Gas Phase Steady State Molecular

Diffusion
Prof. Sampa Chakrabarti
Department of Chemical Engineering
University of Calcutta
 • Steady state – For a system having a steady state, the variables and parameters do not
change with time.
• Steady state and Equilibrium state - In case of ‘Equilibrium’ there is no net transfer of
species from one point to the other whereas in ‘Steady State’ there is a transfer which
does not vary with time.
• Constant area

1 a1 a1 2

L
Variable area
Diffusion of A through Non-diffusing B
Mixture of air and ammonia is bubbled through water.
Ammonia is soluble in water, air is not. So molecules of
ammonia diffuses through the non-diffusing molecules
of air towards water.

1 2
p A1 p A2
Since B is non-diffusing, NB=0
Putting NB=0 in the above equation,
L

DIY
Self Assessment Problem
• Water evaporates from the surface of a pond and the vapour diffuses
through the stagnant film of air of thickness 3 mm. Water
temperature is 23oC and the air temperature is 27oC. Relative
humidity of air is 65%. Average diffusivity is 0.2595 cm2/s. What is
the average flux of evaporation?
Antoine equation constants A=13.8573 B=5160.2
Ans: 1.79×10-6 kmol/m2.s or 0.1159 kg/m2.h
Air

3mm air film

Water
 Equimolar counter diffusion
The number of moles diffusing are the same for both the
components, but the sign is different i.e. the direction is
opposite to one another. If NA stands for the number of
moles of A diffusing per unit area, per unit time and NB for B,
then for equimolar counter diffusion, NA = – NB or NB
= – NA . i.e. NA + NB = 0
C + O2 → CO2
When a particle of carbon burns in air, an air
film surrounds the carbon particle through
which oxygen diffuses to the carbon particle.
CO2 is produced on the surface and it diffuses
back to air. Stoichiometrically, one mole of O2
moves forward and one mole of CO2 moves
backward – this is a case of equimolar
counter diffusion.
Non-equimolar counter diffusion

When a carbon particle is burnt in limited supply of

Now, If we denote O2 by B and CO by
O2, then the following reaction takes place,
A, then,
2C + O2 = 2CO
Here 1 mole O2, diffuses to the carbon particle, and 2 What would be the expression for NA ?
moles of CO diffuses back in the opposite direction. If
we denote O2 by A and CO by B, then
NA = – 2NB i.e. NB = – ½ NA
(10 minutes) DIY
Molecular diffusion through variable area - spherical
surface
Suppose a sphere of a volatile solid (naphthalene or
camphor) is placed in air. Initially, the radius was r1 and
the area through which diffusion occurred was 4π r12.
After sometime, due to evaporation, the sphere
decreased in size and the radius becomes r2 and
eventually the area becomes 4π r22. Hence diffusion
occurs through variable area. This can happen for any
solid shape of the substance.

If the diffusional flux is NA, then the rate = NA × area = N A × 4π r2

As steady state,

4π r2 NA = const = W (say)
 Variable area – spherical surface - continued
For radial diffusion,

NB = 0 Since it is a case of diffusion of A through non-diffusing B

Case-I: when r2 is very large at infinity, pA2 = 0 there

Mole/time

Mass= Mole=
Integrating between p A1 and p A2 at r1 and
r2 respectively Now, the rate of change of moles in the sphere
 Variable area – spherical surface - continued
Therefore, equating the rate of diffusions Case-II: A finite film of thickness δ (say) surrounds the
sphere and beyond that film there is no A
0 at r2

Initially at t = 0, radius of the sphere was r1

After a time t’, the radius has decreased to r’.
r2 r1+ẟ

Knowing the diffusion – time, we can calculate the final

radius and vice-versa
Self Assessment Exercise
• Calculate the time required for 80% sublimation of a sphere of
naphthalene (diameter : 15mm) in air if a 4mm thick stagnant air film
surrounds the sphere.
• Given: Temperature is 35oC and pressure is 1 atm.
Diffusivity under the given conditions is 0.0892 cm2/s,
Density of solid naphthalene is 1140 kg/m3 and mol wt 128
Vapour pr. At 35oC is 0.4303 mm Hg
Gas constant 82.1 [Link]-1

Ans: 91.7 hours

 Variable area- tapered tube
There are two gas-reservoirs ~ (1) and
(2). These two reservoirs are connected
B by a tapered tube. Species A is diffusing
from end (1) to end (2) while species B is
diffusing from end (2) to end (1).
Area through which diffusion is
occurring, is continuously changing from
At steady state, πr12 to πr22 as the length of diffusion
Rate of diffusion (i.e. flux × area) is constant, increases from 0 at end (1) to L at end
(2).
 r 2 NA z   r 2 NA =0
z+z
This is therefore a case of equimolar
Then, counter diffusion through variable area
under steady state.
 The radius of the tube at end (1) is r1 and that Tapered tube
at end (2) is r2. Radius at any distance z from
end (1) is r. At steady state,
Rate of diffusion (i.e. flux × area) is constant,

A B For equimolar counter – diffusion

Here, as described, z is a function of r and vice-versa.

C
From Figure in Δ ABC,

DAB r2  r1 dp A
Rate  Flux  Area    r 2  W (const )
RT L dr
Self Assessment Exercise

• Gases A and B are diffusing towards each other at steady state by

equimolar counter diffusion through a conduit of length L. The partial
pressures of A at the left end is pA1 and at the right end is pA2. The cross
section of the conduit is square, each side of which being 2b1 at the left
and uniformly increasing to 2b2 at the right end. Assuming the
diffusivity to be DAB, derive the formula for the rate of diffusional mass
transfer.
 Experimental determination of gas phase diffusivity- Stefan’s tube method
A volatile liquid A is taken in a tube of say 5 mm
diameter upto a level z1. It is connected to a pipe as in
the figure. Gas B is blown through this pipe. After
sometime, say t min, the volatile liquid will evaporate
and diffuse out through pipe. The level of the liquid is
now z2.
Now let the cross-sectional area of the tube is ‘a’.
The number of moles diffusing in dt time is
With flux NA, number of moles diffusing in
dt time through a area is

=
Now let us assume that in a differential time dt, a level
change of dz is observed. Since the time and level
change by a very small quantity, we assume there is no
change at all -- This is an assumption of pseudo – steady
state. In other words, the change in level (i.e. the rate of
diffusion) is so slow a process that an instantaneous
steady state is achieved.
 Experimental determination of gas phase diffusivity- Twin Bulb Method

At bulb (1), at t = 0, there is no B, At bulb (2), at t = 0, there is no A,

so pA2 = 0, pB2 = P
so pA1 = P, pB1 = 0

At t = t’, there are some A, so

At t = t’, there are some B, so

There are two gas – bulbs – (1) and (2).

Initially (1) is full with gas (A) and (2) is
filled with gas (B). A stopper between two
bulbs was closed

Then the valve on the connecting tube There is pseudo – steady state.
between the two bulbs is opened. Both the If a is the inner cross sectional area of the connecting
gases diffuse to each other and there are tube, then,
binary mixture of A and B in both the bulbs. Rate of diffusion = area × flux
This is a case of equimolar counter diffusion = a . NA
Twin Bulb Method
a . NA

After allowing the gases to diffuse for t’

seconds, concentrations of A in the two
bulbs are measured. Now all other
quantities except DAB are known. So DAB
can be calculated.
 Correlations
Gilliland Equation
Chapman - Enskog Correlation

molar vol. at normal b.p. m3/[Link]

τ AB = collision diameter, m
ΩD = Collision integral based on Lennard
Jones potential. It is a ratio giving the Fuller Equation
deviation of a gas with interactions compared
to a gas of rigid, elastic spheres. This value
would be 1.0 for a gas with no interactions.

Σ υ are the sum of the atomic

volumes of all elements for each
molecule.
 Gas phase multicomponent diffusion
An effective diffusion coefficient in multi-
component gas system can be predicted
by extending Stefan – Maxwell equation.

The method assumes that in a non-

uniform binary mixture of A and B in
which A is diffusing at steady state, the
difference in the partial pressure of A
(ΔpA) over a small distance Δz is
proportional to
(i) Molar concentration of A and B
(ii) Length of Diffusion path.
(iii) Relative velocity is difference in
velocities of A and B.
We may assume that all components except i are non-diffusing i.e. Nj = 0, j ≠ i.

-ve sign as ΔpA decreases with increase in Δz