Subtheme 1 Fick's First Law of Binary Mixtures

Binary mixture in an isothermal and isobaric system

J CDiz X Fick's Law

Diz is the diffusivity or Fick'sdiffusion coeficient

For a binary system Fick's diffusivity D and the MSdiffusivity

D are related by

J CDist X
Diz Dirt
In an ideal system I is equal to 1 meaning that
the diffusivities are identical Diz Dis ideal systems
Fick's diffusivities concentrates on two elements

Importance of an inverse drag coefficient D

Thermodynamic non ideality T

The physical interpretation of Fick's diffusion coefficient is less clea

The constant of proportionality of the FICK's equation Diz is known

as the diffusion coefficient

Die
it It me in earth
 The unit is identical to that of the other fundamental transfer
properties Kinematic viscosity r and thermal diffusivity o

The diffusivity of the mass is given in cm s m s or ft h

The property indicates the ability of a particular compound
to diffuse into the mixture

The diffusive flux J is directly proportional to the diffusion

coefficient Diz Systems with greaterdiffusion coefficients values
indicate that the system presents a greaterfacility towards
the diffusive moleculartransport mechanism Thediffusion
coefficient depends on the pressure temperature and

composition of the system

According to the mobility of molecules the diffusion coefficients

show significant differences depending on the phase in takes place
 Subtheme 3 Diffusion coefficients for binary idealgassystems

For low pressure binary gasmixtures DAB is inversely proportional

to pressure increases with temperature and is almost independent
to composition Combining the principles of kinetic theory and
the corresponding states the following equation has been
obtained to estimate Dais at low pressures
Where
DABdiffusivity cm s

pT t.be jis
DAB a
fate T absolutetemperature
P totalpressure atm
1 1958794 BIKE

YEI.LY To nonpg9.99

At high pressures DAB no longerdecreases linearly with pressure Very

little is known about the variation of diffusivity under pressure
Frequently used correlations are
Gilligand 1934
DAB 4.3 10 9
p v15 3 Ma MB

Where
DABdiffusivity m s

Mmolecularweight kg kgmol
V molarvolume at normal boiling point m kgmol
 Fuller et al 1966
DAB
p É ÉÉÉvj p Ma Mis

Where
EV sum of the atomicvolumes of each molecule m kg

Chapman and Enskog

For the case of a binary mixture of nonpolargases at low pressure
DAB
1.8.58 in M

Where
M molecular weight kg kgmol
TAB collision diameter m

D collision integralfunction F KT Ei j or

RD do Ai in x where or
eat k

1 03587
10 9,36 expid exlp.isda4ix
5x exp1.52996 1
Diffusion coefficients for Binary Ideal Gas systems
 GAB A 0B

EFE 91 1
K Boltzmann's constant 1.38 10 erg k
Ei j molecular interactionenergy is generally estimated by
Lennar Jones parameter as follows

RD AB a TB

EAB FEB
Vc critical volume m kgmol
V normal volume at normal boilingpoint m kgmol
Tc critical temperature K

Ts temperature at normal boiling point K

o collision diameter 10 m

It is recommended to use the method of Tyn and Calus 1975

to calculate the critical volume Vc of the compounds that
are not found in table with the following equation
048
Vb 0 285V1
Otherwise the atomic volume of each element present is

added according to the molecular formula of the compound

 Molar Volumes at Normal Boiling Point

Atomic Volume contributions of the Elements

The following expression is extremely useful when the diffusivity is

known at some pressure and temperature and its desired at
another temperature To extrapolate experimental data at
moderate pressures
5
DAB 0 00266 T
P MI TAB RD
MAB
ME
Where
DAB diffusion coefficient cm4s

MAMB molecular weight of A B

T temperature K

P pressure bar
TAB collision diameter a Lennard Jones parameter A
RD diffusion collision integral dimensionless

The collision integral is a function of the temperature and of

the intermolecular potential field for one molecule of A and
one molecule of B
T
15
In the absence of the experimental values of the Lennard

Jonesparameter for pure components they can be estimated

from the following empiricalequations

0 11 18V13

1 15T

Experimental Values of Dilute Gas Diffusivities at 1 atm

 Wilke and Lee one of the most widely used methods for

estimating DAB in low pressurebinarygas systems

5
103
DAB i

Subtheme 4 Diffusion Coefficients for Dilute Liquids

The diffusion of solutes into liquids is very important especially
in separation operations such as liquid liquidextractions or

solventextractions gas absorption and distillation Its alsofrequent

In nature as in the case of oxygenation of rivers and lakes and

the salt in the blood

The speed of molecular diffusion in liquids is lower than in gases

The molecules of a liquid are very close to each othercompared
to those of a gas therefore the molecules of solute A that diffuse
will collide with the molecules of liquid B more frequently

and they will diffuse more slowly than in gases

The density and resistance to diffuse in liquids are much greate
Due to the proximity of the molecules the attractive forces between
them have an effect on diffusion Thediffusion of liquids is quite

dependent on the concentration of the components that

are diffused
 If its true that the diffusion coefficients obtained when the mole
traction of any of the components in a binary mixture approaches

unity the thermodynamic factor 5 approaches unity and the

Fick diffusivity and the MS diffusivity are equal these are

called infinite dilution diffusion coefficients Do which can be

estimated using the Stokes Einstein equations

Where
DAB
entity Va radius of thesoluteparticles

MBviscosity ofthe solvent

Thisequation is used as the basis for

many empirical correlations ordered in the generalform

DEYB f Vba
Where

f Vba function of the molecular volume of the diffusing solute

Correlations have been derived to predict the diffusion coefficient

of the liquids in terms of the properties of the solute and solvent
One of the modifications was made assuming that all molecules
are equal distributed in a cubic lattice and whose molecular

radius is expressed in terms of molar volume

DAB 9.96 1016T
M
 Where
DAB diffusivity m s

T absolute temperature K

M viscosity of thesolution Pa's

Va molar volume of the solute at itsboiling point m kgmol
For solutes of small molar volumes
7.4 DBMB
DAIMB
1
Where
DAB diffusivity of A in very dilute solution in solvent B cm s

MB molecular weight ofsolvent B kgkgmol

T temperature K

MB viscosity ofsolvent B CP
Vba molar volume of the solute at its normal boilingpoint cm mol

Vba 75.6 cm mol for water as solute

OB association factor of solvent B dimensionless

013 2.26 for water as solvent
1.9 for methanol as solvent
1.5 for ethanol as solvent
1.0 for unassociatedsolvent benzene ether heptane
The associationfactor for a solvent can be estimated only when
diffusivities in that solvent have been experimentally
measured
Hayduk and Minhas reasoned correlations for the binary
diffusion coefficient of infinite dilution and using regression
analysis they proposed some correlations according to the

type of solute solvent system

a For solutes in aqueous solutions

8 52
DAB 1.25 10 V51 0.292 T M

E 915 1 12

Where
DAB diffusivity of A in
very dilute aqueous solution cm s

T temperature K

MB viscosity of water CP
Vba molar volume of the solute at its normal boiling point cm m

b For non aqueous non electrolytic solutions

DAB 1.55 108

Yb
I 915
Where
surfacetension at normal boiling pointtemperature dyn cm

The values of surface tension can be estimated when the values

are unknown limited to non polar liquids
 Tb Pc To 0.132dc 0.278 1 Tbr

Tbr

dc 0.9076 1 Tb In
1 1013

Where
Tb surface tension at normal boilingpointtemperature dyn c
Pc critical pressure bar
Tc critical temperature K

some restriction must be taken into consideration these are

Subtheme 5 Diffusion coefficients for concentrated Liquids

In Vignes proposed an empirical equation that fits for ideal and
non ideal solutions intending to predict D in concentrated

liquid solutions he proposed to combine the diffusioncoefficient

of infinite dilution Diz and Dai in a single compositionfunctio

Diz Diz Dai

D12 Dirt
Subtheme 6 Effectivediffusivities in Multicomponent Mixtures

An effective diffusivity or pseudobinary perspective has beenused

for the diffusion in multicomponent systems
This effective diffusivity is a parameter that quantifies the
transfer of matter encompassing all the diffusion processes that
occur in the system it is assumed that the diffusionspeed
of component i depends only on its own composition

gradient as presented

Ji CDiett Xi 1 1,2 n

Where
Di etf is some characteristic diffusioncoefficient of species i in

the mixture that must be synthesized from the binary

diffusivities of MS

Forgases the binary diffusivities of MS are held typicallythey

are admitted as independent of the composition It is possible
to derive from the MS equation the effective diffusivity
component i in a multicomponent gas mixture resulting
in the following equation

Diett
1 i in YiNi
To use the equation above one must know the relationship
between the N values This relationship is defined as a

bootstrap condition and is generally subject to other

considerations such as the stoichiometry of reaction or

energy balances As an example image the general

chemical reaction written as
i ViA 0

Where
Vi stoichiometric numbers The sign convention is positive t

for products and negative for reactants

Ai chemical formula

In the case of diffusion with a chemical reaction formed

by elements of different class or nature the relationship

between the fluxes are the stoichiometry

defined through
of the reaction such as

An example of this would be the solid catalyst cracking of

ammonia according to the reaction
NHz
1 Nz 3 H2
Under the scenario that NH 111 diffuses to the catalyst
surface and N2 2 and H2 3 diffuse applying the previous

equation we obtain
N2
IN
Nz
3N
Theeffective diffusivity may vary widely along the diffusion

path however a linear variation with distance can

usually be assumed Consequently if Di eff changes
significantly from one end of the diffusion path to the other
the arithmetic average of the two values must be used

In multicomponent diffusion typically it occurs that all N

except N is zero all but component 1 are stagnant

Diett
Eh To i i Ez Yi Die

Where
Yi mole fraction of component i in a freebase of component 1

Yi Iy 1 2,3 n
The demonstration of the equation above is presentedbelow
Where n 4 a 1 1
Diett
ii Elitin yini

Diett
Iii.tw Yi.t E.cklvi nns man t.nu

Where all Nis except N are 0 N2 Ns N4 0

Diett
it
can
It Yani

Diett
µ Yet's ee
y
Dividing numerator and denominator by 1 Y

Diett
a
if Eiji
Where 4
y