Determination of diffusion coefficient for liquids and gases

Research Assistants 2019-2020 Fall Semester

Elvan Konuk
Dr. Gökçe Dicle Kalaycıoğlu

DETERMINATION OF DIFFUSION COEFFICIENT FOR LIQUIDS AND GASES

1. INTRODUCTION

Molecular diffusion, often called simply diffusion, is the thermal motion of all (liquid or gas)
particles at temperatures above absolute zero. The rate of this movement is a function of
temperature, viscosity of the fluid and the size (mass) of the particles, but is not a function of
concentration[1]. Diffusion explains the net flux of molecules from a region of higher
concentration to one of lower concentration, but it is important to note that diffusion also
occurs when there is no concentration gradient. The result of diffusion is a gradual mixing of
material. In a phase with uniform temperature, absent external net forces acting on the
particles, the diffusion process will eventually result in complete mixing2.

The kinetic theory of the gases provides a means of visualizing what occurs, and indeed it was
the success of this theory in quantitatively describing the diffusion phenomena which led to
its rapid acceptance. In the case of a simplified kinetic theory, a molecule is imagined to
travel in a straight line at a uniform velocity until it collides with another molecule,
whereupon its velocity changes both in magnitude and direction. The average distance the
molecule travels between collisions is its mean free path, and the average velocity is
dependent upon the temperature. The molecule thus travels a highly zigzag path, the net
distance in one direction which it moves in a given time, the rate of diffusion, being only a
small fraction of the length of its actual path. For this reason the diffusion rate is very slow,
although we can expect it to increase with decreasing pressure, which reduces the number
collisions and with increased temperature, which increases the molecular velocity4.

Diffusion is quite different from the movement of molecules when a fluid is flowing. In this
case movement is not random; all molecules are moving together and in the same direction[3].

In biological systems, diffusion plays an essential role in the transport, over short distances
(for example across cell membranes), of molecules such as nutrients, respiratory gases
(carbon dioxide and oxygen)3.

1
 Determination of diffusion coefficient for liquids and gases

One application of diffusion is the separation of isotopes, particularly those of uranium. When
uranium hexafluoride diffuses through a porous plate, the ratio of the 235 and 238 isotopes is
changed slightly. With sufficient number of passes, the separation is nearly complete. There
are large plants in the USA and UK for obtaining enriched fuel for fast nuclear reactors and
the fissile uranium-235, originally required for the first atom bombs. Another application is
the diffusion pump, used extensively in vacuum work, in which the gas to be evacuated
diffuses into a chamber from which it is carried away by the vapour of a suitable medium,
usually oil or mercury3.

2. THEORY

When there is a concentration gradient in a multicompartment thermodynamic system, a

matter irreversible flux starts from the high concentrations to the low ones. This flux is called
“diffusion”. The diffusion tends to give back the system to its balance state, of constant
concentrations. Fick’s Law describes the diffusion. For the statement of such law, binary
mixes will be only considered, although possible similar developments are possible for
multicomponent mixes5.

2.1 Fick’s Law of binary mixes

Let’s consider the system represented on Figure 2.1. In such system, a liquid A evaporates in
an insoluble pure gas B.

At the same time, the liquid A comes out very slowly by the bottom of the system, as it is
shown on the Figure 2.1. The vapors do the same when they are removed by a draft of gas B
that flows by the top of the system.

In view of the circumstances regarding the concentration differences of A between the gas in
contact with liquid and the gas in the top, there will be constantly a movement of the vapors
through the narrow tube to the top. On the other hand, the gas mix will move globally to
downward, because of the liquid emptiness. If this emptiness is slow enough, the net
movement of A will be upward.

2
 Determination of diffusion coefficient for liquids and gases

Mix of gases A and B

Gas B
Vapor A
y

x
Liquid A

Figure 2.1. Diffusion of a vapor A in a gas B, which is insoluble in the liquid phase A

The experimental Fick’s Law of the diffusion establishes that in a system in which there is
concentration gradient, the component A moves, regarding the mixing global movement,
taking the direction of the concentration negative gradient CA, according the following
equation.

(1)

where;

JA = Molar flux of A (kgmol A/s.m2)

DAB = Diffusion coefficient, molecular diffusivity of the molecule A in B (m2/s)

CA = Concentration of A (kg/m3)

= gradient operator applied to concentration of A (kg/m3)

The equation (1) is not the usual form of Fick’s Law, since the coefficient DAB depends on the
system concentration. That is why the next expression is preferred:

(2)

3
 Determination of diffusion coefficient for liquids and gases

where;

xA = Mole fraction of A

This equation represents the so-called Fick’s Law of the diffusion. In this case, the diffusivity
DAB is almost independent of the system concentration.

The units in the international system for the diffusion coefficient (DAB) are [m2/s], although,
frequently, they are expressed in [cm2/s]. The diffusivity numerical values change a lot
depending on the system that is considered. This variation is especially important when the
different states of matter aggregation are considered thus, the gases diffusivities change
usually at the interval from 10-5 to 10-4 [m2/s], whereas for the liquids the variation is found at
the interval from 10-10 to 10-9 [m2/s]. The diffusivity in the solids varies a lot from a system to
another one, and their values are much smaller, approximately from 10-34 to 10-14 [m2/s].

Although the gases kinetic theory is very simple, it is interesting to watch a gas diffusivity
dependence with the pressure and the temperature. The diffusion coefficient increases with
the temperature and decreases with the pressure5.

Another very useful form to express the first Fick’s Law is:

(3)

where;

NA, NB = Rate of convective mass transfer (kgmol/m2.s)

xA = Mole fraction of A

DAB = Molecular diffusivity of the molecule A in B (m2/s)

C = Concentration (kg/m3)

Equation (3) is the final general equation for diffusion plus convection which can be said as
the “combined flux”. “ ” term indicates the “convective flux” and “ ”
indicates the “molecular flux”5.

4
 Determination of diffusion coefficient for liquids and gases

2.2 Applicaton of Fick’s Law for the Diffusivity Determination.

The experimental determination of the diffusion coefficient, DAB, for a binary mixing can be
made through a device similar to the one shown on Figure 2.2. Let’s consider indeed a test
tube with dimensions such that the effects in the system limits can be disregarded. Let’s
suppose likewise that inside a pure liquid A evaporates slowly and continuously; and that its
vapors are spreaded over B, an inert gas in rest. During the experiment, the temperature and
the pressure are kept constant. On the other hand, the diffusion will be only considered in the
direction of the axis “z”. If the evaporation is slow enough, the surface level will remain
almost constant, and as a result, there will be no flow of B in the test tube inside. That is to
say, NB=0 (the gas is insoluble in the liquid)4.

Figure 2.2 Device to measure the diffusivity for a gaseous environment

Regarding this hypothesis goodness, it is useful to remind that, in normal conditions, the
volume of an evaporated gaseous mass is hundreds of times above the volume occupied in the
liquid phase.

5
 Determination of diffusion coefficient for liquids and gases

Since there is only flow in the direction of the axis “z”, the matter microscopic balance for the
component A in a stable state and in absence of chemical reaction will be:

(4)

where;

NA = Rate of convective mass transfer (kgmol/m2.s)

z = Distance from reference level z0 (m)

Therefore, integrating;

(5)

And C1 is an integration constant. This last equation shows that the molar flow density of A in
the direction of the axis “z” is constant, that is to say, the molar flow that goes through any
section is the same one.

Applying Fick’s Law in its form expressed by the equation (3) and formulating properly the
component “z” of the gradient operator in cylindrical coordinates, the result will be:

(6)

And, since there is no flow of component B, it will be possible to write:

(7)

Differential equation which integral is obtained directly, being constants NAz and c are ideal
when the pressure and the temperature are constant. Likewise, DAB is considered invariable.
Thus:

(8)

where C2 is a new integration constant. On the other hand, there are two boundary conditions,
which the composition of the gas must comply with; on the first place, if in the interface gas-
liquid it is admitted that the partial pressure of A, pA, the vapor pressure for the pure

6
 Determination of diffusion coefficient for liquids and gases

component to the temperature of the experiment, and Dalton’s Law is fulfilled, the result will
be:

For z= 0 (9)

Being p the total pressure of the system, and

For z= L (10)

Where L (m) is the length of the test tube (Figure 2.2) that contains the gaseous mixing and
xA,L is the molar fraction of A in the gaseous flow A-B that circulates on the upper part of the
test tube.

Applying both boundary conditions to the integrated equation, the result, easily obtained, will
be an expression for the diffusivity calculation:

(11)

If it is preferred, this last equation can be written according to the partial pressures of the
gases A and B. Thus, if it is indicated by pBL to the partial pressure of the gas B on the upper
part of the test tube, and through pB0 to the partial pressure of this gas in the interface, having
in mind that for an ideal gas, C= P/RT, it is obtained:

(12)

This equation allows the estimation of diffusivities if the user knows the total pressure P, the
temperature T and the values for the partial pressures of B in the interface and on the upper
part of the test tube. NAz is calculated by a weighing or by lectures of level from the
evaporated moles of A (mev) (kg) in a certain period of time (tk) (s), knowing, of course, the
cross sectional area of the test tube, A*:

(13)

From the developed equations, it is possible to estimate a concentration profile for A or for B.

7
 Determination of diffusion coefficient for liquids and gases

Applying the first condition limit to the equation (8), it is obtained:

(14)

As a result, it is obtained:

(15)

It is an equation that relates the concentration variation (xA) with the position z. That is to say,
it is the expression of the concentration profile. On the other hand, from the equation (11) and
(14), another alternative expression obtained for such profile:

(16)

As a result, it is deduced that:

(17)

It is a similar equation to the (15) one, but according to both boundary conditions.

Actually, nev is the number of evaporated moles; remaining that the number of moles is the
mass, nev, divided by the molecular weight, MA, and the result is:

(18)

where;

mev = Evaporated mass of A (kg)

MA= Molecular weight of A (kg/kgmol)

And to find the value of the mass:

(19)

8
 Determination of diffusion coefficient for liquids and gases

Finding the value of the unknown quantity and taking the derivative of the mass regarding the
time:

(20)

On the other hand, the mass flow equals:

(21)

where;

= Mass flow rate (kg/s)

= Density (kg/m3)

= Linear velocity (m/s)

As the mass flow is the change of the mass regarding the time, and the velocity is the change
of the space regarding the time, the result will be:

(22)

If the equations (20) and (22) are made equal, the result will be :

(23)

Deleting the common elements:

(24)

Substituting NAZ by its value obtained from the equation (14), ,

the result will be:

(25)

9
 Determination of diffusion coefficient for liquids and gases

Tidying up some terms:

(26)

By grouping the constant terms in only one term which will be called α, and having in mind
that for an ideal gas, C= P/RT, it is obtained :

(27)

Substituting:

(28)

Taking the integral:

(29)

where; z0 is the reference level to compare with the point z.

The result is:

(30)

Operating :

(31)

Developing the differences of squares:

10
 Determination of diffusion coefficient for liquids and gases

(32)

Tidying up:

(33)

Operating

(34)

Grouping again:

(35)

Finally, the final expression is:

(36)

to which corresponds an equation of a straight line with the form:

(37)

Thus, representing regarding (z-zo) the diffusivity DAB can be obtained from the

slope of straight line1.

2.3 Prediction of diffusivities for binary solutions

Since the theory for diffusion in liquids is not very well established as yet, the equations for
predicting diffusivities in liquids are by neccessity semiemprical.

However, we shall review briefly the basic theoretical equations from which the final semi
empirical equations are derived6.

11
 Determination of diffusion coefficient for liquids and gases

Basic theoretical equations for dilute solutions

The Stokes-Einstein equation, one of the first theories, was derived for a very large spherical
solute particle(A) diffusing through a liquid solvent (B) of small particles. The Stokes law
was used to describe the drag on solute particle as it moved. The equation is:

(38)

where DAB is diffusion coefficient (m2/s) B is the viscosity and is the molar volume of
solute.This equation applies very well to very large unhydrated molecules of about 1000
molecular weight and greater in low-molecular-weight solvents or where the molar volume
VA of the solute is above about 500 cm3/g mole7.

The theory of absolute reaction rates by Eyring [Geancoplis, Perkins]. Has also been used to
predict viscosity, thermal diffusivity, and molecular diffusion coefficients. The liquid is
treated as a lattice like array of molecules, which is imperfect because of holes or vacancies.
The solute molecules carrying momentum, heat or mass migrate or jump from one hole or
equilibrium position to another. As the molecules migrates, it actually carries momentum,
heat or thermal energy, and mass-a different species of meterial. However, heat and
momentum can also be transported by an additional mechanism by actual collision with
adjacent molecules without migrating to a “hole”. Hence, the mass diffusivity will be the
smallest of the three. For example, in water at oC the self-diffusivity of water , DAA, is
1.35x10-5 cm2/s, thermal diffusivity α (k/Cp) is 142x10-5 cm2/s, and momentum diffusivity
µ/ is 1790x10-5 cm2/s. Hence, actual migration mechanism is small compared to the collision
mechanism in liquids. The final equations of the absolute reaction rate theory paralel those by
the Stokes-Einstein method and do not predict the diffusivities for liquids well. Arnold
[Thakar] applied the kinetic theory of gases to the liquid state and predicted the diffusivity to
be proportional to 1/µ0.5 [7]

Estimation of diffusion coefficients in dilute solutions

The Wilke-Chang correlation can be used for most general purposes where the solute (A) is
dilute in the solvent (B):

) (39)

12
 Determination of diffusion coefficient for liquids and gases

where MB is the molecular weight of B, µB is the viscosity of B in Eq.(39),VA is the solute

molar volume at the boiling point, and is an “association parameter” of the solvent, where
is 2.6 for water, 1.9 for methanol,1.5 for ethanol and for benzene, ether, heptane or other
unassociated solvents. When molar volumes are above about 400 cm3/g mole in low-
molecular-weight solvent, the Stokes-Einstein equation should be used. When water is the
solute, the values predicted from Eq. (39) should be multiplied by a factor of 1/2.3. Eq. (39)
predicts the diffusivities with a mean deviation of about 10-15 percent for aqueous solutions
but only to about 25 percent in nonaqueous solvents. Data indicate that the temperature
correction of DAB proportional to T/µB may over correct. The equation should be used with
caution outside the temperature range of 5 to 40 oC. Perkins and Geancoplis state that for
highly viscous liquids of hydrocarbons and glycerol the viscosity exponent should be less than
1.0 and between 0.5 and 1.0 [6].

Othmar and Thakar give a simple equation for solutes (A) in water (B), which is as accurate
as Eq. (39)

(40)

where µB is again cp. This equation is often preferred for water as the solvent.

2.4 Prediction of diffusivity for gases

The diffusivity of a binary gas mixture in the dilute gas region, that is low pressures near
atmosheric, can be predicted using the kinetic theory of gases. The gas is assumed to consist
of rigid spherical particles that are completely elastic on collision with another molecule,
which imlies that momentum is conserved8.

In a simplified treatment it is assumed that there are no attractive or repulsive forces between
the molecules. The derivation uses the mean free path λ, which is the average distance that a
molecule has traveled between collisions.

The final equation is:

(41)

13
 Determination of diffusion coefficient for liquids and gases

where DAB is the diffusivity in m2/s, is the average velocity of the molecules8.

Semi emprical method of Fuller, which is more convenient to use is often preferred. The
equation was obtained by correlating many recent data and uses atomic volumes.

The equation is:

(42)

where DAB is the diffusivity in m2/s, T temperature in K, MA is the molecular weight of A in

kg/kg mol MB is the molecular weight of B, P absolute pressure in atm, sum of structural
volume increments for A, sum of structural volume increments for8.

The equation shows that DAB is proportional to 1/P and T1.75. If an experimental value of DAB
is available at a given T and P and it is desired to have a value of DAB at another T and P, one
should correct the experimental value to the new T and P by means of relationship DAB α
T1.75/P [8].

14
 Determination of diffusion coefficient for liquids and gases

3. EXPERIMENTAL PROCEDURE

3.1 SAFETY

People who work in laboratories are exposed to many kinds of hazards. However, many
accidents in the laboratory can be prevented if each person in the laboratory observes a set of
common precautions and uses proper experimental techniques and procedures at all times.

Please follow these instructions when you came across an accidental situation.

 If you receive a chemical splash (NaCl solution) to your eyes, immediately flush eyes with
plenty of water at least 15 minutes, lifting upper and lower eyelids occasionally9.
 Inhalation of asetone vapors irritates the respiratory tract. It may cause coughing,
dizziness, dullness, and headache. If you come across high concentration of acetone vapor,
come out to take fresh air 9.
 Spills shall be cleaned immediately. Small spills of liquids on bench tops shall be cleaned
immediately with personal protective equipment (gloves) to prevent contact with skin or
clothing.
 Your hands and the buttons of the device have to be dry when you working with an
electrical device.
 It is the responsibility of everyone working in the laboratory to make certain that the
laboratory is left clean after work is performed.

3.2 PART I Diffusion in Liquids

The unit which is used during the experiment (Figure 3.1) consists of :
 A liquid tank
 A magnetic stirrer and magnet
 A conductimeter
 A conductivity sensor
 A diffusion cell (capillaries number (N)= 317 capillary length (x)= 5 mm )
 Thermostatic bath (water bath, resistance, level switch, temperature sensor)
 A console that includes:

15
 Determination of diffusion coefficient for liquids and gases

 Temperature digital indicator

 Digital controller of the water temperature
 Heating resistor switch
 Level switch
 Water temperature sensor
 General switch

Figure 3.1. “Liquid diffusion” experiment system

Figure 3.2. Control console of “liquid diffusion” experiment system

The electrical conductivity of aqueous solutions is governed by the presence and

concentration of ions in solution. Therefore, pure water does not conduct an electrical current
well since the concentrations of hydrogen and hydroxide ions are very small. Solutes whose
solutions are conductive are called electrolytes and a solute is considered a strong electrolyte

16
 Determination of diffusion coefficient for liquids and gases

if it dissociates completely into its constituent ions e.g. NaCl and KI aqueous solutions.
Thus, if concentration is higher, that means more ions are present per volume of water and the
conductivity of the solution will be higher. Conductivity has the SI units of siemens per metre
(Sꞏm-1) and cgs units of inverse second (s–1).

Diffusion in Liquids

 Prepare NaCl solutions at different concentrations (2 M, 1.5 M, 1M).

 Fill up completely the cell with NaCl solutions ( 1M, 1.5 M, and 2 M, respevtively) and
clean up any excess from the solution of the cell outside and from the capillary ends using
filter paper.
 Put the cell in its position within the receptacle.
 Fill the receptacle of the measurement unit with distilled water.
 Turn on the magnetic stirrer at a not very quick speed.
 Turn on the conductimeter and measure the conductivity every 30 seconds.
 Turn on the resistance when the cell was filled up with 2M NaCl solutions and repeat the
same measurements at 50oC.

3.3 PART II Diffusion in Gases

The unit which is used during the experiment (Figure 3.3) consists of :
 A narrow vertical tube with a known inside diameter
 An air pump
 An air optical system, with a precision focusing adjustment, assembled on an element
of vertical movement with a capiler and a vernier
 A water bath to regulate the volatile liquid temperature
 A console that includes:
 Temperature digital indicator
 Digital controller of the water temperature
 Heating resistor switch
 Air pump switch
 Level switch
 Water temperature sensor
 General switch

17
 Determination of diffusion coefficient for liquids and gases

Figure 3.3. “Gas diffusion” experiment system

Figure 3.2. Control console of “gas diffusion” experiment system

Diffusion in Gases
 Fill up the capillary tube with acetone and note of acetone highness.
 Take note of acetone highness every 60 minutes.
 Turn on the resistance and perform the same experiment at 45oC.

18
 Determination of diffusion coefficient for liquids and gases

4. CALCULATIONS AND DISCUSSION

4.1 Diffusion in Liquids

 Find the ratio between the concentration of a saline solution and the conductivity
 Calculate the diffusion coefficient of 2M saline solution from the conductivity data.
 Calculate the diffusivity of 1.5 M and 1 M NaCl solutions from the conductivity data.
Compare the diffusion coefficients of saline solutions at different concentrations. Discuss the
effect of concentration in the diffusivity and the reasons of the differences between them.
 Calculate the diffusion coefficient of 2M NaCl solution at 25oC and 50oC. Discuss the
effect of temperature in the diffusivity and the reasons of the differences between them.

4.2 Diffusion in Gases

 Calculate the diffusivity of acetone and discuss the vapor diffusion of acetone in the air.
 Determine the molar flux density of the volatile liquid.
 Calculate the diffusion coefficient of acetone at 25oC, 35oC, 45oC. Discuss the effect of
temperature in the diffusivity and the reasons of the differences between them.
 Represent graphically the concentration profiles of acetone and air.

19
 Determination of diffusion coefficient for liquids and gases

5. NOMENCLATURE

A* : Cross sectional area of the tube (m2)

C : Concentration (kg/m3)
C0 : Initial concentration (kg/m3)
CA : Concentration of A (kg/m3)
C1, C2 : Integration constant
D : Diameter of a capillary (m)
DAB : Molecular diffusivity of the molecule A in B (m2/s)
JA : Molar flux of A (kgmol A/s.m2)
L : Length of test tube (m)
mev : Evaporated mass of A (kg)
MA : Molecular weight of A (kg/kgmol)
MB : Molecular weight of B (kg/kgmol)
nev : Evaporated mole of A (mol)
NA : Rate of convective mass transfer (kgmol A/s.m2)
P : Total pressure (N/m2)
pA : Partial pressure of A (N/m2)
pB0 : Partial pressure of gas in the interface (N/m2)
T : Temperature (K)
tk : Passed time (s)
: Lineer velocity (m/s)
VA : Molar volume of the solute (m3/mol)
: Atomic diffusion volumes for gas A to use with Eq. 42
: Atomic diffusion volumes for gas B to use with Eq. 42

xA : Mole fraction of A
z0 : Reference level of the liquid (m) (Figure 2.2.)
z : Distance from the reference level (m)
ρA : Volatile liquid density (kg/m3)
μ : Viscosity of the solution (kg/m.s)
: Association parameter of the solvent
λ : Mean free path

20
 Determination of diffusion coefficient for liquids and gases

6. REFERENCES

[1] Maton, Anthea; Jean Hopkins, Susan Johnson, David LaHart, Maryanna Quon Warner, Jill
D. Wright. Cells Building Blocks of Life. Upper Saddle River, New Jersey: Prentice Hall.
pp. 66–67 (1997).

[2] D. Brogioli and A. Vailati, Diffusive mass transfer by nonequilibrium fluctuations: Fick's
law revisited, Phys. Rev. E 63, 012105/1-4 (2001).

[3] http://encyclopedia.farlex.com/Fick%27s+first+law+of+diffusion

[4] “Determination of Diffusion Coefficient for Liquids and Gases” Practices Manual,
EDIBON, S.A.
[5] Geankoplis, C. J., “Transport Processes and Unit Operations”, 3rd Edition, p. 388 (2003).
[6] Wilke, C. R., and Chang, Pin. A.I.Ch.E.J., 1, 264 (1995).
[7] Reid, R. C., Prausnitz, J. M., and Sherwood, T.K. The properties of gases and liquids, 3rd
ed. Newyork: McGraw-Hill Book Company, (1977).
[8] Fuller, E.N., Schettler, P.D., and Giddings, J. C. Ind. Eng. Chem., 58, 19 (1996).
[9] http://www.jtbaker.com/msds/englishhtml/a0446.htm

21
 Determination of diffusion coefficient for liquids and gases

Name Surmane: Date:

Asistant:
Group number:

DATA SHEET

1. Determination of Diffusion Coefficients of Gases

T (oC) = T (oC) =

Time (min) Height (mm) Time (min) Height (mm)

2. Determination of Diffusion Coefficients of Liquids

C=1M C = 1,5 M
Time (min) Conductivity Time (min) Conductivity
(S/m) (S/m)

C= 2 M

Time (min) Conductivity

(S/m)

22