Mass Transfer Process Lecture Note PDF

Download as pdf or txt
Download as pdf or txt
You are on page 1of 170
At a glance
Powered by AI
The key takeaways are that this course covers mass transfer operations and processes that are important for separation and purification in many industrial processes. Mass transfer deals with the transportation of components between different phases.

The course will cover topics such as molecular diffusion, mass transfer models, interphase mass transfer, absorption, distillation, liquid-solid extraction, and others.

This course is important for chemical engineers because mass transfer operations make up a large portion of capital invested in chemical plants. Virtually all industrial processes involve purification or separation using mass transfer.

ChE 351

Mass Transfer Processes

E.G. Ankudey, PhD


Dept of Chemical Engineering, KNUST
September 2011

What the course is about

Mass Transfer Operations deal with unit


operations involving mass transfer (a
microscopic process on a macroscopic
scale)

Mass transfer is defined as the transportation


of one (or more) component from one phase
to another

Motivation: in many industrial processes we


use mass transfer to achieve separation
(enrichment or removal) of a substance from
a mixture

Emphasis is placed on separation processes


that involve equilibrium between the phases
2

Course Content

Importance of mass transfer


Molecular diffusion
Diffusion in binary mixtures
Mass transfer models
Interphase mass transfer
Mass transfer with laminar and
turbulent flow

Distillation
Batch distillation
Rayleigh equation
Fractionation
Binary distillation
McCabe-Thiele method
Ponchon-Savarit method

Liquid solid extraction.

Absorption in packed towers


Effect of temperature on absorption

Why is this course important?


Mass transfer operations are largely the
responsibility of chemical engineers
Chemical plants usually have from 50 to
90% of their capital invested in
separation equipment
There is virtually no industrial process
that does not involve purification of raw
materials or final separation of products

Course Objectives
The student is expected to:
Recognize the various modes of mass transfer
Estimate diffusion coefficients under different physical conditions

Determine mass transfer rates using Ficks Law for one-dimensional


molecular diffusion in gases and liquids
Use the rate equation with mass transfer coefficients to determine
mass transfer rates
Calculate film mass transfer coefficients using literature correlations
or material balance findings
Calculate overall mass transfer coefficients from film mass transfer
5
coefficients or material balance findings

Course Objectives contd


Use analogies between heat transfer and mass transfer to
solve problems
Determine liquid and gas flow rates required for
absorption and stripping columns
Calculate the packing height for gas absorption towers.
Obtain practical information including physical properties
from the literature relevant to contactor design
Use McCabe and Thiele method to analyze binary
distillation problems
Determine the number of stages and feed plate location in
simple binary distillations
Understand the concept of tray and overall efficiency and
be able to determine the height of distillation column.
6

References
Principles and Modern Applications of Mass
Transfer Operations 2nd Edition, 2009
Jaime Benitez

Mass Transfer: Fundamentals and


Applications 1985
Anthony L. Hines and Robert N. Maddox

Unit Operations Of Chemical Engineering


2004
Julian C. Smith, Peter Harriott, Warren McCabe

References
Transport Phenomena 2006
Edwin N. Lightfoot, R. Byron Bird, Warren E. Stewart

Separation Process Principles

2005

E.J. Henley, Ernest J. Henley, J. D. Seader

Mass Transfer Operations 3rd Edition 1980


Robert Ewald Treybal

Course Organisation

4 hours regular lectures per week


2 hours tutorial session per week or as needed
8-12 Homework Assignments (10-20%)
Two mid-semester exams
(10-30%)
Final Examination (50-70%)
Office hours: Open door (feel free to walk in)

TA:

Mr. Derrick Amoabeng


9

Prerequisites
What do we need?

Differential Equations (MATH 251/252)


Material/Energy Balances (ChE 251/252)
Thermodynamics (ChE 253/254)
Fluid Transport (ChE 255)
Heat Transfer (ChE 256)
Kinetics and Rate of reactions (??)

10

Introduction to Mass Transfer


Transfer of material from one homogeneous phase
to another
Driving force is concentration difference or
difference in thermodynamic activity
Results from differences in solubility, vapour pressure
or diffusivity.
Independent of density and particle size

Examples are distillation, gas absorption, extraction,


adsorption, dehumidification
A limit to mass transfer occurs when the
concentration in the two phases is the same. i.e. the
two phases come to equilibrium

11

Mass transfer operations


The transfer of mass within a fluid mixture or across a phase boundary is
a process that plays a major role in many industrial processes. Examples
of such processes are:
Dispersion of gases from stacks
Removal of pollutants from plant discharge streams by absorption

Stripping of gases from waste water


Neutron diffusion within nuclear reactors
Air conditioning
12

Mass transfer operations


Many of our day-by-day experiences also involve
mass transfer, for example:
A lump of sugar added to a cup of coffee eventually
dissolves and then eventually diffuses to make the
concentration uniform.
Water evaporates from ponds to increase the
humidity of passing-air-stream
Perfumes presents a pleasant fragrance which is
imparted throughout the surrounding atmosphere.

The mechanism of mass transfer involves both


molecular diffusion and convection.
13

Properties of Mixtures
Mass transfer always involves mixtures.
Consequently, we must account for the variation of
physical properties which normally exist in a given
system.
When a system contains three or more
components, as many industrial fluid streams do,
the problem becomes complicated very quickly.

The conventional engineering approach to


problems of multi-component system is to attempt
to reduce them to representative binary (i.e., two
component) systems.
Explanation of definitions and relations which are
often used to explain the role of components within
a mixture.

14

Concentration of Species
Concentration of species in multi-component mixture can be
expressed in many ways. For species A, mass
concentration denoted by A is defined as the mass of A, mA
per unit volume of the mixture.

mA

.. (1)

The total mass concentration density is the sum of the


total mass of the mixture in unit volume:

where i is the concentration of species i in the mixture.

15

Concentration of species
Molar concentration of, A, CA is defined as the number of
moles of A present per unit volume of the mixture.
By definition,

Number of moles

mass of A
molecular weight of A

.. (2)

Therefore from (1) & (2)

CA

nA A

V MA

... (3)
16

Concentration of Species
For ideal gas mixtures,

pA V

.. (4)

RT

[ from Ideal gas law PV = nRT]

CA

nA
pA

V
RT

where pA is the partial pressure of species A in the mixture. V is the


volume of gas, T is the absolute temperature, and R is the universal gas
constant.
The total molar concentration or molar density of the mixture is given by

C Ci
i

. (5)
17

Velocities
In a multi-component system the various species
move at different velocities
Evaluation of velocity of mixture requires the
averaging of the velocities of each species present.
If I is the velocity of species i with respect to
stationary fixed coordinates, then mass-average
velocity for a multi-component mixture defined in terms
of mass concentration is:

. (6)

18

Velocities
Similarly, molar-average velocity of the mixture * is

C i V
i

. (7)

For most engineering problems, there will be little


difference in * and and so the mass average velocity,
, will be used in most discussions.
The velocity of a particular species relative to the massaverage or molar average velocity is termed as diffusion
velocity
(i.e.) Diffusion velocity = i -
19

Mole and mass fractions


The mole fraction for liquid and solid mixture, x A ,and for gaseous
mixtures, y A, are the molar concentration of species A divided by the
molar density of the mixtures.

CA

C
CA

For solids and liquids


. (8)
For gases

The mole fraction for liquid and solid mixture, x A ,and for gaseous
mixtures, y A, are the molar concentration of species A divided by the
molar density of the mixtures.

y i 1
i

x i 1

.. (9)

Similarly, the mass fraction of component A in the mixture is;

A
wA

.. (10)
20

Example
The molar composition of a gas mixture at 273 K
and 1.5 x 10 5 Pa is:
O2
7%
CO
10%
CO 2
15%
N2
68%
Determine
the composition in weight percent
average molecular weight of the gas mixture
density of gas mixture
partial pressure of O 2.

21

Solution
Consider 1 mole of gas mixture (basis). Then
O2
= 0.07 mol
CO
= 0.10 mol
CO 2
= 0.15 mol
N2
= 0.68 mol
Molecular weight of the constituents are:
O2
= 2 * 16 = 32 g/mol
CO
= 12 + 16 = 28 g/mol
CO 2
= 12 + 2 * 16 = 44 g/mol
N2
= 2 * 14 = 28 g/mol
Weight of the constituents are: (1 mol of gas mixture)
O2
= 0.07 * 32 = 2.24 g
CO
= 0.10 * 28 = 2.80 g
CO 2
= 0.15 * 44 = 6.60 g
N2
= 0.68 * 28 = 19.04 g
Total weight of gas mixture = 2.24 + 2.80 + 6.60 + 19.04
= 30.68 g

22

Composition in weight percent:


O2

2.24

* 100 7.30%
30.68

2.80
CO
* 100 9.13%
30.68
CO 2

6.60

* 100 21.51%
30.68

N2

19.04

* 100 62.06%
30.68

23

Average molecular weight of the gas mixture


M

Weight of gas mixture


Number of moles

30.68
30.68 g mol
1

Assuming that the gas obeys ideal gas law,


PV = nRT

n
P

V
RT

n
molar density m
V
Therefore, density (or mass density) = mM
Where M is the molecular weight of the gas.

Density m

PM
1.5 * 10 5 * 30.68
M

kg m 3
RT
8314 * 273
= 2.03 kg/m 3

24

Partial pressure of O 2 = [mole fraction of O 2] * total pressure

7
* 1.5 * 10 5
100

= 0.07 * 1.5 * 10 5
= 0.105 * 10 5 Pa

25

Modes of mass transfer


Two mechanisms of mass transfer
Molecular diffusion and convective transfer
The total mass (or molar) flux of a given species in a direction
perpendicular to a stationary plane is a vector quantity and is given by
the product of the volumetric average velocity and the mass (or molar)
concentration

N A C A

. (11)

The total flux consists of the diffusive flux and the


convective flux
26

Diffusivities
Diffusivities are defined with respect to a
plane moving at the volume average
velocity v*. By definition,
*

C
i

... (12)

27

Diffusion flux
The total flux of A could be written in terms of diffusion velocity of A,
(i.e., A - ) and average velocity of mixture, , as

NA CA ( A ) CA

By definition

. (13)

C
i

Therefore, equation (13) becomes

CA
NA CA ( A ) Ci i
C i

C A (

) y A C i

.. (14)

i
28

Diffusion flux
For systems containing two components A and B,

N A C A (
C A (
N

A
A

C A (

) y A (C A A C B B )
) y A (N A N B )
A

) y

(15)

... (16)

The first term on the right hand side of this equation is


diffusional molar flux of A, and the second term is flux due to
bulk motion.
29

Flux notations
Mass flux of species i with respect to fixed coordinates

Total mass flux

ni i v i

n v
Molar flux of species i with respect to fixed coordinates

Ni c i v i
Total molar flux

N cv *
30

Flux notations
Mass diffusion flux of species i with respect to the
mass-average velocity

j i i (v i v )
Molar diffusion flux of species i with respect to molaraverage velocity

Ji ci (v i v *)
The mass flux ni is related to the molar diffusion flux by

ni j i i v j i i n
The molar flux is related to the molar diffusions flux by

Ni Ji ci v * Ji y i N

Ficks law
An empirical relation for the diffusional molar flux, first postulated by
Fick and, accordingly, often referred to as Ficks first law, defines the
diffusion of component A in an isothermal, isobaric system. For diffusion
in only the Z direction, the Ficks rate equation is

JA D

AB

dC A
dZ

. (17)

where D AB is diffusivity or diffusion coefficient for component A


diffusing through component B, and dCA / dZ is the concentration
gradient in the Z-direction.
A more general flux relation which is not restricted to isothermal,
isobaric system could be written as

J A C DA B

dy A
dZ

. (18)

using this expression, Equation (16) could be written as

C DA B

dy A
y
dZ

AN

. (19)
32

Relation among molar fluxes


For a binary system containing A
and B,

y AN

. (20)

J A N A y AN

.... (21)

J B N B y BN

. (22)

Addition of Equation (21) & (22) gives,

N A N B (y

A y B) N

(23)
33

Relation among molar fluxes


By definition N = N A + N B and y A + y B = 1.
Therefore equation (23) becomes,
JA+JB=0
J A = -J B
CD
From

AB

dy A
dyB
C D BA
dz
dZ

... (24)

............... (25)

y A + y B- = 1

dy A = - dy B
Therefore Equation (25) becomes,
D AB = D BA

.. (26)
... (27)

This leads to the conclusion that diffusivity of A in B is


equal to diffusivity of B in A.

34

Diffusivity
Ficks law proportionality, D AB, is known as mass diffusivity
(simply as diffusivity) or as the diffusion coefficient.
D AB has the dimension of L 2 / t, identical to the fundamental
dimensions of the other transport properties:

Kinematic viscosity, = ( / ) in momentum transfer


Thermal diffusivity, (= k / C ) in heat transfer.
Diffusivity is normally reported in cm2 / sec; the SI unit being
m2 / sec.
Diffusivity depends on pressure, temperature, and composition
of the system.

In literature, some values of DAB are given for a few gas, liquid,
and solid systems.
35

Table 1a: Mass Diffusivities in Gases


System
Air
Ammonia
Aniline
Benzene
Bromine
Carbon dioxide
Carbon disulfide
Chlorine
Diphenyl
Ethyl acetate
Ethanol
Ethyl ether
Iodine
Methanol
Mercury
Naphthalene
Nitrobenzene
n-Octane
Oxygen
Propyl acetate
Sulfur dioxide
Toluene
Water
Ammonia
Ethylene
Carbon dioxide
Benzene
Carbon disulfide
Ethyl acetate

T, K D AB P,m -Pa/s
273
298
298
293
273
273
273
491
273
298
293
298
298
614
303
298
298
273
315
273
298
298

2.006
0.735
0.974
0.923
1.378
0.894
1.256
1.621
0.718
1.337
0.908
0.845
1.641
4.791
0.870
0.879
0.610
1.773
0.932
1.236
0.855
2.634

293

1.793

318
318
319

0.724
0.724
0.675

System
Carbon dioxide
Ethanol
Ethyl ether
Hydrogen
Methane
Methanol
Nitrogen
Nitrous oxide
Propane
Water
Hydrogen
Ammonia
Argon
Benzene
Ethane
Methane
Oxygen
Pyridine
Nitrogen
Ammonia
Ethylene
Hydrogen
Iodine
Oxygen
Oxygen
Ammonia
Benzene
Ethylene
Water

T, K

D AB P, m -Pa/s

273
273
273
273
298.6
298
298
298
298

0.702
0.548
5.572
1.550
1.064
1.672
1.185
0.874
1.661

293
293
273
273
273
273
318

8.600
7.800
3.211
4.447
6.331
7.061
4.427

293
298
288
273
273

2.441
1.651
7.527
0.709
1.834

293
296
293
308.1

2.563
0.951
1.844
2.857

36

Table 1b: Mass Diffusivities in Liquids at Infinite Dilution


5

System

T,K

Chloroform (solvent B)

Acetone
Benzene
Ethanol
Ethyl ether
Ethyl acetate
MEK

298
288
288
298
298
298

2.35
2.51
2.20
2.13
2.02
2.13

Acetic acid
Acetone
Ethyl benzoate
MEK
Nitrobenzene
Water

293
293
293
303
293
298

2.18
3.18
1.85
2.93
2.25
3.20

1.49
2.00
1.92
1.25
2.00
1.02
1.00
1.20
2.10
0.58
0.92
1.64

298
298
298
281
298
288
298
353
303
281
298
281

2.09
1.96
1.38
1.45
2.09
2.25
2.28
4.25
2.09
1.19
1.85
1.77

Methane
Air
Carbon dioxide
Chlorine
Argon
Benzene
Ethanol
Ethane
Oxygen
Pyridine
Aniline
Ammonia

298
298
298
298
298
298
288
298
298
288
293
298

1.87

2.92
4.04
2.62
3.77
4.56

Ethylene
Allyl alcohol
Acetic acid
Benzoic acid
Propionic acid
Vinyl chloride

298

288
313
298
298
298

288
293
298
298
298

0.90
1.19
1.00
1.06
1.34

298
298

1.81
1.24

Ethylbenzene
Sulfuric acid
Nitric acid

293
298
298

0.81
1.73
2.60

Ethanol (solvent B)

Benzene
Water

Water (solvent B)

Acetone (solvent B)

Acetic acid
Acetic acid
Benzoic acid
Formic acid
Water

System T,K DAB x 10 , cm/s

Ethyl acetate (solvent B)

Benzene (solvent B)

Acetic acid
Aniline
Benzole acid
Bromobenzene
Ciclohexane
Ethanol
Formic acid
n-Heptane
MEK
Naphthalene
Toluene
Vinyl chloride

DAB x10 , cm /s

Sources: Reid, R. C., J. M. Prausnitz, and B. E. Poling,


The Properties of Gases and
Liquids,4th ed., McGraw
-Hill, New York (1987); Cussler, E.Diffusion,
L.,
Mass Transfer
in Fluid Systems,
2nd ed., Cambridge University Press, Cambridge, UK (1997).

37

Table 1c: Mass Diffusivities in the Solid State


System

Hydrogen in iron

Hydrogen in nickel
Carbon monoxide in nickel
Aluminum in copper
Uranium in tungsten
Cerium in tungsten
Yttrium in tungsten
Tin in lead
Gold in lead
Gold in silver
Antimony in silver
Zinc in aluminum
Silver in aluminum
Bismuth in lead
Cadmium in copper
Carbon in iron
Helium in silica
Hydrogen in silica
Helium in Pyrex

T, K

283
323
373
358
438
1223
1323
1123
293
2000
2000
2000
558
558
1033
293
773
323
293
293
1073
1373
293
773
473
773
293
773

1.66 x 10 -9
11.4 x 10-9
124 x 10-9
-9
11.6 x 10
-8
10.5 x 10
4.00 x 10-8
-8
14.0 x 10
-9
2.20 x 10
1.30 x l0- 30
1.30 x 10 - 11
-11
95.0 x 10
-11
1820 x 10
-10
1.60 x 10
- 10
4.60 x 10
3.60 x 10- 10
3.60 x 10- 10
-9
2.00 x 10-9
1.20 x 10
l.l0 x 10 -16
2.70 x 10 -15
1.50 x 10 -8
-8
45.0 x 10
4.00 x 10- 10
7.80 x 10 -8
6.50 x 10- 10
1.30 x 10 -8
4.50 x 10-11
2.00 x 10 -8

Sources: Barrer, R. M., Diffusion in and through Solids, Macmillan, New York
(1941); American Society for Metals, Diffusion, ASM (1973); Cussler, E. L.,
Diffusion, Mass Transfer in Fluid Systems, 2nd ed., Cambridge University Press,
Cambridge, UK (1997).

38

Diffusivities

Diffusivities of gases at low density are almost composition


independent, increase with the temperature and vary
inversely with pressure.

Liquid and solid diffusivities are strongly concentration


dependent and increase with temperature.

39

Range of diffusivity values


Gases :
Liquids :
Solids :

5 X 10 6
10 6
5 X 10 14

----------------

1 X 10-5
10-9
1 X 10-10

m2 / s.
m2 / s.
m2 / s.

In the absence of experimental data, semi theoretical


expressions have been developed which give approximation,
sometimes as valid as experimental values, due to the
difficulties encountered in experimental measurements

40

General expressions
Pressure dependence of diffusivity is given by

AB

(for moderate ranges of


pressures, up to 25 atm).

. (28)

And temperature dependency is according to

AB

. (29)

In the absence of experimental data, semi-empirical


expressions have been developed to estimate the diffusion
coefficient.
41

Binary ideal gas systems


Based on the kinetic theory of gases at moderate
pressures
Lennard-Jones potential to evaluate the influence of
intermolecular forces
The diffusion coefficient of gas pairs of nonpolar, nonreacting molecules is given by the Wilke-Lee equation

DAB

3.03 0.98 / M 10 T

1/ 2
AB
2
AB

PM 1AB2 D

42

M AB

DAB
MA, MB
T
P
AB
D

1
1
2

M A MB

= diffusion coefficient, cm2/s


= molecular weights of A and B
= temperature, K
= pressure, bar
= collision diameter, a Lennard-Jones parameter,
= diffusion collision integral, dimensionless

The collision integral is a function of the temperature and the


intermolecular potential field for one molecule of A and one molecule of
B. It is tabulated as a function of T* = T/AB, where is the Boltzmann
constant and AB is the energy of molecular interaction of the binary
system A and B
43

An approximation of D is obtained from the


following equation
D

* b

c
e
g

*
*
exp dT
exp fT
exp hT *

where T* = T/AB
a
b
c
d

=
=
=
=

1.06036
0.15610
0.19300
0.47635

e
f
g
h

=
=
=
=

1.03587
1.52996
1.76474
3.89411
44

For a binary system of nonpolar molecular pairs,


the Lennard-Jones parameters may be obtained
as follows

AB

A B
2

AB A B

The interaction parameters for the pure


components are usually obtained from viscosity
data

45

In the absence of experimental data, the values of the


parameters for pure components may be estimated from
the following empirical correlations

1.18V

13
b

1.15Tb

where Vb is the molar volume of the substance as a


liquid at its normal boiling point in cm3/gmol and Tb is the
normal boiling temperature

Vb 0.285Vc

1.048

where Vc is the critical volume


Otherwise, the atomic volume of each element present
are added together
46

47

48

49

50

Fuller method
DAB

1.00 x107T 1.75(1 / M A 1 / MB )0.5

P v A v B
1

vA = sum of structural volume increments of


component A
P = absolute pressure in atm

51

52

Multi-component diffusivities
Diffusivity of a component in a mixture of components
can be calculated using the diffusivities for the various
binary pairs involved in the mixture. The relation given
by Wilke is

D 1 mixture

1
y 3
y 2
y n

...........
D1 2
D1 3
D1 n

. (30)

Where D 1-mixture is the diffusivity for component 1 in the gas


mixture; D 1-n is the diffusivity for the binary pair, component 1
diffusing through component n; and yn is the mole fraction of
component n in the gas mixture evaluated on a component 1
free basis, that is

y 2

y2

y2
y 3 ....... y n

... (31)
53

Example
Determine the diffusivity of CO 2 (1), O 2 (2) and N 2
(3) in a gas mixture having the composition:
CO2 : 28.5 %, O2 : 15%, N 2 : 56.5%,

The gas mixture is at 273 K and 1.2 * 10 5 Pa. The


binary diffusivity values are given as: (at 273 K)
D 12 P = 1.874 m 2 Pa/s
D 13 P = 1.945 m 2 Pa/s
D 23 P = 1.834 m 2 Pa/s

54

Diffusivity of CO 2 in mixture

D 1m

y 2

where

D 12

y2

y 2

y2 y3
y3

y 3
Therefore

y3

D 1m P

y 3
D 13

0.15

0.21
0.15 0.565
0.565

0.79
0.15 0.565
1
0.21
0.79

1.874
1.945

= 1.93 m 2.Pa/s

Since P = 1.2 * 10 5 Pa,

D1m

1.93
5
2

1.61
*
10
m
s
5
1.2 * 10

55

Diffusivity of O 2 in the mixture,

D 2m

y 1
D 21

Where

y 1

y1
y1 y3

D 23

0.285

0.335
0.285 0.565

(mole fraction on-2 free basis).

y3

y3
y1 y3

0.565

0.665
0.285 0.565
56

D 21 P = D 12 P = 1.874 m 2.Pa/sec
Therefore

D 2m P

0.335
0.665

1.874
1.834

= 1.847 m 2.Pa/sec

D 2m

1.847
1.2 * 10 5

1.539 * 10 5 m 2 sec

By similar calculations Diffusivity of N 2 in the mixture can be


calculated, and is found to be, D 3m = 1.588 * 10 5 m 2/s.
57

Diffusivity in liquids
Diffusivities in liquids are nearer to 10-5 cm2 / s, and
about ten thousand times slower than those in dilute
gases.

This characteristic of liquid diffusion often limits the


overall rate of processes occurring in liquids
Diffusivity limits the rate of acid-base reactions;
Diffusion is responsible for the rates of liquid-liquid
extraction.

58

Diffusivity in liquids
Certain compounds diffuse as molecules,

Others which are designated as electrolytes ionize in


solutions and diffuse as ions.
For example, sodium chloride (NaCl), diffuses in water as ions
Na + and Cl-.
Though each ions has a different mobility, the electrical
neutrality of the solution indicates the ions must diffuse at the
same rate;
It is not possible to speak of a diffusion coefficient for
molecular electrolytes such as NaCl.
59

Diffusivity in liquids
If several ions are present, the diffusion rates of the individual
cations and anions must be considered, and molecular
diffusion coefficients have no meaning.
Diffusivity varies inversely with viscosity when the ratio of
solute to solvent ratio exceeds five.
In extremely high viscosity materials, diffusion becomes
independent of viscosity.

60

Diffusivity in solids:
An outstanding characteristic of these values is their
small size, usually thousands of times less than those in
a liquid, which are in turn 10,000 times less than those
in a gas.
Diffusion plays a major role in catalysis and is important
to the chemical engineer. The diffusion of atoms within a
solid is of prime importance to the metallurgist.

61

Steady State Diffusion


Steady-state molecular mass transfer through simple
systems in which the concentration and molar flux are
functions of a single space coordinate.

In a binary system, containing A and B, the molar flux in


the direction of z, as given by Eqn (16) is

dy A
C D AB
y
dz

(N

N B)

. (32)

62

Diffusion through a stagnant gas film


The diffusivity or diffusion coefficient for a gas can be measured
experimentally using a diffusion cell. This cell is illustrated
schematically in figure.
Gas B

z=z2

z=z1=0

Liquid A

Liquid A maintained at constant temperature and pressure

Gas B has negligible solubility in A

No reaction between A and B


63

Diffusion through a stagnant gas film


Component A vaporizes and diffuses into the gas phase
The rate of vaporization may be physically measured and may
also be mathematically expressed in terms of the molar flux.
Consider the control volume S z, where S is the cross
sectional area of the tube.
Mass balance on A over this control volume for a steady-state
operation yields

64

Mass balances
[Moles of A leaving at z + z] [Moles of A entering at z] = 0. (i.e.)

SN

A z z

SN

A z

0.

. (32)

Dividing through by the volume, SZ, and evaluating in the limit as Z


approaches zero, we obtain the differential equation

dN A
0
dz

.. (33)

This relation stipulates a constant molar flux of A throughout the gas


phase from Z1 to Z2.
A similar differential equation could also be written for component B as,

d N B
d Z

0,

.. (34)

and accordingly, the molar flux of B is also constant over the entire
diffusion path

65

Mass balances
Since the gas B is insoluble in liquid A, we realize that NB, the net
flux of B, is zero throughout the diffusion path; (alternatively, B is a
stagnant gas along the diffusion path).
From equation 19

NA

dy A
C D AB
y A (N A N B ) . (35)
dz
Since N B = 0,

NA

dy A
C D AB
y AN
dz

.. (36)

Rearranging,

NA

C D AB d y A

1 y A d z

. (37)
66

Mass balances
This equation may be integrated between the two
boundary conditions:
at z = z1
YA = yA1
And
at z = z2
YA = yA2
Assuming the diffusivity is to be independent of
concentration, and realizing that NA is constant along the
diffusion path, by integrating equation (37) we obtain

Z2

y A2

Z1

A1

d z C DAB

NA

CD
Z

AB

Z1

dy A
1 y
A
y

. (38)

1 y A2

ln
1 y A1

. (39)
67

The log mean average concentration of component B is


defined as
y B 2 y B1
. (40)
y B, lm

ln y B 2

B1

y B 1 y A

Since

y B, lm

(1 y

) (1 y

ln y B 2

y
B1

A2

A1

A1

A2

ln y B 2

y
B1

. (41)

Substituting Eqn (41) in Eqn (39),

C D AB ( y A1 y A2 )

Z 2 z1
y B, lm

. (42)
68

Diffusion flux of gases


For an ideal gas
n
p
C

V
RT

. (43)

and for mixture of ideal gases

p A

. (44)

Therefore, for an ideal gas mixture equation. (6) becomes

D AB
( p A1 p A2 )
RT ( z2 z1)
p B, lm

. (45)

This is the equation of molar flux for steady state


diffusion of one gas through a second stagnant gas
69

Applications
Many mass-transfer operations involve the diffusion
of one gas component through another non-diffusing
component; absorption and humidification are typical
operations defined by these equation.

The concentration profile (yA vs. z) for this type of diffusion is


shown in figure:

70

Example
Oxygen is diffusing in a mixture of oxygen-nitrogen at 1 std
atm, 25C. Concentration of oxygen at planes 2 mm apart
are 10 and 20 volume % respectively. Nitrogen is nondiffusing.
Derive the appropriate expression to calculate the flux
oxygen. Define units of each term clearly.
Calculate the flux of oxygen. Diffusivity of oxygen in
nitrogen = 1.89 * 10 5 m 2/s.

71

Solution
Let us denote oxygen as A and nitrogen as B. Flux of A
(i.e.) N A is made up of two components, namely that
resulting from the bulk motion of A (i.e.), Nx A and that

resulting from molecular diffusion J A:

N A Nx A J A

. (a)

72

Solution
From Ficks law of diffusion,

D AB

dCA

. (b)

dz

Substituting this equation (a)

N A Nx A D AB

dCA

. (c)

dz

Since N = N A + N B and x A = C A / C equation (c) becomes

NA NA NB

CA
C

D AB

dCA
dz

.. (d)
73

Rearranging the terms and integrating between the planes


between 1 and 2,
C A2
dC A
dz
cD AB C A1 N AC C A N A N B

. (e)

Since B is non diffusing N B = 0. Also, the total concentration C remains


constant. Therefore, equation (4) becomes
C A2
dC A
z

C A1 N C N C
CD AB
A
A
A

C C A2
1

ln
NA
C C A1
Therefore,

NA

CD AB
z

ln

. (f)

. (g)
C C A2
C C A1

. (h)
74

Replacing concentration in terms of pressures using Ideal gas law,


equation (h) becomes

NA

D AB P t
RTz

ln

P t P A2
P t P A1

. (i)

where
D AB = molecular diffusivity of A in B
P T = total pressure of system
R = universal gas constant
T = temperature of system in absolute scale
z = distance between two planes across the direction of diffusion
P A1 = partial pressure of A at plane 1, and
P A2 = partial pressure of A at plane 2

Given:
D AB = 1.89 * 10 5 m2/sec
P t = 1 atm = 1.01325 * 10 5 N/m 2
T = 25C = 273 + 25 = 298 K
z = 2 mm = 0.002 m
P A1 = 0.2 * 1 = 0.2 atm (From Ideal gas law and additive pressure rule)
75

Substituting these in equation (i)

1.89 * 10 5 1.01325 * 10 5 1 0.1


NA
ln

83142980.002

1 0.2

= 4.55 * 10 5 kmol/m 2.s

76

Equimolar counter diffusion


A physical situation which is encountered in the distillation
of two constituents whose molar latent heats of
vaporization are essentially equal

The flux of one gaseous component is equal to but acting


in the opposite direction from the other gaseous
component; that is, NA = - NB.
The molar flux NA, for a binary system at constant
temperature and pressure is described by

N
or

C D AB

dy A
y
dz

D AB

dC A
y
dz

(N A N B )

(N A N B )

. (48)
. (49)

with the substitution of NB = - NA, Equation (49) becomes,

AB

dC A
dz

. (50)

For steady state diffusion Equation. (50) may be integrated, using


the boundary conditions:
at z = z1
CA = CA1
and z = z2
CA = CA2
Giving,
Z2

NA d z D

C A2

dC

AB

Z1

C A1

from which

AB

z 2z1

(C

A1

A2 )

. (51)

For ideal gases,

nA
pA

V
RT

. Therefore Equation. (51) becomes

N
.

D AB

(P
R T (z 2 z 1)

A1

A2 )

. (52)

This is the equation of molar flux for steady-state equimolar counter


diffusion.
Concentration profile in these equimolar counter diffusion may be
obtained from,

d
(N
dz

A)

. (53)

(Since NA is constant over the diffusion path).


And from equation. (50)

AB

d CA
dz

. (54)

Therefore

d
dz

or

d CA
D AB
0
dz

CA

dz

0.

. (55)

This equation may be solved using the boundary conditions to give

C
C

A
A1

A1

A2

zz1

z1 z 2

.. (56)

Equation, (56) indicates a linear concentration profile for equimolar


counter diffusion.

Example
Methane diffuses at steady state through a tube containing helium. At
point 1 the partial pressure of methane is p A1 = 55 kPa and at point 2,
0.03 m apart, P A2 = 15 KPa. The total pressure is 101.32 kPa, and the
temperature is 298 K. At this pressure and temperature, the value of
the diffusivity is 6.75 * 10 5 m 2/s.
Calculate the flux of CH 4 at steady state for equimolar counter
diffusion.
Calculate the partial pressure at a point 0.02 m apart from point 1.

Solution
For steady state equimolar counter diffusion, molar flux is given by

NA

D AB
RT z

p A 1 p A 2

Therefore;

NA

6.75 * 10 5
55 15 kmol

8.314 * 298 * 0.03


m 2 . sec

3.633 * 10

kmol

sec

And from (1), partial pressure at 0.02 m from point 1 is:

3.633 * 10

6.75 * 10 5

55 p A
8.314 * 298 * 0.02

p A = 28.33 kPa

Example
In a gas mixture of hydrogen and oxygen, steady state
equimolar counter diffusion is occurring at a total pressure
of 100 kPa and temperature of 20C.

If the partial pressures of oxygen at two planes 0.01 m


apart, and perpendicular to the direction of diffusion are 15
kPa and 5 kPa, respectively and the mass diffusion flux of
oxygen in the mixture is 1.6 * 10 5 kmol/m 2.s, calculate the
molecular diffusivity for the system.

Solution
For equimolar counter current diffusion:

NA

D AB
RTz

p A1 p A2

where
N A = molar flux of A (1.6 * 10 5 kmol/m 2.s):
D AB = molecular diffusivity of A in B
R = Universal gas constant (8.314 kJ/kmol.k)
T = Temperature in absolute scale (273 + 20 = 293 K)
z = distance between two measurement planes 1 and 2 (0.01 m)
P A1 = partial pressure of A at plane 1 (15 kPa); and
P A2 = partial pressure of A at plane 2 (5 kPa)

Substituting these in the equation

1.6 * 10

D AB

8.3142930.01

Therefore, D AB = 3.898 * 10 5 m 2/sec

15 5

Example
A tube 1 cm in inside diameter that is 20 cm long is filled with
CO2 and H2 at a total pressure of 2 atm at 0C. The diffusion
coefficient of the CO2 H2 system under these conditions is
0.275 cm2/sec. If the partial pressure of CO2 is 1.5 atm at one
end of the tube and 0.5 atm at the other end, find the rate of
diffusion for:
steady state equimolar counter diffusion (N A = - N B)
steady state counter diffusion where N B = -0.75 N A, and
steady state diffusion of CO2 through stagnant H2 (NB = 0)

dy

N A C D AB

dz

N A N B

Given N B = - N A

N A C D AB
(For ideal gas mixture

CA

dyA
dz

D AB

dCA
dz

pA
RT

where pA is the partial pressure of A; such that p A + p B = P)


Therefore
d pA RT
N A D A B

dz

For isothermal system, T is constant


Therefore
D AB d p A
NA
RT
dz

(i.e.)

Z2

N A d z
Z1

NA

D AB
RT

P A2

d pA

P A1

p A1 p A2
RT z
D AB

where Z = Z 2 Z 1
Given: D AB = 0.275 cm 2/sec = 0.275 * 10 4 m 2 /sec ; T = 0C = 273 K

NA

0.275 * 10 4

1.5 * 1.01325 * 10 5 0.5 * 1.01325 * 10 5


8314 * 273 * 0.2
6.138 * 10

k mol
m

sec

Rate of diffusion = N A S
Where S is surface area
Therefore rate of diffusion = 6.138 * 10-6 * r 2
= 6.138 * 10 6 * (0.5 * 10 2) 2
= 4.821 * 10 10 k mol/sec
= 1.735 * 10 3 mol/hr.

ii)

C D AB

dy

dz

N A

NB

given: N B = - 0.75 N A
Therefore

N A C D AB
C D AB

dy

dz
dy

dz

N A 0.75 N A

0.25 y

N A 0.25 y A N A C D AB

N A d z C D AB
for constant N A and C

dy

dyA
dz

1 0.25 y

NA

Z2

y A2

Z1

1 0.25 y
A
y A1

N A d z CD AB

dy

dx
1
a b x b ln a b x

N A z C D AB

NA

ln 1 0.25 y A
0.25

4 CD AB
z

yy

1 0.25 y A 2

ln
1 0.25 y A 1

A2
A1

Given:

p
2 * 1.01325 * 10 5
C

0.0893 K mol m 3
RT
8314 * 273

A1

A2

p A1
P

1.5

0.75
2

p A2
P

0.5

0.25
2

Substituting these in equation (2),

NA

4 * 0.0893 * 0.275 * 10 4

0.2
7.028 * 10 6

kmol

1 0.25 * 0.25
ln 1 0.25 * 0.75

m 2 sec

Rate of diffusion = N A S = 7.028 * 10 6 * * (0.5 * 10 2) 2


= 5.52 * 10 10 kmol/sec
= 1.987 * 10 3 mol/hr.

iii)

N A CD AB

Given: N B = 0

dyA

N A CD AB

dz

dyA
dz

Z2

N A d z CD AB
Z1

y A N A NB

yA NA
y A2

dy

1 y
A
y A1

1 y A2

ln
1 y A1
Z

0.0893 * 0.275 * 10 4

0.2
CD AB

1.349 * 10
Rate of diffusion

1 0.25
ln 1 0.75

kmol
m 2 . sec

= 1.349 8 10 5 * * (0.5 * 10 2) 2
= 1.059 Kmol / sec
= 3.814 mol/hr

Practice Problem

Water in the bottom of a narrow metal tube 0.5 cm in


diameter is held at a constant temperature of 293K.
The total pressure of air is 1 atm and the temperature
is 293K. Water evaporates and diffuses through the air
in the tube and the diffusion path (z2-z1) is 0.1524 m
long. Calculate the rate of evaporation at steady state
in kmol/s. The diffusivity of water at 293K and 1 atm is
0.250x10-4 m2/s. Assume that the system is
isothermal.

93

Diffusion into an infinite standard


medium

Problems involving diffusion from a spherical particle into an


infinite body of stagnant gas.

How to set up differential equations that describe the diffusion


in these processes.

The solutions developed here for these problems actually

represent a special case of the more common situation


involving both molecular diffusion and convective mass
transfer.

Evaporation of a spherical Droplet


Consider the evaporation of spherical droplet such as a raindrop
or sublimation of naphthalene ball. The vapor formed at the
surface of the droplet is assumed to diffuse by molecular motions
into the large body of stagnant gas that surrounds the droplet.

At any moment, when the radius of the drop is r 0, the flux of water
vapor at any distance r from the center is given by

N A C D AB

dy
dr

N A N B

.. (57)

Here N B = 0 (since air is assumed to be stagnant)

N A C D AB

dy

dr

NA

Rearranging,

C D AB d y
1 y A

.. (58)

dr

The flux N A is not constant, because of the spherical geometry; it


decreases as the distance from the center of sphere increases. But
the molar flow rate at r and r + r are the same.
This could be written as,

AN A
4 r 2 N A

AN A

4 r 2 N A

r r

r r
r

.. (59)

where A = surface area of sphere at r or r + r.


Substituting for A = 4 r 2 in equation (59),
r 2NA
lim

r r

r 0

d
r
dr
Integrating,

r 2NA

N A constant

.. (60)
... (61)

From equation (61),

r 2 N A r 02 N A 0
Substituting for N A from equation (58),

r 02

C D AB d y A
r 02 N A 0
1 y A
dr
NA 0

dr
r

C D AB

dy

1 y

... (62)

Boundary condition :
At r = r 0
y A = y AS
and
At r =
y A = y A
Therefore equation (62) becomes,

r 02

NA 0

1
C D AB ln 1 y A

r r0

y A
y AS

Simplifying,

NA 0

C D AB
r0

1 y A
ln
1 y A S

... (63)

Time required for complete evaporation of the droplet may be


evaluated from making mass balance.

Moles of water diffusing moles of water leaving the droplet

unit time
unit time
4

r 02

N A0

d

dt

4
L
3
r

0
3
M A

4 r0

... (64)

d r0

MA dt

Substituting for N A0 from equation (63) in equation (64),

C D AB
r0

1 y A L d r 0

ln
1 y AS
MA dt

... (65)

Initial condition :
When t = 0

r0 = r1

Integrating equation (65) with these initial condition,


t

d t M
A
0

1
C D AB

L
MA

0
1
r0 d r0
1 y A r 1

ln
1 y A S

1
2 C D AB

r 12
1 y A

ln
1 y A S

... (66)

Equation (66) gives the total time t required for complete evaporation of
spherical droplet of initial radius r 1.

Combustion of a coal particle


The problem of combustion of spherical coal particle is similar to
evaporation of a drop with the exception that chemical reaction
(combustions) occurs at the surface of the particle. During combustion of
coal, the reaction
C + O2
CO 2 occurs.
According to this reaction for every mole of oxygen that diffuses to the
surface of coal (maximum of carbon), react with 1 mole of carbon,
releases 1 mole of carbon dioxide, which must diffuse away from this
surface. This is a case of equimolar counter diffusion of CO 2 and O 2.
Normally air (a mixture of N 2 and O 2) is used for combustion, and in this
case N 2 does not take part in the reaction, and its flux is zero. N N 0

The molar flux of O 2 could be written as

N O 2 C D O 2 gas
Where

D O 2 gas

d yO2
dr

y O 2 N O 2 N CO 2 N N 2

is the diffusivity of O 2 in the gas mixture.

... (67)

NN2 0

Since

and from stoichiometry

equation (67) becomes

N O 2 N CO 2
N O 2 C D O 2 gas

For steady state conditions,

d
r
dr

d yO2
dr

... (68)

NO2 0

Integrating,

N O 2 constant r 02 N O 2 s

... (69)

Where r 0 is the radius of coal particle at any instant,

and

N O 2 s is the flux of O 2 at the surface of the particle.

Substituting for N O ,
2
2

r C D O 2 gas

d yO2
dr

r 02 N O 2 s ... (70)

Boundary condition s:
At

r r0

y O2 y O2 s

yO2 yO2

With these boundary condition, equation (70) becomes

r 02 N A 0

r0

which yields

NO2 s

dr
r

C D O 2 gas

C D O 2 gas
r0

O2s

y O2

d yO2

y O2s

yO2

... (71)

For fast reaction of O 2 with coal, the mole fraction of O 2 at the surface of
particle is zero. (i.e.,) y
O2 s 0

And also at some distance away from the surface of the particle

y O 2 y O 2 0.21
(because air is a mixture of 21 mole % O 2 and 79 mole % N 2)
With these conditions, equation (6) becomes,

NO2s

0.21 C D O 2 gas
r0

... (72)

Example
A sphere of naphthalene having a radius of 2mm is
suspended in a large volume of shell air at 318 K and 1
atm. The surface pressure of the naphthalene can be
assumed to be 0.555 mm Hg. The D AB of naphthalene
in air at 318 K is 6.92 * 10 6 m 2/sec. Calculate the rate
of evaporation of naphthalene from the surface.

Solution
Steady state mass balance over a element of radius r and r + r
leads to

SNA

SNA

rr

where S is the surface are (= 4 r 2)


dividing (1) by Sr, and taking the limit as r approaches
zero, gives:

d r 2N A
dr

Integrating r 2 N A = constant (or) 4 r 2 N A = constant


We can assume that there is a film of naphthalene vapor /
air film around naphthalene through which molecular
diffusion occurs.
Diffusion of naphthalene vapor across this film could be
written as,

N A CD AB

dyA
dr

y A N A NB

N B = 0 (since air is assumed to be stagnant in the film)

N A CD AB

dyA

N A CD AB

N A CD AB

dr
d
dr

yA NA
yA

1 y A

d ln 1 y A
dr

W A = Rate of evaporation = 4 r 2 N A R = constant.

WA
WA

4 r 2 CD AB d ln 1 y A
dr

dr
r

4 D AB C d ln 1 y

Boundary condition:
At r = R

yA

0.555

7.303 * 10 4
760
ln (1 y A) = - 7.3 * 10 4
y A = 0 ln (1-y A) = 0

At r =
Therefore

WA

dr
r

4 D AB C

d ln 1 y A

7.3 *10 4

WA

1
r 4 D AB C ln 1 y A
R

0 7.3 *10

1
W A 0 4 D AB C 0 7.3 * 10 4
R

W A = 4 R D AB C * 7.3 * 10 4

P
1.01325 * 10 5
C

R *T
8314 * 318
= 0.0383 kmol/m 3
Therefore W A = 4 * 3.142 * 2 * 10 3 * 6.92 * 10 6 * 0.0383 *
7.3 * 10 4
= 4.863 * 10 12 kmol/s
= 1.751 * 10 5 mol/h.

Diffusion in Liquids
The equations derived for diffusion in gases equally apply to diffusion in
liquids with some modifications. Mole fraction in liquid phases is normally
written as x (in gases as y). The concentration term C is replaced by
average molar density,


M av

For steady state diffusion of A through non diffusivity B:


N A = constant , N B = 0

NA

D AB
z x BM

x A1 x A2

M av

.. (73)

where Z = Z 2 Z 1, the length of diffusion path; and

X BM

X B 2 X B1
X B2

ln

X
B 1

. (74)

For steady state equimolar counter diffusion :


N A = - N B = const

NA

D AB
Z

C A1 C A2

D AB

x A1 x A 2

Z M av

(75)

Example
Calculate the rate of diffusion of butanol at 20C under
unidirectional steady state conditions through a 0.1 cm
thick film of water when the concentrations of butanol at
the opposite sides of the film are, respectively 10% and
4% butanol by weight.

The diffusivity of butanol in water solution is 5.9 * 10 6 cm


2/sec. The densities of 10% and 4% butanol solutions at
20C may be taken as 0.971 and 0.992 g/cm3
respectively. Molecular weight of butanol (C 4 H 9 OH) is
74, and that of water 18.

Solution
For steady state unidirectional diffusion,

NA

D AB

x A1 x A2
C

z
x B, lm
where C is the average molar density.


M avg
Conversion from weight fraction the Mole fraction:

x A1
x A2

0.1 74

0.1 74

0.9 18

0.04 74

0.04 74

0.96 18

0.026
0.010

Average molecular weight at 1 & 2:

M1

M2

1
19.47 kg Kmol
0.1 74 0.9 18

18.56 kg Kmol
0.04 74 0.96 18

1 M1 2 M 2


2
M avg

0.971 19.47 0.992 18.56


2
= 0.0517 gmol / cm 3
= 51.7 kmol/m 3

x B,lm

x B 2 x B1

ln x B 2 x B1

1 x A2 1 x A1

1 x A2

ln
1 x A1

(i.e.)

x B,lm

Therefore

NA

1 0.01 1 0.026
1 0.01
ln

1 0.026
0.016
0.982
0.0163

D AB


2 M avg

x A1 x A2
x B, lm

5.9 * 10 6 * 10 4 * 51.7
0.1 * 10

4.97 * 10

0.026 0.010

kmol
m2 s

0.982

1.789

gmol
m 2 .hr .

1.789 * 74

132.4

g
m 2 . hr .

g
m 2 . hr .

Mass diffusion with homogeneous


chemical reaction
Absorption operations involve contact of a gas mixture with a
liquid and preferential dissolution of a component in the
contacting liquid.
Depending on the chemical nature of the involved molecules,
the absorption may or may not involve chemical reaction.

The following analysis illustrates the diffusion of a


component from the gas phase into the liquid phase
accompanied by a chemical reaction in the liquid phase.

Mass diffusion with homogeneous


chemical reaction
Consider a layer of absorbing medium (liquid) as shown in
diagram.

At the surface of the liquid, the composition of A is CA 0. The


thickness of the film, is so defined, that beyond this film
the concentration of A is always zero ; that is C A = 0. If
there is very little fluid motion within the film,
118

N A D AB

dCA
dz

CA
C

N A N B

.. (76)

If the concentration of A in the film, C A is assumed small, equation (76)


becomes

N A D AB

dCA
dz

... (77)

The molar flux N A changes along the diffusion path. This change is due to
the reaction that takes place in the liquid film. This changes could be
written as
d

N A r A
dz

. (78)

where rA is the rate of disappearance of A. For a first order


reaction,
k

A B

r A kC A

.. (79)

with the substitution from equation (79) and (77) in equation (78),

dCA
d
D AB
kCA 0

dz
d z
For constant Diffusivity,

D AB

CA

dz2

kCA 0

.. (80)

.. (81)

which is a second order ordinary differential equation. The


general solution to this equation is

C A C 1 cos h

D AB

z C 2 sin h

D AB

z .. (82)

The constants of this equation can be evaluated from the


boundary conditions:
at
Z=0
C A = C A0
And
at
Z=
C A = 0.

The constant C 1 is equal to C A0 , and C 2 is equal to

C A0

tan h

D AB

.. (83)

with this substitution equation (82) becomes,

C A C A 0 cos h

D AB

C A 0 sinh k
z
D AB

tan h k

D
A
B

.. (84)

This equation gives the variation of concentration of A


with z (i.e concentration profile of A in the liquid). The molar
flux at the liquid surface can be determined by differentiating
equation (84), and evaluating the derivative

dCA
dz

at z 0

Differentiating C A with respect to z,

dCA
dz

C A0

D AB

sin h k
z
D AB

C A0

D AB

tan h

cos h k
z
D AB

k

D AB

.. (85)

Substituting z = 0 in equation (85) and from equation (77),

NA

Z0

D AB C A 0

D AB


tan h
D
AB

.. (86)

For absorption with no chemical reaction, the flux of A is obtained


from equation (77) as

D AB C A 0

. (87)

which is constant throughout the film of liquid. On comparison of equation


(86) and (87), it is apparent that the term

kD
AB

tan h

D AB

shows the influence of the chemical reactions. This term is a dimensionless


quantity, is often called as Hatta Number.

Diffusion in solids
In certain unit operations of chemical engineering such as in drying or
in absorption, mass transfer takes place between a solid and a fluid
phase. If the transferred species is distributed uniformly in the solid
phase and forms a homogeneous medium, the diffusion of the species
in the solid phase is said to be structure independent. In this case
diffusivity or diffusion coefficient is direction independent.
At steady state, and for mass diffusion which is independent of
the solid matrix structure, the molar flux in the z direction is :

N A D AB

dCA
dz

constant

.. (88)

as given by Ficks law. Integrating the above equation,

NA

D AB C A1 C A2
z

.. (89)

which is similar to the expression obtained for diffusion in a stagnant


fluid with no bulk motion (i.e. N = 0).

Diffusion in process solids


In some chemical operations, such as heterogeneous
catalysis, an important factor, affecting the rate of reaction
is the diffusions of the gaseous component through a
porous solid.

The effective diffusivity in the solid is reduced below what it


could be in a free fluid, for two reasons.
tortuous nature of the path increases the distance,
which a molecule must travel to advance a given
distance in the solid.

free cross sectional area is restricted. For many


catalyst pellets, the effective diffusivity of a gaseous
component is of the order of one tenth of its value in a
free gas.

Diffusion in process solids


If the pressure is low enough and the pores are small
enough, the gas molecules will collide with the walls more
frequently than with each other.
This is known as Knudsen flow or Knudsen diffusion.
Upon hitting the wall, the molecules are momentarily
absorbed and then given off in random directions.
The gas flux is reduced by the wall collisions.
By use of the kinetic flux, the concentration gradient is
independent of pressure ; whereas the proportionality
constant for molecular diffusion in gases (i.e. Diffusivity) is
inversely proportional to pressure.
Knudsen diffusion occurs when the size of the pore is of the
order of the mean free path of the diffusing molecule.

Transient Diffusion
Transient processes, in which the concentration at a
given point varies with time, are referred to as unsteady
state processes or time dependent processes. This
variation in concentration is associated with a variation
in the mass flux. These generally fall into two
categories:
the process which is in an unsteady state only
during its initial startup
the process which is in a batch operation throughout
its operation.

Transient Diffusion
In unsteady state processes there are three variablesconcentration, time, and position
Therefore the diffusion process must be described by partial
rather than ordinary differential equations
Many solutions are for one-directional mass transfer as
defined by Ficks second law of diffusion

C A
t

D AB

2 C A
z2

... (90)

Transient Diffusion
This partial differential equation describes a physical situation in
which there is no bulkmotion contribution, and there is no
chemical reaction.
This situation is encountered when the diffusion takes place in
solids, in stationary liquids, or in system having equimolar counter
diffusion.
Due to the extremely slow rate of diffusion within liquids, the bulk
motion contribution of flux equation (i.e., y A N i) approaches the
value of zero for dilute solutions ; accordingly this system also
satisfies Ficks second law of diffusion.

Transient Diffusion
The solution to Ficks second law usually has one of the two
standard forms.
It may appear in the form of a trigonometric series which
converges for large values of time, or it may involve series
of error functions or related integrals which are most
suitable for numerical evaluation at small values of time.
These solutions are commonly obtained by using the
mathematical techniques of separation of variables or
Laplace transforms.

Convective Mass Transfer Coefficient


Diffusive mass transfer results from a concentration
gradient within a system.
In systems involving liquids or gases, however, it is
very difficult to eliminate convection from the overall
mass-transfer process.
Mass transfer by convection involves the transport of
material between a boundary surface (such as solid or
liquid surface) and a moving fluid or between two
relatively immiscible, moving fluids.

131

Convective Mass Transfer Coefficient


There are two different cases of convective mass
transfer:
Mass transfer takes place only in a single phase
either to or from a phase boundary, as in sublimation
of naphthalene (solid form) into the moving air.

Mass transfer takes place in the two contacting


phases as in extraction and absorption.

132

Convective Mass Transfer Coefficient


In the study of convective heat transfer, the heat flux is connected to
heat transfer coefficient as

Q A q h ts tm

.. (91)

The analogous situation in mass transfer is handled by an


equation of the form

N A k c C As C A

.. (92)

The driving force is the difference between the concentration at the


phase boundary, CAS and the concentration at some arbitrarily
defined point in the fluid medium, C A .
The convective mass transfer coefficient kC is a function of
geometry of the system and the velocity and properties of the fluid
similar to the heat transfer coefficient, h.
133

Significant Parameters in Convective


Mass Transfer
Dimensionless parameters are often used to correlate
convective transfer data.
Reynolds number and friction factor (momentum transfer)
Prandtl and Nusselt numbers (convective heat transfer)
The molecular diffusivities of the three transport process
(momentum, heat and mass) have been defined as:
Momentum diffusivity

Thermal diffusivity

Mass diffusivity D AB

k
Cp

.. (93)

134

Correlations
It can be shown that each of the diffusivities has the dimensions
of L2 / t, hence, a ratio of any of the two of these must be
dimensionless.
The ratio of the molecular diffusivity of momentum to the
molecular diffusivity of heat (thermal diffusivity) is designated as
the Prandtl Number

Cp
Momentum diffusivity

Pr

Thermal diffusivity

.. (94)

The analogous number in mass transfer is Schmidt number given as

Momentum diffusivity

Sc

Mass diffusivity
D AB
D AB

.. (95)
135

Correlations
The ratio of the molecular diffusivity of heat to the molecular diffusivity of
mass is designated the Lewis Number, and is given by

Thermal diffusivity

k
Le

Mass diffusivity
D AB
C p D AB

.. (96)

Lewis number is encountered in processes involving simultaneous


convective transfer of mass and energy.

136

Sherwood number
Consider the mass transfer of solute A from a solid to a fluid
flowing past the surface of the solid.
For such a case, the mass transfer between the solid surface
and the fluid may be written as

N A k c C As C A

.. (97)

137

Sherwood number
Since the mass transfer at the surface is by molecular diffusion,
the mass transfer may also be described by

N A D AB

dCA
dy

.. (98)
y 0

When the boundary concentration, CAs is constant, equation (98) may


be written as
d C A C As
.. (99)
N A D AB
dy

y 0

Equation (97) and (99) may be equated, since they define the same flux of
component A leaving the surface and entering the fluid

k c C A s C A D AB

d
C A C As
dy

.. (100)
y 0

138

Sherwood number
This relation may be rearranged into the following form:

kc
D AB

d C A C As d y

C A C A

.. (101)
y 0

Multiplying both sides of equation(100) by a characteristic length, L we


obtain the following dimensionless expression:

kc L
D AB

d C A C As d y

C A S C A L

y 0

.. (102)

The right hand side of equation (102) is the ratio of the concentration
gradient at the surface to an overall or reference concentration
gradient; accordingly, it may be considered as the ratio of molecular
mass-transport resistance to the convective mass-transport resistance
of the fluid. This ratio is generally known as the Sherwood number, Sh
and analogous to the Nusselt number Nu, in heat transfer.
139

Application of Dimensionless
Analysis to Mass Transfer
One of the methods of obtaining equations for predicting
mass-transfer coefficients is the use of dimensionless
analysis.
Dimensional analysis predicts the various dimensionless
parameters which are helpful in correlating experimental
data.
There are two important mass transfer processes, which we
shall consider, the transfer of mass into a stream flowing
under forced convection and the transfer of mass into a
phase which is moving as the result of natural convection
associated with density gradients.
140

Transfer into a stream flowing under


forced convection
Consider the transfer of mass from the walls of a circular conduit to a
fluid flowing through the conduit. The mass transfer is due to the
concentration driving force
C As C A .
The important variables, their symbols and their dimensions are listed in
the table
Variable

Symbol

Dimensions

tube diameter

fluid density

M/L3

fluid viscosity

M/Lt

fluid velocity

L/t

mass diffusivity

DAB

L2/t

mass-transfer coefficient kc

L/t

141

Dimensionless analysis
These variables include terms descriptive of the system
geometry, the flow and fluid properties and the quantity of
importance, k c.
By the Buckingham method of grouping the variables, the
number of dimensionless groups is equal to the number of
variables minus the number of fundamental dimensions.
Hence the number of dimensionless group for this problem
will be three.
With D AB, and D as the core variables, the three groups to
be formed are

142

1 DABa

Dc kc

.. (102)

2 DABd e D f v

.. (103)

.. (104)

and

g
3 DAB

Di

Substituting the dimensions for ,

1 DABa
L
1
t

M
3
L

Dc kc

L c

L

t

.. (105)
143

Equating the exponents of the fundamental dimensions on both


sides of the equation, we have
L : 0 = 2a 3b + c + 1
t : 0=a1
M: 0=b
Solving these equations,
a = 1, b = 0 and
c=1
Thus

kc D

.. (106)

D AB

which is the Sherwood number.


The other two groups could be determined in the same manner,
yielding
and

D AB

D AB

which is termed as Schmidt Number

... (107)

Sc

.. (108)
144

Dividing 2 by 3, we get

2 D

3 D AB

D AB

D

Re

.. (109)

which is the Reynolds Number


The result of the dimensional analysis of mass transfer by forced
convection in a circular conduit indicates that a correlating relation could
be of the form,

Sh Re, Sc

.. (110)

Which is analogous to the heat transfer correlation

Nu Re, Pr

.. (111)

145

Mass Transfer with natural


convection
Natural convection currents develop if there exists any
variation in density within the fluid phase.
The density variation may be due to temperature differences
or to relatively large concentration differences.
In the case of natural convection involving mass transfer from
a vertical plane wall to an adjacent fluid, the variables of
importance are listed in the table.

146

Mass Transfer with natural convection


According to Buckingham theorem, there will be three
dimensionless groups. Choosing D AB, L and as the core
variables, the groups to be formed are

1 DABa L b c k c

.. (112)

2 DABd Le f

.. (113)

g
3 DAB

.. (114)

L gA
h

147

Table
Variable

Symbol

Dimensions

characteristic length

fluid density

M/L3

fluid viscosity

M/Lt

buoyant force

gA

M/L2t2

mass diffusivity

DAB

L2/t

kc

L/t

mass-transfer coefficient

148

Solving for the dimensionless groups, we obtain

and

kc L

Sh

D AB
D AB

Sc

L3 g A

D AB

.. (115)
.. (116)
.. (117)

With the multiplication of 2 and 3, we obtain a dimensionless


parameter analogous to the Grashof number in heat transfer by natural
convection

2 3

D AB

L 3 g A

D AB

L3 g A

Gr AB

.. (118)

149

The result of the dimensional analysis of mass transfer by


natural convection indicates that a correlating relation could
be of the form,

Sh Gr AB , Sc

.. (119)

150

Analogies among Mass, Heat and


Momentum Transfer
Analogies among mass, heat and momentum transfer have their
origin either in the mathematical description of the effects or in
the physical parameters used for quantitative description.
To explore those analogies, it could be understood that the
diffusion of mass and conduction of heat obey very similar
equations. In particular, diffusion in one dimension is described
by the Ficks Law as

D AB

dCA

.. (120)

dz

Similarly, heat conduction is described by Fouriers law as

dT
q k
dz
where k is the thermal conductivity.

.. (121)
151

Mass, Heat and Momentum Transfer


The similar equation describing momentum transfer as given by
Newtons law is

d

dz

.. (122)

where is the momentum flux (or shear stress) and is the viscosity of
fluid.
It has become conventional to draw an analogy among mass, heat
and momentum transfer. Each process uses a simple law combined
with a mass or energy or momentum balance.
The analogies are useful in understanding the transfer phenomena
and as a satisfactory means for predicting behaviour of systems for
which limited quantitative data are available.
152

Mass, Heat and Momentum Transfer


The similarity among the transfer phenomena and accordingly the
existence of the analogies require that the following five conditions exist
within the system

The physical properties are constant

There is no mass or energy produced within the system. This


implies that there is no chemical reaction within the system

There is no emission or absorption of radiant energy.

There is no viscous dissipation of energy.

The velocity profile is not affected by the mass transfer. This


implies there should be a low rate of mass transfer.
153

Reynolds Analogy
The first recognition of the analogous behaviour of mass,
heat and momentum transfer was reported by Osborne
Reynolds in 1874. Although his analogy is limited in
application, it served as the base for seeking better
analogies.

Reynolds postulated that the mechanisms for transfer of


momentum, energy and mass are identical. Accordingly,

kc

Cp

.. (123)
154

Reynolds Analogy
Here h is heat transfer coefficient
f is friction factor
is velocity of free stream

The Reynolds analogy is interesting because it suggests


a very simple relation between different transport
phenomena.
This relation is found to be accurate when Prandtl and
Schmidt numbers are equal to one.

This is applicable for mass transfer by means of turbulent


eddies in gases. In this situation, we can estimate mass
transfer coefficients from heat transfer coefficients or
from friction factors.
155

Chilton-Colburn analogies
Chilton and Colburn, using experimental data, sought
modifications to the Reynolds analogy that would not have the
restrictions that Prandtl and Schmidt numbers must be equal to
one. They defined a j factor for mass transfer as

jD

kc

Sc 2 3

.. (124)

The analogous j factor for heat transfer is

j H St Pr
where St is Stanton number =

23

Nu
h

Re Pr v C p

.. (125)
.. (126)

Based on data collected in both laminar and turbulent flow


regimes, they found

jD jH

.. (127)
156

This analogy is valid for gases and liquids within the range of 0.6 < Sc
< 2500 and 0.6 < Pr < 100.
The Chilton-Colburn analogy has been observed to hold for many
different geometries for example, flow over flat plates, flow in pipes,
and flow around cylinders.

157

Example
A stream of air at 100 kPa pressure and 300 K is flowing on the top
surface of a thin flat sheet of solid naphthalene of length 0.2 m with a
velocity of 20 m/sec. The other data are:
Mass diffusivity of naphthalene vapor in air = 6 * 10 6 m 2/sec
Kinematic viscosity of air = 1.5 * 10 5 m 2.sc
Concentration of naphthalene at the air-solid naphthalene
interface = 1 * 10 5 kmol/m3
Calculate:

the average mass transfer coefficient over the flat plate

the rate of loss of naphthalene from the surface per unit width
Note: For heat transfer over a flat plate, convective heat transfer
coefficient for laminar flow can be calculated by the equation.

Nu 0.664 Re1L 2 Pr 1 3
You may use analogy between mass and heat transfer.

158

Solution
Given: Correlation for heat transfer

Nu 0.664 Re1L 2 Pr 1 3
The analogous relation for mass transfer is

Sh 0.664 Re1L 2 Sc 1 3
where
Sh = Sherwood number = kL/D AB
Re L = Reynolds number = L/
Sc = Schmidt number = / ( D AB)
k = overall mass transfer coefficient
L = length of sheet
D AB = diffusivity of A in B
= velocity of air
= viscosity of air
= density of air, and
/ = kinematic viscosity of air.

159

Substituting for the known quantities in equation (1)

k 0.2
6 * 10 6

0.220
0.664

5
1.5 * 10

12

1.5 * 10 5

6 * 10 6

13

k = 0.014 m/s
Rate of loss of naphthalene = k (C Ai C A)
= 0.014 (1 * 10 5 0)
= 1.4024 * 10 7 kmol/m 2 s
Rate of loss per meter width = (1.4024 * 10 7) (0.2)
= 2.8048 * 10 8 kmol/m.s
= 0.101 gmol/m.hr.
160

Convective Mass Transfer Correlations


Extensive data have been obtained for the transfer of
mass between a moving fluid and certain shapes, such
as flat plates, spheres and cylinders.
The techniques include sublimation of a solid,
vapourization of a liquid into a moving stream of air and
the dissolution of a solid into water.
These data have been correlated in terms of
dimensionless parameters and the equations obtained
are used to estimate the mass transfer coefficients in
other moving fluids and geometrically similar surfaces.

161

Flat Plate
From the experimental measurements of rate of evaporation from a liquid
surface or from the sublimation rate of a volatile solid surface into a
controlled air-stream, several correlations are available. These correlation
have been found to satisfy the equations obtained by theoretical analysis
on boundary layers,

Sh 0.664 ReL1 2Sc 1 3 laminar Re L 3 * 10 5 .. (128)


Sh 0.036 ReL0.8 Sc 1 3 turbulent Re L 3 * 10 5 .. (129)
Using the definition of j factor for mass transfer in equation (128) and
(129) we obtain

j D 0.664 ReL 1 2 laminar Re L 3 * 10 5 .. (130)


J D 0.037 ReL 0.2 turbulent Re L 3 * 10 5 .. (131)
These equations may be used if the Schmidt number is in the
162
range 0.6 < Sc < 2500.

Single Sphere
Correlations for mass transfer from single spheres are
represented as addition of terms representing transfer by purely
molecular diffusion and transfer by forced convection, in the form

Sh Sh o C Re m Sc

.. (132)

Where C, m and n are constants, the value of n is normally taken


as 1/3
For very low Reynolds number, the Sherwood number should
approach a value of 2. Therefore the generalized equation
becomes

Sh 2 C Re

Sc

1 3 .. (133)

163

Single Sphere
For mass transfer into liquid streams, the equation given by Brain and
Hales

Sh 4 1.21

2 3 12
Pe AB

.. (134)

correlates the data that are obtained when the mass transfer
Peclet number, Pe AB is less than 10,000. This Peclet number is
equal to the product of Reynolds and Schmidt numbers (i.e.)

Pe AB Re Sc

.. (135)

For Peclet numbers greater than 10,000, the relation given by


Levich is useful

Sh 1.01 Pe1AB3

.. (136)
164

Single Sphere
The relation given by Froessling

Sh 2 0.552 Re 1 2 Sc 1 3

.. (137)

correlates the data for mass transfer into gases Reynolds numbers
ranging from 2 to 800 and Schmidt number ranging 0.6 to 2.7.
For natural convection mass transfer the relation given by Schultz

Sh 2 0.59 Gr AB Sc

1 4

.. (138)

is useful over the range


2 * 10 8 < Gr AB Sc < 1.5 * 10 10

165

Example
The mass flux from a 5 cm diameter naphthalene ball placed
in stagnant air at 40C and atmospheric pressure, is 1.47 * 10
3 mol/m 2. sec. Assume the vapor pressure of naphthalene
to be 0.15 atm at 40C and negligible bulk concentration of
naphthalene in air. If air starts blowing across the surface of
naphthalene ball at 3 m/s by what factor will the mass transfer
rate increase, all other conditions remaining the same?
For spheres :

Sh = 2.0 + 0.6 (Re) 0.5 (Sc)0.33


Where Sh is the Sherwood number and Sc is the Schmids
number. The viscosity and density of air are 1.8 * 10 5
kg/m.s and 1.123 kg/m 3, respectively and the gas constant is
82.06 cm 3 . atm/mol.K.
166

Flow Through Pipes


Mass transfer from the inner wall of a tube to a moving fluid has
been studied extensively. Gilliland and Sherwood, based on the
study of rate of vapourisation of nine different liquids into air given
the correlation

Sh

p B, l m
P

0.023 Re

0.83

Sc

0.44

.. (140)

Where p B, lm is the log mean composition of the carrier gas,


evaluated between the surface and bulk stream composition. P is
the total pressure. This expression has been found to be valid over
the range
2000 < Re < 35000
0.6 < Sc < 2.5
167

Linton and Sherwood modified the above relation making it suitable for
large ranges of Schmidt number. Their relation is given as

Sh 0.023 Re 0.83 Sc 1 3

.. (141)

and found to be valid for


2000 < Re < 70000
and
1000 < Sc < 2260

Example
A solid disc of benzoic acid 3 cm in diameter is spun at 20 rpm and
25C. Calculate the rate of dissolution in a large volume of water.
Diffusivity of benzoic acid in water is 1.0 * 10 5 cm 2/sec, and
solubility is 0.003 g/cc. The following mass transfer correlation is
applicable:
Sh = 0.62 Re Sc 1/3
Where
D2
and is the angular
Re
speed in radians/time 168

Solution

.. (*1)
Dissolution rate = N A S
Where N A = mass flux, and
S = surface area for mass transfer
N A = k c (C As C A )
.. (*2)
Where C As is the concentration of benzoic acid in water at the surface
of the dose.
C A is the concentration of benzoic acid in water for an infinite
distance from the surface of the disc.
Given:
Sh = 0.62 Re Sc 1/3

(i.e.)
kc D
D AB

D 2
0.62

1
2

1
3

D AB

.. (*3)

20

* 2 radian sec
60
1 rotation = 2 radian
Therefore 20 rotation per minute = 20 * 2 radian/min

169

For water = 1 g/cm 3 = 1 centipoise = 0.01 g/cm.sec.


From (*3),
1
1
3
2

k c 0.62 D AB
D AB

0.62 * 1.0 * 10

40 60 * 1
*

0.01

1
2

0.01

5
1 * 1.0 * 10

1
3

= 8.973 * 10 4 cm/sec.
From (*2),
N A = 8.973 * 10 4 (0.003 0)
= 2.692 * 10 6 g/cm 2.sec
From (*1),
N A S = N- A * (2r 2)
= 2.692 * 10 6 * (2 * 1.5 2)
= 3.805 * 10 5 g/sec
= 0.137 g/hr.

170

You might also like