Diffusion of Liquids Through Stagnant, Non-Diffusing Air

P. L. Z. Soriano, V. D. C. Aguilar, A. C. G. Cabotaje, S. Y. Naval, G. R. Olvido

Chemical Engineering Department, Adamson University

Abstract: In this experiment, the mass diffusivities of three different volatile liquids were
determined with the aid of capillary tubes. The capillary tubes were filled with the three liquids.
For a certain time of 10 mins and 15 mins, the initial and final height of the liquid were measured
after the elapsed time. This was performed using varying temperatures of 50°C, 65°C and 80°C. It
has been found out that diffusivity is a function of temperature and molecular weight. The
diffusivities calculated using the empirical equations were compared to the data available in
Perry’s Handbook. It was observed that diffusivities obtained from Chapman-Enskog were the
nearest to the available data. The diffusivities from the experiment using the capillary method
exhibited a greater deviation from the available data.

Introduction

Diffusion is the movement, under the

influence of a physical stimulus, of an
individual component through a mixture.
Although the usual cause of diffusion is a
concentration gradient, diffusion can also be
caused by an activity gradient, as in reverse
osmosis, by a pressure gradient, by a
temperature gradient, or by the application of
Figure 1 shows that diffusion is the
an external force field, as in centrifuge.
movement of a solute from an area of high
Diffusion is not restricted to molecular
concentration to low concentration.
transfer through stagnant layers of solid or
fluid. Sometime diffusion process is In this experiment the diffusion
accompanied by bulk flow of the mixture in involves the mass transfer of a volatile
a direction parallel to the direction of component A through a non-diffusing
diffusion, and it is often associated with heat stagnant B. The most common cause
flow. In all mass-transfer operations, diffusion is concentration gradient of the
diffusion occurs in at least one phase and diffusing components. A concentration
often in both phases. (“Diffusion of liquids,” gradient tends to move the component in such
2014) a direction as to equalize concentration and
destroy the gradient while the gradient is
maintained by constantly supplying the

1

Figure 2. Diffusion of Liquid through
Stagnant Non-diffusing Air
diffusing component to the high
concentration end of the gradient and
removing it at low-concentration end. There
is steady state reflux of the diffusing
component. This is characteristics of many
mass transfer operations.
Diffusion is not restricted to
molecular transfer through stagnant layers of
solid or fluid. It also takes place when fluids
of different compositions are mixed. The first
step in mixing is often mass transfer called by
eddy motion characteristics of turbulent flow.
This is called eddy diffusion. The second step
is molecular diffusion between and inside the
very small eddies. Sometimes the diffusion
process is accompanied by bulk flow of the
mixture in a direction parallel to the direction
of diffusion.
Diffusivity is a proportionality
constant between the mass flux due to
molecular diffusion and the gradient in the
concentration of the species. It should be
2
apparent that the rate of molecular diffusion
in liquids is considerably slower in gases.
The molecules in liquid are very close
together compared to a gas. Hence, the
molecules of the diffusing solute A will
collide with molecules of liquid B more often
and diffuse more slowly that in gases. In
general, the diffusion coefficient in a gas will
be in order of magnitude of about times
greater than in liquid There are several
equations used for the prediction of
diffusivities. It is usually computed using the
general formula;
RTpBm ρA (z ′2 − zo 2 )
DAB =
2PMA (pA1 − pA2 )t ′
where pBm = log mean partial pressure of B
T= Absolute Temperature
ρA = Density of Liquid A
z ′ = height from the end of the tube at time t ′
zo = initial height from the end of the tube
pA1= partial pressure of A at point 1
pA2= partial pressure of A at point 2
Shown below are the empirical
correlations used for the prediction of
diffusivities.
Chapman-Enskog Equation. A more
accurate and rigorous treatment must
consider the intermolecular forces of
attraction and repulsion between molecules
and also different sizes of molecules A and B.
Chapman and Enskog solved the Boltzmann
equation, which does not utilize the mean
free path λ but uses a distribution function.
To solve the equation, a relation between the
attractive and repulsive forces between a V = atomic diffusion volumes
given pair of molecules must be used. For a
pair of nonpolar molecules a reasonable The general objective of this
approximation to the forces is the Lennard- experiment is to determine the mass
Jones function. The final relation to predict diffusivities of volatile organic liquids in air
the diffusivity of a binary gas pair of A and B at different temperature using capillary
molecules is method. Furthermore, the specific objective
3⁄ 1⁄ is to compare the results with those obtained
1.8583𝑥10−7 𝑇 2 1 1 2
from empirical equations (Chapman-Enskog
𝐷𝐴𝐵 = ( + )
𝑃𝜎𝐴𝐵 2 Ω𝐷 𝑀𝐴 𝑀𝐵 and Fuller) and available data from books.

DAB = diffusivity Materials

T = temperature
ΩD = collision integral The materials used in this experiment
P = pressure are as follows: three volatile organic liquids
MA,B = molecular weight (methanol, ethanol, and acetone), threes
σAB = Lennard-Jones force constant capillary tubes, timer, water bath, digital
thermometer, small electric fan and ruler.
Fuller et al., equation. Chapman Enskog
equation is relatively complicated to use and
often some of the constants such as σAB are Experimental Procedures
not available or difficult to estimate. Hence,
the semiempirical method of Fuller et al, Each capillary tube with equal length
which is much more convenient to use, is and diameter was filled with organic liquids
often utilized. The equation was obtained by and sealed at each end to prevent
correlating many recent data and uses atomic contamination with water. A water bath was
volumes which are summed for each gas prepared and allowed to reach a temperature
molecule. The equation is (Tompson & of 50°C. Each height of the liquid from the
Loyalka, 1987) end of the tube was measured using a ruler
before dipping in the bath. Receptacles were
1
1.0133x10−7 T1.75 1 1 2 also constructed to hold each tube during the
DAB = 2 [ + ] experiment. A gentle stream of air was
1 1 MA MB
3 3 passed above each tube using a fan while in
P [(∑ v)A + (∑ v)B ]
the water bath. Each height of the liquid from
T = temperature, in K the end of the tube was measured after 10
minutes and after 15 minutes. Same
MA and MB = molecular weights of A and B procedures were observed for trials 2 and 3
respectively but using temperature of 65°C and 80°C
respectively.
P = total pressure, in bar.

3
 Results and Discussion diffusivity calculated using fuller et al. is the
nearest to the data from Perry’s Handbook. It
was also observed that at 50°C, methanol is
Diffusivities in cm2/s at T = 50°C (323.15 K)
the fastest to diffuse among the three liquids.
This is due to that methanol has the lowest
Liquid 10 15 mins. Chapman-
molecular weight among the three, which is
mins. Enskog
32 g/mol. Like acetone, the diffusivity of
Ethanol 0.1436 0.1613 0.1383
methanol computed using fuller et al. is the
nearest to the true value from Perry’s
Acetone 0.0993 0.1132 0.1201
Handbook. These behaviors are affected by
different parameters such as temperature,
Methanol 0.1097 0.1351 0.1788
molecular weight, and structure. Compounds
having higher molecular weights have higher
Fuller Available
et. al Data atomic diffusion volumes. The atomic
Ethanol 0.1421 0.1369 diffusion volumes are used in the calculation
of diffusivities using fuller et. al.
Acetone 0.1219 0.1460

Methanol 0.1863 0.1771 Diffusivities in cm2/s at T = 65°C (338.15 K)

Table 1: Diffusivities at 50°C Liquid 10 15 mins. Chapman-

mins. Enskog
Table 1 showed the experimental and Ethanol 0.1502 0.1688 0.1481
computed diffusivities using different
empirical equations; Capillary Tube method, Acetone 0.1041 0.1185 0.1284
Chapman-Enskog, Fuller et. al. The available
data from Perry’s Chemical Engineers Methanol 0.1148 0.1414 0.1914
Handbook 8th edition were also utilized. For
ethanol, the experimental diffusivities Fuller Available
et. al Data
calculated using Capillary Tube Method were
Ethanol 0.1538 0.1482
quite near to the computed diffusivities using
the empirical equations. The diffusivity
Acetone 0.1400 0.1584
calculated using Chapman-Enskog is the
nearest to the data from Perry’s Handbook for
Methanol 0.2016 0.1918
ethanol. Acetone exhibited lower
diffusivities than ethanol and methanol. This Table 2: Diffusivities at 65°C
is due to that acetone has the highest
molecular weight among the three. The
higher the molecular weight of the liquid, the
slower it is to diffuse. In acetone, the

4
 Referring to table 2, it was observed that the
diffusivities from the experiment and
empirical equations are higher at 65°C than in
50°C. This simply implies that changing the
temperature will change the computed
diffusivity because it is a function of
temperature. According to table 2,
diffusivities of ethanol and methanol
calculated using Chapman-Enskog are the
nearest to the true value from the Perry’s
Handbook. On the other hand, diffusivity of
acetone calculated using Fuller et al. is the
nearest to the available data.
Diffusivities in cm2/s at T = 80°C (353.15 K)
Liquid 10 15 mins. Chapman-
mins. Enskog
Ethanol 0.1569 0.1763 0.1580
also observed that there is a noticeable
deviation of the experimental diffusivities
from the available data. These values show a
certain amount of deviation from the
available data garnered which can be
explained by the errors and shortcomings
accumulated throughout the experiment.
Comparing the results from the three
tables presented, it is found out that mass
diffusivities obtained from empirical
equations are precise to the available source.
It can be observed that there is a certain
deviation from the experimental and
available data which can be accounted by the
errors accumulated in the experiment such as
the non-available source of constant
magnitude and uniform distribution of hot air
through the use of a mini fan.

Acetone 0.1158 0.1267 0.1370 The diffusivities from Chapman-

Enskog is closer to the available data due to
Methanol 0.1199 0.1477 0.2043 the fact that most of its constants were
evaluated through previous experimental
Fuller Available data and covers parameters that are within the
et. al Data molecular level. Furthermore, the increase in
Ethanol 0.1659 0.1599 temperature also increases the mass
diffusivity. This is evident in the data
Acetone 0.1424 0.1709
presented because molecules exhibit a higher
kinetic energy at higher temperatures, thus
Methanol 0.2176 0.2069
making it to diffuse faster.
Table 3: Diffusivities at 80°C

In table 3, there is an observable

difference in the diffusivities computed due
to the increase in temperature. Since
diffusivity is a function of temperature,
higher values are probable. In this table, the
diffusivities of ethanol and methanol
calculated using Chapman-Enskog are once
again the nearest to the available data. It was
Conclusion
In this experiment, several findings
have been concluded. These are as follows:
 The diffusivity is a function of
temperature and molecular weight. It
is directly proportional to the

5
 temperature and inversely
proportional to the molecular weight.
 Capillary method can be used to
experimentally determine mass
diffusivity. The initial height and
final height of the liquid used is an
important parameter in this method.
On the other hand, diffusivities
calculated using this method have a
greater deviation from the available
data than those of empirical
equations, due to experimental errors.

 Among the empirical equations used,

Chapman-Enskog is closer to the
available data because its constants
were evaluated through previous
experimental data and covers
parameters that are within the
molecular level.

References

 Brown G. G. (1962). Unit Operation,

Asia publishing house, Calcutta, 117
 Coulson J. M., Richardson J. F.,
Buckhurst J., Harker J. H. (1962).
Coulson & Richardson Chemical
Engineering (volume 2), Asian Book
Pvt. Ltd., 131
 Diffusion of liquids. (2014), (1).
 Green & Perry. Perry’s Chemical
Engineers’ Handbook. McGraw-Hill
Inc.; New York, 8th Edition.