CHE 304

Chapter 22. Fundamentals of Mass

Transfer
Sagheer Onaizi
 DISCLAIMER!

 The materials presented in the handouts are

only brief summaries of the text materials
 The material covered in these handouts can
be incomplete
 Students will be examined according to
the materials presented in the textbook
and handouts

2
 Transport Phenomena
Transport Phenomena Driving Force

 Momentum Transfer  Velocity/Pressure Gradient

 Heat Transfer  Temperature Gradient

 Mass Transfer  Concentration Gradient (of

a given species in a
mixture)

3
 Definition

 Mass Transfer: Natural tendency to transfer a

given component in a mixture from a region of
high concentration (the source) to a region of
low concentration (the sink) in order to establish
a uniform equilibrium condition
Examples:
 Water evaporation from wet clothes
 Sugar dissolving in coffee AND diffusing uniformly
through the coffee
 Perfume diffusing to the surrounding atmosphere
4
 Modes of Mass Transfer
1. Molecular Mass Transfer (Diffusion): Mass transfer
through a quiescent or unstirred medium (fluid or
liquid) due to random molecular motion; Analogous
to heat transfer by conduction.
[Link]

2. Convective Mass Transfer: Mass transfer promoted

by fluid flow. It can occur in a single phase or
between two distinct phases. It is analogous to heat
transfer by convection.
5
 Examples on MT Modes
 Sugar dissolving in unstirred
water {molecular diffusion}
 slow process
 If you stir the water with a
spoon  Sugar dissolving in
Mechanically agitated
medium  {convective MT}
 faster process

6
 Molecular Mass Transfer
(Diffusion) is slower
 Someone putting a strong perfume stands in
front of you  it takes some time to smell
the perfume {molecular diffusion of perfume}

 If that person runs in front of you  the

movement agitates (and carries away) the
volatile perfume molecules faster  you smell
the perfume in a shorter time

7
 Limitations of Mass Transfer

 Mass transfer occurs only in mixtures, and

the effect of each component must be
examined

 Each component may have different mobility

(diffusion) Mixture velocity is evaluated by
the average velocity of all components
present
8
 Concentration of species “A”

Concentration of “A” = [amount of A (moles or

mass)]/[unit volume]

 Types of Concentration:

1. Mass Concentration: “ρA”

2. Molar Concentration: “cA”

9
 Mass Concentration
For “n” species in a mixture:
 Total mass density:
n
m1 m2 m m
ρ= ∑ ρi =
i =1
+
V V
+ + n =
V V

 Mass Fraction of “A”:

ρA
ρ A mA V mA
w
=A = = =
n
ρ mV m
∑ i
ρ
i
n

∑ w = 1.0
i =1
i
10
 Molar Concentration

nA ρ A kg m 3 kmole
cA = = [ =] =
V MA kg kmole m3
For gases (if ideal gas law applies):
nA p A
cA = = ( p A = partial pressure of A)
V RT
 For “n” components in a mixture:
n
n1 n2 n3 n P
c = ∑ ci = + + +  = =
i =1 V V V V RT
11
 Mole Fraction
 Mole fraction of “A” in liquids and solids:
cA nA V nA moles of A
xA = = = [=]
c nV n total moles

 Mole fraction of “A” in gases:

c A nA V p A RT p A
y=
A = = =
c nV P RT P
n n

∑x i = 1 .0 ∑y
i =1
i = 1.0
i =1 12
 Interrelations
ρA ρA wA
cA = ; wA = ⇒ cA = ρ
MA ρ MA

For a binary mixture of “A” and “B”:

cA ρA / M A wA ρ / M A
{x A or y A } = = =
c ρ A / M A + ρ B / M B wA ρ / M A + wB ρ / M B
⇓

wA / M A
{x A or y A } =
wA / M A + wB / M B 13
 Class Exercise:
 For a binary mixture of “A” and “B”, Verify that:
 xAM A yAM A 
wA =  or 
 x A M A + xB M B y A M A + yB M B 

 Hint:
 Start with the definition of wA in terms of ρA
 then convert ρA to cA

14
 ρA ρA cAM A
wA = = =
ρ ρ A + ρ B c A M A + cB M B

=
(c A / c )M A
=
xAM A
(c A / c )M A + (cB / c )M B x A M A + xB M B

15
 Some Relations

Table 22.1 Concentrations in a binary mixture of A and B

16
 Example 1

Solution

m3

17
Solution (2)

2.72×10-3 0.555

18
 Example 2

Solution

19
 Mixture Velocity

 Different components in a mixture can move

in different velocities and directions
 The whole mixture can be assumed to move
with an average velocity

20
 Mixture Average Velocities
n n

∑ (ρ v ) ∑ (ρ v ) i i i i n

 Mass-Average Velocity: v= i =1
= i =1
= ∑ (wi vi )
n
ρ
∑ ρi
i =1
i =1

n n

∑ (c v ) ∑ (c v )
i i i i n n

 Molar-Average Velocity: V = i =1
n
= i =1
= ∑ ( yi vi ) or ∑ (x v ) i i
c
∑c i =1
i
i =1 i =1

 Diffusion velocities:
 (vi – v) = diffusion velocity of species i relative to
mass-average velocity of the mixture
 (vi – V) = diffusion velocity of species i relative to
molar-average velocity of the mixture 21
 Fluxes
 In mass transport, each component has its
own mobility under a given driving force
 Flux: Amount of a substance crossing unit
area per unit time, [in molar flux units:
Moles A/(time x area)]
 Flux is expressed according to coordinates:
 Relative to fixed coordinates  NA (analogy: police car
on bridge)
 Relative to coordinates moving with bulk  JA
22
(analogy: police car driving through traffic)
 Fick’s First Law (for
Molecular Mass Transfer) (1)
 Fick’s First Law: For isothermal, isobaric system 
JA is proportional to concentration gradient:
J A = − DAB ∇c A [=] m mole A/m = mole A (transferred)
2 3

s m m 2 .s
dc A
J A, z = − DAB (1-Dimensional Flux in z only)
dz
Why do we need the negative sign?

 JA,z = Molar flux of A relative to moving coordinate

 DAB = Diffusion coefficient of A in a binary mixture
of (A+B), m2/s
23
 Point 1: Point 2:
z1 = 0 m z2= 10 m
Point 3: cA1= 5 mol/m3 cA2= 2 mol/m3
z3= 10 m
cA3= 1 mol/m3

24
 Fick’s First Law (for Molecular
Mass Transfer) (2)
 Flux is proportional to (diffusion coeff.)x(conc. gradient)
 If total concentration is constant (isobaric and isothermal
conditions):
dy A
J A, z = −cDAB
dz
 Mass Flux:
dρ A
j A, z = − DAB
dz
dwA
= − ρDAB For constant ρ
dz 25
 Fick’s First Law (for
Molecular Mass Transfer) (3)
 Recalling the diffusion velocities which define the
movement of a certain component relative to the average
velocity of the mixture 

 The molar flux of A (JA) relative to coordinates moving with

the molar average velocity (V):
dy A
J A, z = c A (v A, z − Vz ) = −cDAB
dz

What are the units of JA,z according to the equation given above?
26
 Fluxes Relative to Fixed
Coordinates (1)
 Flux relative to fixed coordinates can be calculated
from the absolute velocity of each component. For
example: N = J + c V = c v = −cD
dy A
+c V
A, z A, Z A Z A A, Z AB A Z
dz
dy B
and N B , z = J B , Z + cBVZ = cB vB , Z = −cDAB
dz
+ cBVZ

 With the summation for the total flux:

n n
ci
n

∑N
i =1
i,z = ∑ ci vi , z = c ∑ vi , z = cVZ
i =1 i =1 c

27
 Fluxes Relative to Fixed
Coordinates (2)
 For a binary system:
c A v A, Z + c B v B , Z
VZ = or c AVZ = y A (c Av A, Z + cB vB , Z )
c

 Therefore:
dy A
c Av A, Z = −cDAB + y A (c A v A, Z + c B v B , Z )
dz
dy A
= −cDAB + y A ( N A, Z + N B , Z )
dz 28
 Fluxes Relative to Fixed
Coordinates (3)
Generally:
 For a binary system:

N A = −cDAB ∇y A + y A ( N A + N B )

 For multicomponent mixtures

N A = −cDAM ∇y A + y A ∑ N i
i= A

29
 Fluxes Relative to Fixed
Coordinates (4)
 Recalling that:
cA
J A, z = c A (v A, z − Vz ) = c A v A, z − (cVz )
c
n
 We conclude: J = N − c V = N − y ∑ N
A, z A, z A z A, z A i,z
i =1
n
 Or N A, z = J A, z + c AVz = J A, z + y A ∑ N i , z
i =1
 Substituting Fick’s first law for JA,Z:
dy A dy A n
N A, z −cDAB
= + c AVz =
−cDAB + y A ∑ Ni, z
dz dz i =1
30
 Mass Flux
 For a binary system:
n A = − DAB ∇ρ A + wA (n A + nB )

 Under isothermal and isobaric conditions:

n A = − ρDAB ∇wA + wA (n A + nB )
= − ρDAB ∇wA + wA ( ρ Av A + ρ B vB )

31
 Summary (1)

Molar Fluxes Mass Fluxes

J i = ci (vi − V )
Coordinates

ji = ρ i (vi − v )
Moving

= −cDim∇yi
= − ρDim∇wi
Or = − cDim∇xi
N i = ci vi
ni = ρi vi
Coordinates

n
=−cDim∇yi + yi ∑ N i n
Fixed

i =1 =− ρ Dim∇wi + wi ∑ ni
n i =1
Or =− cDim∇xi + xi ∑ N i
i =1
Summary (2)

33
 Molecular Mass Transfer in
Non-ideal Systems
 The molar diffusion velocity of A in terms of Chemical
Potential: d µc DAB d µc
v A, Z − VZ =
−u A =−
dz RT dz
 The molar flux of A: Mobility of
D AB dµ c component A
J A, Z = c A ( v A, Z − VZ ) = − c A
RT dz

 The chemical potential of A in homogeneous ideal

solution at constant T and P:
dµ c RT dc A
µ c = µ + RT ln c A ⇒
0
=
dz c A dz 34
 How Can You Deal with the
Term ΣNi?
Several situations regarding the other fluxes (other than NA).
For example:

 Unimolecular diffusion (UMD): Only “A” is diffusing  ΣNi =

NA n n 0 0 0
∑ N = ∑c v
i =1
i
i =1
i i = c Av A + cB vB + cC vC +  = c Av A = N A

 Equimolar countercurrent diffusion (EMCD): positive fluxes

= negative fluxes  ΣNi = 0 k n m
N = cv − c v =0 ∑
i =1
i ∑
i =1
i i,z ∑
j =1
j j

 Diffusion through stagnant mixture: V = 0  ΣNi = c V = 0

 Reacting mixture of diffusing components: fluxes relate

according to the reaction stoichiometry
35
 Reacting Mixture of Diffusing
Components
Example: Consider the gas phase diffusion with the
reaction:

A (g) + b B (g)  c C (s)+ d D (g)

 NB = b NA
 NC = 0 (why?)
 ND = - d NA

 ΣNi = NA (1 + b - d)
36
 Molecular Diffusion
Coefficient (Diffusivity)

 Measure of speed of diffusion

 Diffusion of A through B (DAB) is a function of T, P,
molecular weights of A and B, environment
 Gas Phase DAB ~ 10-1 cm2/sec
 Liquid Phase DAB ~ 10-6 ̶ 10-5 cm2/sec
 Solid Phase DAB ~ 10-10 ̶ 10-6 cm2/sec

 Molecular diffusion alone is efficient for small

distances (~5 µm)
 Takes hours for longer distances (0.2 cm) 37
 Estimation of Molecular
Diffusion Coefficients
 Experimental values for the diffusion
coefficients of gases, liquids and solids are
listed in Appendices J.1, J.2 and J.3,
respectively.

 When experimental values are not available,

semi-theoretical expressions are available to
give approximations
38
39
40
 Estimation of Gas-Phase
Diffusion Coefficients

Observations from the kinetic theory of gases:

Observation Explanation

DAB α T3/2 As T, molecules move faster

As P, molecules become more crowded & increased

DAB α (1/P)
collisions slow the movement down

DAB α (1/MA)1/2
As MA or MB, heavier molecules move slower
DAB α (1/MB)1/2

41
 Hirschfelder Correlation (valid only
for non-polar, non-interacting GAS pairs)
1.5
0.0018583T 1 1
DAB = +
PΩ Dσ AB
2
MA MB
 Units:
 DAB = Diffusion coefficient (cm2/s)
 T = Temperature (K)
 P = Absolute Pressure (atm)
 MA, MB = Molecular weights of A and B,
respectively
 σAB and ΩD = Lennard-Jones (L-J) parameters 
42
 Lennard-Jones (L-J) Parameters
 σAB = Collision diameter (Å) = (σΑ + σΒ)/2

 ΩD = Diffusive collision integral (dimensionless), tabulated in

Appendix K as a function of (κT/εAB)

 εAB/κ = [(εA/κ).(εB/κ)]0.5 = molecular interaction energy (K)

 If not available in Appendix K, Lennard-Jones parameters can be

approximated from critical and boiling point properties:

 σ (Å) ≈ 1.18 Vb1/3 ≈ 0.841 Vc1/3 ≈ 2.44 (Tc/Pc)1/3

Vb = molecular volume at normal boiling point (cm3/mol)
Vc = critical molecular volume (cm3/mol)
Tc = critical temperature (K)
Pc = critical pressure (atm)
 εA/κ (K) ≈ 0.77 Tc ≈ 1.15 Tb 43
Tc and Tb = critical and normal boiling point temperatures (K)
Lennard-Jones (L-J) Constants (1)

44
Lennard-Jones (L-J) Constants (2)

45
 Other Gas-Phase Correlations
(1)
 Extrapolation of Gas-Phase Diffusivities: ~1
1.5
 T   P0   Ω D (T0 ) 
DAB (T , P ) = DAB (T0 , P0 )     
 T0   P   Ω D (T ) 
1.5
 T   P0 
≅ DAB (T0 , P0 )   
 T0   P 
 DAB = DBA (only for gases)

46
 Other Gas-Phase Correlations
(2)
 Fuller-Schettler-Giddings:
 Can be used when L-J parameters are not available
 Valid for non-polar gases at low pressure
1.75
0.001 T 1 1
DAB = +
[
P (∑ vi )A + (∑ vi )B
1/ 3
]
1/ 3 2 MA MB

Atomic diffusion volume

47
 Atomic Diffusion Volume

Deuterium. It is an isotope of hydrogen containing a neutron in their nucleus

(hydrogen atoms do not contain any neutrons in their nucleus).
Some Heterocyclic Compounds

49
 Diffusion of Polar-nonpolar
and Polar-Polar Gases
1.5
0.0018583T 1 1
DAB = +
PΩ Dσ AB
2
MA MB
1/3
 1.585Vb 
σ AB = (σ Aσ B )
0.5
σ = 2 
* κT  1 + 1.3δ 
T =
ε AB 0.196δ 2 δ AB = (δ Aδ B )0.5
Ω D = Ω D,0 + AB
T*
0.5 1.94 ×103 µ p2
ε AB  ε A ε B  δ=
=  VbTb
κ κ κ 
A C E G
Ω D,0 = + + +
ε
= 1.18(1 + 1.3δ 2 )Tb (T )
* B ( ) ( ) (
exp DT * exp FT * exp HT * ) Dipole
κ moment
Constants values are listed on page 395
 Multicomponent Gas Systems
(Wilke’s Equation)

1
Diffusion
D1− M = '
coefficient of
y2 / D1− 2 + y3' / D1−3 +  + yn' / D1− n
component 1
in the
Binary diffusion
mixture
coefficient of
component 1 into 2
y2 y2
y2′ =
y2 + y3 +  + yn 1 − y1
Mole fraction of component 2
on component (1)-free basis
51
 Example 3

1.5
0.0018583T 1 1
DAB = +
PΩ Dσ AB
2
MA MB

0.001 T 1.75 1 1
DAB = +
[
P (∑ vi )A + (∑ vi )B
1/ 3
]
1/ 3 2 MA MB
52
Solution (1)

53
Solution (2)

54
 Solution (3)
Let CO = A and Ethylene = B.
K.2: ε A / κ 110
From Appendix Table= = K ; ε B / κ 205
At 273 K:
(κ / ε A × κ / ε B ) (1/110 × 1/ 205)
1/2 1/2
κ T / ε AB= T= × 273= 1.817981
⇒ Ω D =1.112 (by interpolation)

At 350 K:
(κ / ε A × κ / ε B ) (1/110 × 1/ 205)
1/2 1/2
κ T / ε AB= T= × 350= 2.330745
⇒ Ω D =1.022 (by interpolation)
55
 Solution (4)
From Appendix Table J.1: DABP of CO in ethylene at 273 K is:
0.151 cm2 atm/s

56
 Solution (5)

If we neglect the effect of temperature on ΩD:

1.5
 350 
0.151 
 273 
DAB = = 0.110 cm 2 / s
2
The error is ~ 8%

57
 Liquid Diffusivity

Stokes-Einstein Equation:
 cm 2  κT
DAB   =
 s  6π (rA µ B )
 κ = Boltzmann constant (1.38064852 × 10-23 [Link].s-2.K-1)
 rA = radius of the solute (A) as a hard sphere
 µB = viscosity of the solvent (B), Pa.s (= N.s/m2 = 10 Poise =
1000 cP)

 DAB α T, DAB α (1/µB), DAB α (1/rA). Explain these behaviors

 Does DAB = DBA? 58
 κT
DAB =
6π (rA µ B )
κT
DBA =
6π (rB µ A )
κT
DAB 6π (rA µ B ) rB µ A
= =
DBA κT rA µ B
6π (rB µ A )
 rB  µ A 
DAB = DBA   
 rA  µ B 
Generally DAB ≠ DBA for liquid-phase diffusion 59
 Correlations for Liquid-Phase Binary
Diffusion Coefficients with Non-Ionic
Solutes (1)

Wilke-Chang Correlation:
 cm 2  −8 T Φ B M B
DAB   = 7.4 ×10 For very dilute
 s  µ BVA
0.6
systems

µB = viscosity of solvent B (cP)

T = Absolute temperature (K)
ΦB = Association parameter for solvent B [page 399]
VA = Molecular volume of solute at normal boiling
point [Tables 22.4 and 22.5] (cm3/mole)
Molecular volume Atomic volume
Table 22.4

If a molecule is
listed here, take VA
directly from this
table
Table 22.5
Otherwise, add
atomic volumes
from this table; pay
attention to
adjustments for
cyclic rings (see next
slide)
Association
2.6 Parameters

Nonpolar solvents
 Corrections for Molar Volume When the
Molecule Contains Cyclic Ring

Ethylene oxide

Cyclobutane

Furan Pyridine ring Naphthalene ring Anthracene ring

 What if Molecular Volume or
Atomic Volume is not Available?

If VA cannot be calculated using this

method, use:
 cm 3 
V A   = 0.285Vc1.048
mol 
 
Where Vc is the critical volume of
the solute (in cm3/mol)

63
 Correlations for Liquid-Phase Binary
Diffusion Coefficients with Non-Ionic
Solutes (2)
Hayduk-Laudie Correlation:
 cm 2  −1.14 − 0.589

DAB  
 = 13.26 ×10 µ B VA
−5

 s 

Hayduk-Laudie Correlation is used for infinite dilution of

non-electrolytes in water

64
 Correlations for Liquid-Phase Binary
Diffusion Coefficients with Non-Ionic
Solutes (3)
Scheibel-Wilke-Chang Correlation:

 cm 2  T × K For very dilute

DAB   = (1 3 )
 s  µ V
B A systems

  3V  2 3 
K = 8.2 ×10 −8 1 +  B  
  VA  

 If benzene is the solvent and VA < 2 VB  K = 18.9x10-8

 For other organic solvents, if VA < 2.5 VB  K = 17.5x10-8
 For water as solvent, if VA < VB  K = 25.2x10-8

65
 Liquid-Phase Binary Diffusion
Coefficients for Concentrated Solutions

For diffusion in concentrated solutions, the Vignes

equation can be used:
Mole fraction of A
Mole fraction of B in the concentrated solution
in the concentrated solution

DAB = (DAB ) (DBA )

xB xA

The diffusion coefficient of the

infinitely dilute solute B in solvent A
The diffusion coefficient of the infinitely dilute solute A in solvent B
66
 Liquid-Phase Binary Diffusion
Coefficients for Ionic Solutes
For diffusion of ions pair in solvent B at infinite dilution:

Valence of cation Valence of anion

 cm 2  (1 / n + + 1 / n − ) RT
DAB   =
 s  (1 / λ0
+ + 1 / λ0
− ) F 2

Faraday’s constant

Limiting (zero concentration)

conductance of anion
Limiting (zero concentration) conductance of cation
67
Table 22.6 Selected limiting ionic conductances in water at 25°C,
reported in units of A.cm2/[Link]
 Example 4

Solution

9.480 × 10-6 cm2/s

(1.306)
69
Solution (2)

2.139 × 10-5 cm2/s

(1.394)
 Pore Diffusivity

 Diffusion of gases or liquids into the

pores of solids is common (Note: the
pores contains gas or liquid not solid)
 Practical examples:
 mass transfer in porous catalysts
 mass transfer in porous adsorbents
 mass transfer in porous membranes

71
 Knudsen Diffusion (1)
If:
 the pore diameter is smaller than the
mean free path of the diffusing gas
molecules and
 the density of the gas is low

The gas molecules will collide with the

pore walls more frequently than with each
other.

This process is known as

Knudsen flow or Knudsen
diffusion. 72
 Knudsen Diffusion (2)
Knudsen number
λ mean free path lenght of the diffusing species κT
Kn = = λ=
d pore pore diameter 2πσ A2 P
Knudsen diffusion coefficient of species A:
cm
d pore 8κNT T
DKA = DKA ( cm 2 / s ) = 4850 d pore
3 πM A MA

If Knudsen diffusion and molecular diffusion compete with

each other, then the effective diffusion of A in a gas mixture of
A and B within the pore of the solid is:
1 1 − αy A 1
κ = 1.38064852 × 10-23 [Link].s-2.K-1 = +
DAe DAB DKA
73
 Knudsen Diffusion (3)

1 1 − αy A 1 For diffusion in straight,

= +
DAe DAB DKA cylindrical pores

NB
α = 1+ Generally:
NA
DAe ≠ DBe
If pores are
ε=
volume occupied by pores
volume of ( solid + pores )
D = ε DAe
'
Ae
2
tortuous

For large pore diameters

D = ε DAB
'
Ae
2
or high pressure 74
 Example 5

Solution For CO2:

75
Solution (2)

76
Solution (3)

77
End of Chapter 22

78