Lecture 13

1. Interphase mass transport-mechanisms

2. Individual mass transfer coefficients
3. Equilibrium at the interface
Interphase mass transfer
Unit operation for mixture separations (absorption, distillation,
extraction) are based on mass transfer of one component (or a group of
components) from one phase into another phase – interphase mass
transfer.
The surface where the two phases are in contact is called
interface.
The interphase mass transport mainly contains three stages: the
transport of the component through the initial phase up to the interface,
the transport through the interface, the transport from the interface
through the second phase.
The driving force for the transport through each phase is the difference
of concentration between the concentration in the phase and the
concentration at the interface.
The main problem in the interphase mass transfer theory is the
clarification of phenomena that occur at the interface.
Interphase mass transport mechanism
In order to define relations that can express the flux of
transferred component, some physical transport models for the
processes in the vicinity of the interface have been proposed in literature.

The simplest one is “the film model” (Whitman, 1923). Other models
are: “penetration theory” (Higby, 1935), “random renewal theory”
(Danckwerts, 1951), “film - penetration theory” (Marchello, 1958).
The most applied theory is the film model (two film model, double layer
model) as it is intuitive . This theory can lead to the formulation of kinetic
equation of mass transfer known as interphase mass transfer equations.

These models are used to calculate convective mass transfer

coefficients ( and also heat transfer coefficients).
 Film model
The film theory is based on the hypothesis that there is a fluid film in
laminar flow next to the boundary which gives the same resistance to
mass transfer as the entire flowing fluid. This hypothesis was valid also
in the heat transfer theory.
In this fictitious film the mass transport is realized by molecular
mechanism.
The film thickness is denoted by , but its value can not be measured.
The film thickness must extend beyond the laminar sublayer to include
an equivalent resistance encountered as the concentration changes in
the turbulent core.
For diffusion through a stagnant layer this theory predicts the mass
transfer coefficient in the gaseous phase as:
D AB p pBm is the mean concentration
k  of component B that does not
 p Bm
diffuse.
Discussion: In the diffusion through a stagnant layer of gas, of thickness
 with concentrations at the limits, p A and pAi , the molar flux of component
A that diffuses in B is:
c DAB y A,1  y A,2
DABp 1 N A, z  
NA   * ( p A  p A ,i ) z 2 z1 yB ,m
 pBm RT
DAB p p A  p Ai
N A,Z   
k  RT pBm

For the hypothesis of countercurrent equimolar diffusion, the mass

transfer coefficient is calculated as:
cDAB
N A, z   y A1  y A2  lecture 11
0 D NA 
DAB
( p A  p Ai ) 
1 z2  z1
k  AB δ RT
δ p DAB D 1
N AZ   ( y A  y Ai )  AB ( p A  p Ai ) 
k0 RT δ δ RT
Coefficient k has higher values than
k0 as in the hypothesis of diffusion
N A  cDAB y A  y A  N A  N B 
through stagnant layer the molar
flux, NA includes also the bulk motion
Pure diffusion bulk motion
In the film theory the mass transfer coefficient directly depends on the
diffusion coefficient.
The fictitious film thickness, , can not be measured. Because of this,
and because of its apparent inadequacy, other theoretical models
have been postulated.
 Penetration model (Higby)
This model was initially proposed by Higby to explain the mass transfer
in a liquid at the absorption of gases. It has been applied by Danckwerts
when the diffusing component only penetrates a short distance into the
phase of interest because of its rapid disappearance by chemical
reaction or its short time of contact. Higby considered mass to be
transferred into the liquid phase by unsteady state molecular transport.
The mass flux at the interface is calculated with the relation:
DAB
NA  (c A ,i  c A , L )
πτ r

r is the period during which the fluid elements remain at interface, c Ai is
the conentration at the interface, cAL is the concentration in the fluid.
 Danckwerts applied this unsteady-state concept to the absorption of
gases in turbulent liquid flow. The motion of the liquid is assumed to
constantly bring new eddies at the interface from the interior of the
liquid which replace the previous elements. The rate of mass
transfer depends on the exposure time. Each fluid element can be
considered to have the same exposure period before it is replaced.

Random renewal model

Danckwerts modified the assumption that all eddies have the same
exposure period at the interface and proposed an infinite range of
periods for the elements at the interface surface. Surface age distribution
is introduced. A surface-renewal factor is defined, S. This model was very
successful in describing the convective mass transport accompanied by
chemical reaction.

N A  DAB S c A,i  c A, L 
 Film penetration model
Marchello pointed out that the penetration concept of Danckwerts is
valid only when the surface renewal is rapid. For older elements of fluid
at the interface surface a steady-state concentration gradient can be
defined as predicted by the film theory.
When conditions are such that the surface contains both young and
older elements, the transfer characteristics are intermediate between the
film and penetration models. The convective mass transfer coefficient
will be proportional to a power of the molecular mass diffusivity between
0.5 and 1.0. These conclusions have been supported by experimental
data.
In these models (film and penetration) the mass transfer involves an
interface between two fluids (applied in absorption, distillation,
extraction). When one of the phases is a solid, the velocity of the fluid at
the interface is zero. In this case a boundary layer model is necessary
(solid dissolution into a liquid).
For diffusion through a laminar boundary layer, the average mass
transfer coefficient is calculated by criterial relations. This is the
engineering approach.
 Two-resistance theory and individual mass
transfer coefficients
The three steps of mass transfer through an interface is best explained in
the frame of two-film model.
The two-resistance theory, initially given by Whitman, is often used to
explain the process. The main principles of this theory are:
-The rate of mass transfer between the two phases is controlled by the
rates of diffusion through the phases on each side of the interface
-No resistance to the mass transfer is present by crossing the interface.
The transfer of a component A from gaseous phase is transferred to the liquid
phase.
A very important point is to correctly express the driving force in the two phases,
that is the concentration of component A in gaseous phase and liquid phase.
For the gaseous phase, the concentration can be expressed in partial pressures,
pA molar fraction, yA or molar ratios, YA.
In the liquid phase the concentration is expressed in molar concentration (mol/m 3),
cA molar fraction, xA or molar ratio, XA.
If no resistance to mass transfer exists at the interfacial surface, p Ai and CAi
are relate by the equilibrium concentration( thermodynamic relations).
The partial pressure, pA is grater than the partial pressure at the interface if the
mass transfer of component A occurs from the gas bulk to the interface.
If the transfer occurs from the liquid phase to the gaseous phase, the
concentration cA will be greater that cAi and pAi will be grater than pA.
 Individual mass transfer coefficients

The two-film model consider that the whole resistance to mass transfer is
concentrated in the two films. The flow in these layers is laminar and the
interface is fixed (due to friction). Outside the layers the flow is turbulent.
This model may represent the reality if the time required to establish the
concentration gradient is small compared to the contact time. The
transport in the layers is by diffusion at steady state.
The rates o diffusion in the two films can be described by the relations:

N A kG  p A, g  p Ai 
N A k L c A,i  c A, L 
pAG and cAL are the bulk concentration in the gas and liquid phase
kG is the convective mass transfer coefficient in the gas phase, and
represents mol of A /time/interfacial area/  units of concentration
kL is the convective mass transfer coefficient in the liquid phase and
represents
mol A transferred/ time/ interfacial area/ units of concentration
Note that the concentration in the two phases may not be expressed
in the same units.

The driving force is the difference of pressure (pA,G-pA,i) in the gaseous

phase and the difference (cA,i-cA,L) in the liquid phase.
The driving force is in this case expressed in units of pressure, for the
gas film and in mol/m3 for the liquid film.
At steady state, the fluxes in the two phases are equal:

N A kG  p A,G  p A,i  k L c A,i  c A, L 

N A kG  p A,G  p A,i   k L c A, L  c Ai 

From these equations, the ratio between the two mass transfer coefficients is
given by the relation:

k L p A ,G  p A ,i
 
kG c A , L  c A ,i
Point M represents conditions found at one plane within a mass exchanger.

pA
kL
slope 
kG
pA,G M

Equilibrium curve
pA,i
Mi

cA,L cA,i cA

The equilibrium curve is a relation of the type: p A  f (c A )

y A  f ( xA )
YA  f ( X A )
 As component A concentration may be expressed in various ways, the
corresponding mass transfer coefficient must also have accordingly units
(see table for specific units for mass transfer coefficients).

Note: The mass transfer coefficient that are calculated from criterial relations
(the engineering approach), results in m/s, as the driving force was expressed in
molar concentration and Sh is a dimensionless number.  L  SI m
kL  DAB SI m2 / s
Sh  Nu AB 
DAB
 k  m / s
 Transformation relations
According to the units the driving force is expressed in, the mass transfer
coefficient must be consistent with it. When the difference of concentration is
given in mol fraction, the mass transfer coefficient is k y(mol/(area*time*y), if
the concentration difference is in molar concentration, the mass transfer
coefficient must be given as kc(m/s), when the driving force is in partial
pressure, the mass transfer coefficient is kG(mol/area*time*p).
Considering the relations between different modalities of expressing concentration
in the gaseous film:
p A  py A RTc A

The molar N A kG  p A,G  p A,i  kG p  y A,G  y A,i  kG RT cA,G  cA,i 
flux of A is
kc=kGRT
ky=kGp

p
ky
kc As from criterial relations kc is calculated, ky  kc
kG   this value must be transformed according to
RT
k
p RT driving force units. kG  c
RT
 For the liquid film:

N A k L c A,i  c A, L  k L c  x A,i  x A, L 

kx = k L c
k L c k x
Where c is the total molar concentration (total number of moles in the mixture
referred to unit volume, which can be calculated as the ratio between the density
of the liquid phase and the mean molar mass).

If the difference of concentration in the liquid phase is given in molar concentrations

( mol/m3), then the mass transfer coefficient must be as kL(m/s, as calculated from
criterial relation. But in the case when the difference of concentration is given in
molar fraction, the units must be transformed correspondingly, and obtain k x
(mol/(area*time*x).
 Example

•The absorption of NH3 in water takes place in a packed bed column at 20C and
105Pa.
The calculation of the individual mass transfer coefficient in the gaseous phase
was performed using the equation: 0.8 0.33
Sh 0.11 Re Sc

The values for Reynolds and Schmidt numbers are: Re=329, Sc=9.5
The characteristic length in Sh expression is the equivalent diameter of the
packed bed particle, d=15.5 mm. The diffusion coefficient of ammonia in air at
working conditions is D=0.254 cm2 /s.

Calculate the individual mass transfer coefficient in gaseous phase in mol/m 2sy.

momentum difusivity ν η kL k
Sc   Sh  Nu AB  
mass difusivity DAB ρDAB DAB DAB / L
Solution

Sh 0.11 3290.8 9.50.33 23.86

k l
Sh 
D
Sh D Sh D 23.86 0.254 10 4
k   0.0391m / s
l d 0.0155

k from criterial relation is kc because:

N A kc c
mol
 N A SI 
m2 s
dimensional analysis:
mol mol
 c 3
k 
m 2s m
 kc SI m / s
 The value for ky will be:

p
k y kc * , as from pv=RT, v=RT/p so, p/RT=1/molar_volum
RT
or equilavent
kc kc
ky   , the units are mol/m 2sy
molar _ volum 22.41 10 3 T p0
T0 p

The volume of a mol in normal conditions is 22.41L, 22.41*10-3 m3

105 mol
k y k p / RT 0.0391  1.65 2
8.31 293 m sy

Please note that 22.41 is L/mol, T0 and p0 are the standard temperature and
pressure (273K and 1 atm). For the first relation, if we use R=8.31 J/mol K, then
the mass transfer coefficient is in mol/(m2sy). If we use R=8314 J/kmolK then kY
is in kmol/(m2sy).
Example 2

The individual mass transfer coefficient in liquid phase is needed for the
absorption of NH3 in water. The criterial relation recommended is:

Sh 0.006 Re0.78 Sc 0.5

For the absorption conditions, the Re and Sc numbers are:

Re=10
Sc=405

The diffusion coefficient DNH3-water=1.76*10-9 m2/s and the characteristic length is

L=4*10-5 m.

Calculate the individual mass transfer coefficient in the liquid phase expressed in
mol/m2sx considering that the concentration of ammonia in water is very small, and
the mean density and molar mass of the liquid are those for water at 20 C.
Solution

Using criterial relations, the mass transfer coefficient is calculated:

Sh 0.006 9.20.78 4050.5 0.68

Sh D 0.68 1.7 10 9 4
k  5
0.29 10 m/s
L 4 10

Note: the mass transfer coefficient in Sherwood is kc, expressed in m/s

(mol/m2s)/(mol/m3)
The density of the liquid is 1000 kg/m3 and the molar mass M=18 kg/kmol

 1000
k x k  0.29 10 4  1.6 10  3 kmol/m 2 sx 1.6mol / m 2sx
M 18
 Equilibrium at the interface
As at the interface equilibrium between the two phases is considered, the
equilibrium model is important in the calculation of overall mass transfer
coefficients

A recall of main phase equilibrium considerations is necessary.

Gas-liquid equilibrium
Henry’ law:
p A H A x A
HA
yA  xA
p

Heny’s law is correct for ideal systems, for nonideal systems H varies with the
concentration of A. The more Henry’s constant is smaller the more the gas is
soluble in the liquid phase.
 Liquid-vapour equlibrium
In a multi-component mixture, the relation between the concentration in
the vapour phase and liquid phase of component i is:

yi ki xi
Where ki is the equilibrium constant
Rault’s law, valid for ideal mixtures gives the relation between partial pressure
and vapour pressure:
pi  xi Pi pi is partial pressure of component i

Pi vapour pressure of component i,

Consequently, for ideal mixture:
function of temperature
Pi
yi  xi p is total pressure
p
pi=yip
 Non-ideal mixtures

For non ideal mixture, the equilibrium relation becomes

i Pi
yi  x i
p

ki

Where i is the activity coefficient reflecting the non-ideality in liquid phase

and is a function of temperature and composition.

That is, the equilibrium constant ki depends also on liquid composition.

Mixtures as alcohols, water and alcohols are characterized by interactions

between molecules and activity coefficients must be calculate.
Some wellknown models for liquid vapour equilibrium are Van Laar, Margules,
Wilson, UNIQUAC, NRTL
The correct calculation of the equilibrium constant is of most importance in the
design of apparatus as distillation columns, absorption columns, extraction
columns. In these cases, both mass transfer and equilibrium must be considered.
Simplified approach using the relative volatility

The volatility is defined as the ratio of partial pressure of component i in the

vapour phase and molar fraction in the liquid phase. For ideal mixtures the
volatility is equal to the vapour pressure.
pi Px
i i

pi
volatility  Pi
xi
The relative volatility between component i and j is defined as:

pi / xi yi p / xi  yi / xi 
i, j   
pj / xj yj p / xj yj / xj 

 
 y /x  i i
i, j
y j
/ xj 

i is the lighter component and j is the heavier component, y and x

stand for molar fraction , at equilibrium, in the vapour and liquid
phase
For ideal mixtures (A,B), the relative volatility is defined as the ratios of the vapour
pressures of the two components. A is the light component, having a bigger vapour
pressure at a given temperature.
PA PB
As for ideal mixture (Rault law) : y A  xA yB  xB
p p

 AB 
PA
/ p x A  / x A PA

PA
α A, B 
P
B
/ p xB  / xB PB PB
For binary mixture (A,B with A the light component), with constant relative volatility,
the equilibrium equation can be expressed as Fenske equation that allows to
calculate the vapor composition, yA using the molar fraction of A in liquid phase (xA)
y A / xA αx A
 A, B  yA 
(1  y A ) /(1  x A ) 1  (α  1) x A
The equilibrium may be given by experimental data sets ( y, x). Sometimes is
possible to corelate this data by simple regression curves, but equilibrium models
are recommended.
Basicaly, the equilibrium in a vapour-liquid system or a gas-liquid system is
expressed as yA=f(xA) or pA=f(cA)