Therefore
3k g C Ag
t   rc 3 
= 1 −  This gives the rate of change of core radius with time when the diffusion
τ   R  
through gas film is rate limiting.

Note that the conversion of the solid is given by XB. We can calculate it as

4 π r3 3
volume of unreacted core 3 c  rc 
1− X B = = = 
original volume of the unreacted particle 4 π R 3  R 
3

t
Therefore, = XB
τ

The reaction rate is a constant and thus the conversion is a linear function of time. After a time τ
the conversion will be more than 1 (as per the formula), but there is no un-reacted solid B present
after that time τ and hence conversion is meaningless at that stage. This is essentially a zero
order reaction wrt B.
(b) What if the diffusion through pore is rate limiting?

First we see a qualitative description. In the previous case, the gas film thickness (boundary layer
thickness) does not change with time. So, we were able to use the mass transfer coefficient kg
instead of D and δ. In the present case, the core radius (rc) changes with time. As time progresses,
the shell thickness changes. Therefore, the gas has to diffuse through a thicker film (thickness of
the shell = R-rc). Here, we have to use the effective diffusivity De and the shell thickness. We
can not get away with using a mass transfer coefficient.

Initially the shell thickness will be zero, so the overall reaction will be fast. At later times, the
shell thickness will be more and hence the overall reaction rate will slow down. Therefore, we
can expect that the conversion will not be a linear function of time. A plot of XB vs time will
show that initially it will rise quickly with time and later it will flatten and slowly come to 1.

Here, we assume that D and ks are very large and De is small. Consider a given moment when
there is partial conversion, i.e. there is a shell and a core. We will assume that the total amount of
A diffusing into the particle (per unit time), on the shell surface is the same as the total amount of
A reacting (per unit time) on the core surface and at any intermediate location (rc ≤ r ≤ R), the
total quantity of A diffusing (per unit time) is the same.
− dN A ∂C A
i.e. = De 4π r 2 = constant
dt ∂r

This means
 − dN A  ∂r
4π De ∂C A =   2
 dt  r
 − dN A   1 1 
4π De ( C A− shell − surface − C As ) =   − 
 dt   rc R 

Therefore,
 − dN A  4π De ( C A− shell − surface − C As )
 =
 dt  1 1
 − 
 rc R 

*Note: We used this expression in the earlier case of ‘rate limiting step is diffusion through gas
film’.

Till now, we have assumed that the rate of diffusion of A into the particle does not change. Now
we will say that it changes with time.

Since surface reaction and diffusion through gas film are very fast, the diffusion through the
porous shell is the rate limiting step. This means CAs is approximately zero, and CAg = CA-shell-
surface.
4π De ( C A− shell − surface − C As ) 4π DeC Ag
≃
1 1 1 1
 −   − 
 rc R   rc R 

Rate of consumption of A = Rate of consumption of B

4π DeC Ag − dN B dr
= = − ρ B 4π rc2 c
1 1 dt dt
 − 
 rc R 

Therefore,
 rc2  − DeC Ag
 rc −  drc = dt
 R ρB

rc = rc  rc2  − De C Ag t =t
∫rc = R
 c

r −
R
 drc =
ρB ∫ t =0
dt

Integrate and apply the limits to get

 rc2 rc3 R 2 R 3  − DeC Ag
 − − + = t
 2 3R 2 3R  ρB

On the LHS, multiply and divide by R2 /6 to get a more elegant form

 rc2 rc3  R 2 − DeC Ag

 3 2 − 2 3 − 1 = t
 R R  6 ρB
ρB R2   rc  
2 3
 rc 
i.e. t = 1 − 3   + 2   
6 DeC Ag  R R 

ρB R2
Time taken for complete conversion is τ =
6 DeC Ag


t r 
2
r  
3

Therefore, = 1 − 3  c  + 2  c  
τ  R  R  

3
r 
Note that 1 − X B =  c  and hence the above equation can also be written in terms of XB.
R
We get this expression by assuming that at any given time, rate of diffusion of A in the shell is
independent of the location (rc ≤ r ≤ R), and then by using the mass balance (i.e rate of
consumption of A = rate of consumption of B).

(c) Surface reaction is rate limiting.

In this case, the diffusion through gas film and through the pore are very fast and only the
surface reaction is slow. Therefore, CAg = CAs.

− dN A −dN B dr
The rate of consumption of A is given by = 4π rc2 ks C Ag = = 4π rc2 ρ B c
dt dt dt

Therefore,
drc k s C Ag
= , and the initial condition is at t = 0, rc = R
dt ρB

( R − rc ) = B 1 − c 
ρB ρ R r
t=
k s C Ag k s C Ag  R 

3
ρB R r 
Noting that the time for complete conversion is τ = and 1 − X B =  c  we can write
ks C Ag R
 r 
( 1
t = τ 1 − c  = τ 1 − (1 − X B ) 3
 R
)
drc
When reaction is rate limiting, rate of change of radius is a constant.
dt

Note:

1. The relationship between the particle radius (R) and the time for complete conversion (τ) is
given by
(a) when diffusion through gas film is rate limiting , or when surface reaction is rate limiting,
τα R
(b) when diffusion through porous shell is rate limiting, it is τ α R 2
Thus by varying the particle size and measuring the time for complete conversion, we can
identify ‘shell diffusion’ vs ‘other’.

2. An increase in gas flow velocity will change the mass transfer coefficient, but it will not affect
the effective diffusivity or surface reaction rate.

3. An increase in temperature will cause the surface reaction rate to increase dramatically, but
will increase the diffusivities to a lesser extent. If the overall reaction rate increases dramatically
with temperature, then the rate limiting step is surface reaction.
In case all the three steps contribute equally to the net rate, we can write
dr C Ag driving force
− ρ B 4π rc2 c = ≡ ,
dt 1 1 1 resistance
+ +
k g 4π R 2 4π De ks 4π rc2
1 1
 − 
 rc R 
which simplifies to
C Ag
dr ρB
− c = 2
dt rc
+
( R − rc ) rc + 1
2
kg R RDe ks