API TECHNICAL DATA BOOK – 9th Edition

API® TECHNICAL DATA BOOK

9th Edition

CHAPTER 13

DIFFUSIVITY

Revised Chapter 13 to First Edition (1966),

Second Edition (1970), Third Edition (1976),
Seventh Edition (2005) and Ninth Edition (2013)

© 2013, The American Petroleum Institute and EPCON International - All Rights Reserved
 API TECHNICAL DATA BOOK – 9th Edition

Table of Contents
Chapter 13 – Diffusivity

Page

13-0 Introduction ............................................................................................................................................. 13-1

13A Diffusivity in Liquid Systems ................................................................................................................ 13-4

13A1 Binary Liquid Systems ................................................................................................................ 13-4
Procedure 13A1.1 Diffusion Coefficients in Binary Nonpolar Liquid Systems
at High Dilution of the Solute ................................................................. 13-4
Table 13A1.2 Radius of Gyration .................................................................................. 13-6
Procedure 13A1.3 Diffusion Coefficients in Binary Polar or Associating Liquid
Systems at High Dilution of the Solute ................................................... 13-7
Figure 13A1.4 Diffusion Coefficients in Dilute Binary Polar or
Associating Liquids ............................................................... …………..13-9
Procedure 13A1.5 Diffusion Coefficients in Concentrated Binary Nonpolar
Liquid Systems .................................................................................... ..13-11

13A2 Multicomponent Liquid Systems ............................................................................................... 13-13

Procedure 13A2.1 Diffusion Coefficients in Multicomponent Liquid
Systems ............................................................................................ ….13-13

13B Diffusivity in Gas Systems ................................................................................................................... 13-15

13B1 Binary Gas Systems ................................................................................................................... 13-15
Procedure 13B1.1 Diffusion Coefficients in Binary Gas Systems at Low
Pressures - Hydrocarbon/Hydrocarbon Systems .................................... 13-15
Procedure 13B1.2 Diffusion Coefficients in Binary Gas Systems at Low
Pressures - Air/Hydrocarbon Systems .................................................... 13-17
Table 13B1.3 Atomic Diffusion Volumes for Use in Procedure
13B1.2 …….. ........................................................................................ 13-19
Figure 13B1.4 Diffusion Coefficients in Gas Binary Systems at
High Pressure ......................................................................................... 13-20

13B2 Multicomponent Gas Systems ................................................................................................... 13-22

Procedure 13B2.1 Diffusion Coefficients in Multicomponent Gas Systems ....................... 13-22

13C Diffusivity in Gas-Liquid Systems....................................................................................................... 13-24

13C1 Gas – Liquid Systems .................................................................................................................. 13-24
Procedure 13C1.1 Diffusion Coefficients of Dissolved Gases in Liquids........................... 13-24

Bibliography ...................................................................................................................................................... 13-26

13 - ii © 2013, The American Petroleum Institute and EPCON International - All Rights Reserved
 API TECHNICAL DATA BOOK – 9th Edition

Preface
Chapter 13 – Diffusivity
(1980)

The previous edition of Chapter 13, "Diffusivity," was prepared in 1964. Since that time considerable new data
and a number of improved correlation techniques for diffusivities have been published. Thus, in 1979 a complete
revision of the diffusivity chapter was begun. The resulting chapter covers more types of systems and provides
methods which yield lower average errors than the previously published chapter. Detailed results for all the
evaluations used to select the contents of the revised chapter are given in Documentation Report 13-80 available
from University Microfilms, Ann Arbor, Michigan.
Work on this chapter was carried out by N. O. Umesi in conjunction with the project directors, R. P. Danner and
T. E. Daubert. The work was reviewed periodically by the Subcommittee on Technical Data of the American
Petroleum Institute.

© 2013, The American Petroleum Institute and EPCON International - All Rights Reserved 13 - iii
 API TECHNICAL DATA BOOK – 9th Edition

13-0 Introduction
Chapter 13 – Diffusivity
(1980)
Diffusion of mass is the dissipation of a the liquid phase. The correlations that were
concentration (or chemical potential) gradient by developed from these data necessarily reflect this
molecular transport without overall mass flow. The experimental uncertainty. Although the reliability
diffusion coefficient is the proportionality constant quotations for the correlations in this edition are not
between the diffusion rate and the concentration low, they are superior to those quoted in the last
gradient causing diffusion. It is generally defined by edition. The estimated diffusion coefficients are
Fick's first law for unidirectional, binary-system suitable for most engineering applications.
diffusion. This equation, which follows, is valid only
in nonturbulent systems: Liquid Systems

dC1 Because no single correlation is satisfactory for

W1   D1,2 (13-0.1) estimating properties for all systems involving
dL
liquid solutions, two correlations are presented for
the estimation of diffusion coefficients in dilute
Where: binary liquid mixtures. In some molecules, uneven
spatial distribution of electronic charges about the
W1 = molar flow of solute species 1 in lb-mole positively charged nucleus gives these molecules
per (hour) (ft2) permanent dipoles. These molecules are, by
D1,2 = diffusion coefficient of solute species 1 in definition, polar. Typical polar compounds include
solvent species 2 in ft2/hr acetone, acetaldehyde, and dimethyl ether. The
C1 = concentration of solute species 1 in lb- dipole-dipole interaction from these permanent
mole/ft3 dipole moments is not always an adequate
L = distance in ft description of the interactions between polar
molecules because other localized effects do occur.
On differentiation, the rate of change of local For instance, the molecules of some pure polar
concentration with time (Fick's second law) results liquids such as water, carboxylic acids and alcohols
in: occur in aggregated forms due to specific forces
such as hydrogen bonding. These compounds are
dC1 d 2C1 (13-0.2) called associating compounds.
 D1,2 Symmetrical molecules, such as carbon
dt dL2
tetrachloride, benzene and methane have dipole
Where: t = time in hours moments of zero. These substances are called
nonpolar molecules. Using these definitions,
Although other diffusion coefficients have been Procedure 13A1.1 is presented for the estimation of
defined, the Fick diffusion coefficient is the most diffusion coefficients in binary nonpolar liquid
common measure of diffusional tendency. systems at high dilution of solute. Procedure 13A1.3
Accordingly, only Fick ordinary diffusion is given for the estimation of diffusion coefficients
coefficients are predicted by the methods in this in binary systems at high dilution of solute when
chapter. either the solvent is polar for associating and the
The primary application of diffusion coefficients solute is nonpolar or both components are polar (or
in chemical engineering calculations is in the associating). Procedures 13A1.5 and 13A2.1 are
Schmidt number, μ/ρD, which is used generally to presented as the recommended methods for
correlate mass transfer properties. The equations are estimating diffusivities in concentrated solutions
useful in reaction rate calculations for which the rate and liquid mixtures of more than two components,
of diffusion to a catalyst is important. respectively. These procedures are semiempirical in
Unfortunately, experimental diffusion nature. The more theoretical methods are not
coefficients are usually not precise, particularly in sufficiently accurate for general utility.

© 2013, The American Petroleum Institute and EPCON International - All Rights Reserved 13 - 1
 API TECHNICAL DATA BOOK – 9th Edition

13-0 Introduction
Chapter 13 – Diffusivity
(1980)
Gaseous Systems diffusion coefficients can be combined, using
Procedure 13B2.1 to obtain a multicomponent
Two semiempirical correlations are given for diffusion coefficient value that is satisfactory for
estimating diffusion coefficients in low-pressure most engineering applications.
binary gaseous systems. Procedure 13B1.1 is
recommended for binary hydrocarbon-hydrocarbon Gas-Liquid Systems
systems and Procedure 13B1.2 is recommended for
estimating diffusion coefficients in binary Procedure 13C1.1 is recommended for
air-hydrocarbon systems. estimating the diffusion coefficients of dissolved
A correction chart for high pressures, Figure gases diffusing into liquids. It is the same method
13B1.4, is more empirical in nature and is retained used in Procedure 13A1.1. The correlation was
from the last edition. This pressure correction term, developed for use with gas-liquid systems and with
although based on very few data, predicts diffusion nonpolar liquid-liquid systems (at high dilution of the
coefficient values that are suitable for most solute).
engineering applications. The binary gaseous

13 - 2 © 2013, The American Petroleum Institute and EPCON International - All Rights Reserved
 API TECHNICAL DATA BOOK – 9th Edition

Computer Methods
Chapter 13 – Diffusivity
(2005)

Program Procedure Description

Diffusion 13A1.1 Diffusion Coefficients in Binary Nonpolar Liquid
Coefficients Systems at High Dilution of the Solute
Diffusion 13B1.1 Diffusion Coefficients in Binary Gas Systems at Low
Coefficients Pressures Hydrocarbon/Hydrocarbon Systems
Diffusion 13B2.1 Diffusion Coefficients in Multicomponent Gas
Coefficients Systems

© 2013, The American Petroleum Institute and EPCON International - All Rights Reserved 13 - 3
 API TECHNICAL DATA BOOK – 9th Edition

13A Diffusivity in Liquid Systems

13A1 Binary Liquid Systems

Procedure 13A1.1 – Diffusion Coefficients in Binary Nonpolar Liquid Systems at High Dilution of
the Solute

(1980)

Discussion
The following equation is recommended for estimating diffusion coefficients in binary nonpolar liquid systems
at high dilution of solute.

8 T R2 (13A1.1-1)
D1,2  5.922  10 2/3
 R1

Where:

D1,2 = diffusion coefficient of solute (component 1) in solvent (component 2) in ft2/hr

T = temperature in degrees Rankine

μ = viscosity of solution (considered as pure solvent) in centipoise

R 2 = radius of gyration of solvent in Ångstrom units

R 1 = radius of gyration of solute, in Ångstrom units

Procedure

Step 1: Obtain R 1 and R 2 from Table 13A1.2. In the absence of experimental data, estimate μ using the Pure
Component Properties program.
Step 2: Calculate D1‚2 from equation (13Al.l -1).

13 - 4 © 2013, The American Petroleum Institute and EPCON International - All Rights Reserved
 API TECHNICAL DATA BOOK – 9th Edition

Comments on Procedure 13A1.1

(1980)

Purpose
Procedure 13A1.1 is presented primarily for the estimation of diffusion coefficients in dilute (less than 5 mole
percent solute) binary nonpolar liquid solutions (for example, solutions of hydrocarbons and solutions of carbon
tetrachloride and hydrocarbons). Indirectly, it is also used to estimate diffusion coefficients in nonpolar concentrated
solutions (with Procedure 13A1.5) and nonpolar multicomponent liquid solutions (with Procedure 13A2.1).

Limitations
Equation (13A1.1-1) was developed from hydrocarbon/hydrocarbon and hydrocarbon/carbon tetrachloride data
at a solute concentration less than 5 mole percent. For systems involving water and other polar or associating
compounds, use Procedure 13A1.3.

Reliability
For the systems to which this equation applies, experimental data are reproduced to within an average of 16
percent with occasional errors of 30 percent. Umesi (11) has shown that this equation gives good results for polar
solutes diffusing into nonpolar solvents. For polar-polar systems, however, the calculated coefficients tend to be 15
to 50 percent too low.

Special Comments
For concentrated solutions (more than 5 mole percent solute) the coefficients from this procedure must be
combined using Procedure 13A1.5. For multicomponent solutions the use of Procedure 13A2.1 is recommended. A
table of R values is given in Reid et al. (7) for 250 compounds.

Literature Source
Equation (13A1.1-1) was developed by Umesi (M.S. Thesis, The Pennsylvania State University, 1980).

Example
Estimate the diffusion coefficient of n-decane in a binary liquid solution of n-decane and n-heptane at 77 °F.
The viscosity of the solution (considered as the viscosity of pure n-heptane) at 77 °F is 0.3984 centipoise.
From Table 13A1.2, the radius of gyration of n-heptane is 4.2665 Å and that of n-decane is 5.5390 Å. The
diffusion coefficient is calculated using equation (13A1.1-1).

D 1,2 
5.922 10   77  459.7  4.2665  1.09 10
8
4
ft 2 per hr
 0.3984 5.5390
2/3

An experimental value of 1.19×10–4 ft2/hr is available. The predicted value represents an absolute deviation of
8.4 percent from the experimental value.

© 2013, The American Petroleum Institute and EPCON International - All Rights Reserved 13 - 5
 API TECHNICAL DATA BOOK – 9th Edition
Table 13A1.2 – Radius of Gyration
(1980)
Radlus of Gyration Radius of Gyration
(Angstroms) (Angstroms)
Paraffins
Melhane 1.1234 !-Butene 2.7458
Ethane 1.8314 CIS-2-Butene 2.7765
Propane 2.4255 rrans-2-Butene 2.7123
n-Butane 2.8885 2-Melhylpropene 2.8281
2-Methylpropane 2.8962 1-Pentene 3.1956

n-Pentane 3.3850 cis-2-Pentene 3.2763

2-Methylbutane 3.3130 rrans-2-Pentene 3.2826
2.2-Dimelhylpropane 3.1530 2-Methyl-1-butene 3.2239
n-Hexane 3.8120 2-Methyl-2-butene 3.2301
2-Methylpentane 3.8090 1-Hexene 3.6472

3-Methylpentane 3.6797 rrans-2-Hexene 3.6964

2,2-Dimethylbutane 3.4846 1-Heptene 4.0971
2,3-Dimethylbutane 3.5209 1-0ctene 4.5342
n-Heptane 4.2665 1-Nonene 4.9687
2-Methylhexane 4.2779 1-Decene 5.4017

3-Methylhexane 4.1454 1-Undecene 5.8389

2.2-Dimethylpentane 4.0001 1-Dodecene 6.2966
2,3-Dimethylpentane 3.9210 1-Tridecene 6.7514
2,4-Dimethylpentane 3.9634 1-Tetradecene 7.2244
3.3-Dimethylpentane 3.7952 I -Pentadecene 7.6969

2 ,2,3-Trimethylbutane 3.6960 1-Hexadecene 8.1872

n-Octane 4.6804 1,2-Butadiene 2.7497
2-Methylheptane 4.7401 2.3-Pentadiene 3.0552
3-Methylheptane 4.5932 Ethyne (acetylene) 1.1095
4-Methylheptane 4.5581 Propyne 1.8864
1-Buryne 2.7130
2,2-Dimethylhexane 4.4956
2.3-Dimethylhexane 4.4084 Aromatics
2,4-Dimethylhexane 4.3463 Benzene 3.0037
2,5-Dimethylhexane 4.5932 Methyl benzene 3.4431
3,3-Dimethylhexane 4.3197 Ethyl benzene 3.8211
1.2-Dimethylbenzene 3.7889
3,4-Dimethylhexane 4.4000
1,3-Dimethy!benzene 3.8966
2,2.3-Trimethylpentane 4.1618
2,2 .4-Trimethylpentane 4.1714 I ,4-Dimelhyl benzene 3.7962
2,3,3-Trimethylpentane 4.0859 lsopropylbenzene 4.1870
2,3,4-Trimethylpentane 4.2052 1-Methyl-2-ethylbenzen�: 4.1296
I -Methyl-3-elhylbenzen�: 4.2845
n-Nonane 5.1263
1-Methyl-4-ethylbenzenc: 4.1662
2.2.3,3-Tetramethylpentane 4.1556
n-Decane 5.5390 1.2.3-Trimethylbenzene 4.0996
n-Undecane 5.9867 1.2,4-Trimethylbenzene 4.1678
n-Dodecane 6.4321 1.3,5- Trimelhylbenzene 4.3408
n-Tetradecane 7.3578 1-Methyl-3-isopropylbemzeoe 4.5790
n-Pentadecane 7.8387 1-Methyl-4-isopropylbenzeoe 4.5231
n-Hexadecane 8.3180
1,2,3,4-Tettamethylbem�ne 4.3779
Napthenes 1,2,3,5-Tetramethylbem:ene 4.4920
Cyclohexane 3.2605 1.2.4.5-Tetramethylben;:ene 4.4512
Methylcyclohexane 3.7467
H alogenared Compounds
1.1-Dimethylcyclohexane 4.0925
Trichloromethane 3.1779
cis-1,2-Dimethylcyclohexane 4.0612
Tetrachloromelhaoe 3.4581
rrans-1.2-Dimethylcyclohexane 4.1814
Chloroethane 2.2812
cis-! ,3-Dimethylcyclohexane 4.0549 I. I-Dichloroelhane 2.9945
rrans-1.3-Dimethylcyclohexane 4.1462 1,2-Dichloroethane 2.8510
cis-1,4-Dimethylcyclohexane 4.1446
I , I, I- Trichloroethane 3.3566
rrans-1.4-Dimethylcyclohexane 4.1670
Chloroethene 2.1220
0/efins cis-1.2-Dichloroethene 3.0132
Ethene 1.5382 trans-1.2-Dichloroethen,e 3.0132
Propene 2.2283 Trichloroethene 3.7592

13 - 6 © 2013, The American Petroleum Institute and EPCON International - All Rights Reserved
 API TECHNICAL DATA BOOK – 9th Edition

Procedure 13A1.3 – Diffusion Coefficients in Binary Polar or

Associating Liquid Systems at High Dilution of the Solute
(1980)

Discussion
The following equation is recommended for estimating diffusion coefficients in dilute binary liquid systems
involving polar or associating solvents. The equation may be used directly or approximated conveniently with the
nomograph in Figure 13A1.4.

2 M 2  T
1/2
(13A1.3-1)
D 1,2  1.59 107
 V1 
0.6

Where:

2 = association parameter for the solvent

M2 = molecular weight of the solvent
V 1 = molar volume of solute at the normal boiling point given by V1= 0.285 Vc11.048 in cc/g-mole
Vc = critical volume in cc/g-mole
μ = viscosity of the pure solvent in centipoise
T = temperature in degrees Rankine

Procedure
Step1: Obtain M2 and Vc1 from the Pure Component Properties program and V1 from the equation given above.
In the absence of experimental data, estimate μ using the procedures of Chapter 11.
Step 2: Determine the association parameter, 2 , for the solvent from the tabulation below:

2
Ethyl Alcohol 1.5
Methyl Alcohol 1.9
Water 2.6
Other Associating Solvents 1.0 -2.6

Step 3: Obtain D1,2 directly from equation (13A1.3-1) or with nomograph, Figure 13A1.4.

© 2013, The American Petroleum Institute and EPCON International - All Rights Reserved 13 - 7
 API TECHNICAL DATA BOOK – 9th Edition

Comments on Procedure 13A1.3

(1992)

Purpose
Procedure 13A1.3 is presented primarily for the estimation of diffusion coefficients in dilute (less than 5 mole
percent solute) polar or associating binary liquid solutions. These coefficients are also required to estimate
diffusivities in polar or associating concentrated solutions (with Procedure 13A1.5) and polar or associating
multicomponent liquid solutions (with Procedure 13A2.1).

Limitations
For most associating systems other than methanol, ethanol and water, the association parameter is arbitrarily
assigned according to an estimate of the degree of association of the solvent (see step 2). The diffusion coefficient
estimated with this equation is not very accurate when more than 5 mole percent solute is present in the solution.

Reliability
Equation (13A1.3-1) reproduces experimental data of associating binary liquid systems with an average
deviation of 20 percent with occasional errors of 35 percent. It should not be used for systems in which the solvent is
a hydrocarbon because calculated coefficients are too small when compared to the experimental values. Higher
errors occur when the diffusing molecule is much smaller than the solvent.

Special Comments
For concentrated solutions (more than 5 mole percent solute) of polar or associating systems, the coefficients
from this procedure must be combined using Procedure 13A1.5. For multicomponent solutions of polar or
associating systems, Procedure 13A2.1 is appropriate.

Literature Sources
Equation (13A1.3-1) was developed by Wilke and Chang, AIChE Journal 1 264 (1955). The correlation
between molar volume at the normal boiling point and the critical molar volumes was developed by Tyn and Calus,
Hydrocarbon Processing 21 [4] 16 (1975).

Example
Estimate the diffusion coefficient of benzene in a binary liquid solution of benzene and water at 68 °F.
The viscosity of the solution (considered as the viscosity of pure water) at 68 °F is 0.95 centipoise.
From the Pure Component Properties Program, M2 =18.015. The molar volume of benzene at the normal boiling
point is 96.35 cc/g-mole. For water, the association factor is 2.6. The diffusion coefficient is calculated using
equation (13A1.3-1)

D1‚2 = 1.59×10–7 (2.6 × 18.015)1/2 (68 +459.7)

0.95 (96.35)0.6

= 3.90×10–5 ft2/hr

An experimental value of 4.57×10–5 ft2/hr is available; this difference translates to an absolute deviation of 14.7
percent.

13 - 8 © 2013, The American Petroleum Institute and EPCON International - All Rights Reserved
 API TECHNICAL DATA BOOK – 9th Edition
Figure 13A1.4 – Diffusion Coefficients in Dilute Binary Polar or Associating Liquids
(1980)

FIGURE 13A1.4
DIFFUSION COEFFICIENTS IN
DILUTE BINARY POLAR OR
1
ASSOCIATING LIQUIDS
Created: June 1980 Remastered : 2012

TECHNICAL DATA BOOK VERSION 9

1.5
Copyright © 2013. The American Petroleum Institute and EPCON International, Inc.

0.04 2

M O L A R V O L U M E A T T H E N O R M A L B O I L I N G POINT , cm /gmole)
0.03
(SOLVENT MOLECULAR WEIGHT) (ASSOCIATION PARAMETER)

0.02

3
/T, (VISCOSITY OF SOLUTION) / (TEMPERATURE), cP / deg R

D 1, 2 ‘ D I F F U S I O N C O E F F I C I E N T , s q f t / h r x 1 0
0.01 4
0.009
0.008
0.007
1000 5
0.006
800
0.005 600 6
0.004
400
7
0.003
REFERENCE

200 8
0.002 9
100 10
80
60
0.001 40
0.0009
0.0008
0.0007 15
20
0.0006
0.0005
0.0004 10 20

0.0003
2’
∅

0.0002
M2

V1 , ( S O L U T E

30

0.0001 40

50

60
NOTE: PART OF PROCEDURE 13A1.3
NOMOGRAPH REPRESENTS EQ. (13A1.3-1) 70

80
90
100

© 2013, The American Petroleum Institute and EPCON International - All Rights Reserved 13 - 9
 API TECHNICAL DATA BOOK – 9th Edition

Comments on Figure 13A1.4

(1980)

Purpose
Figure 13A1.4 is a convenient representation of equation (13A1.3-1) for estimating diffusion coefficients in
dilute (less than 5 mole percent solute) polar or associating binary liquid solutions.

Limitations
For most associating systems other than methanol, ethanol and water, the association parameter has to be
arbitrarily assigned according to an estimate of the degree of association of the solvent (see step 2 of Procedure
13A1.3). The diffusion coefficient estimated with this equation is not very accurate when more than 5 mole percent
solute is present in the solution.

Notation
μ = viscosity of solution (considered as pure solvent) in centipoise
T = temperature in degrees Rankine
M2 = molecular weight of solvent
2 = association parameter for the solvent
V1 = solute molar volume at the normal boiling point in cc/g-mole
D1‚2 = diffusion coefficient of solute (component 1) in solvent (component 2) in ft2/hr

Reliability
Figure 13A1.4 is a representation of equation (13A1.3-1) which reproduces experimental data for polar or
associating binary liquid systems to an average deviation of 25 percent with occasional errors of 50 percent. An
additional uncertainty of approximately 5 percent is introduced with the nomograph. This figure should not be used
for systems in which the solvent is a hydrocarbon.

Special Comments
For concentrated solutions (more than 5 mole percent solute) of polar or associating systems, the coefficients
from this figure must be combined using Procedure 13A1.5. For multicomponent solutions of polar or associating
systems, Procedure 13A2.1 is necessary.

Literature Source
This nomograph was adapted from Kuong, Chem. Eng. 68 [12] 258 (1961).

Example
Estimate the diffusion coefficient of benzene in a binary liquid solution of benzene and water at 68 °F. To
obtain a solution, follow the steps outlined in Procedure 13A1.3 with the values of the parameters as given. The
nomograph parameters are calculated below:

 0.95 cP
  0.00180
T 527.7 deg R
2 M 2  46.84
V1  96.38 cc/g-mole

Connect 0.00180 on the /T scale of Figure 13A1.4 with 46.84 on the M 22 scale, and locate a point on the
reference line. Connect this reference line with 96.38 on the V1 scale and read the diffusion coefficient on the D1‚2
scale.
The estimated coefficient is 4.02×10–5 ft2/hr, which compares favorably with the value of 3.90×10–5 ft2/hr
calculated from equation (13A1.3-1). The experimental value is 4.57×10–5 ft2/hr.

13 - 10 © 2013, The American Petroleum Institute and EPCON International - All Rights Reserved
 API TECHNICAL DATA BOOK – 9th Edition

Procedure 13A1.5 – Diffusion Coefficients in Concentrated Binary Nonpolar Liquid Systems

(1980)

Discussion
For concentrated binary liquid solutions (more than 5 mole percent solute) Procedures 13A1.1 and 13A1.3 are
not accurate. The following equation should be used:

o
D1,m  x1D2,1  1 – x1  D1,2
o
(13A1.5-1)

Where:

D = diffusion coefficient (D1, m for the concentrated solution, and D°1‚2 and D°2‚1 for the dilute
solution), in ft2/hr
x = mole fraction
Subscripts 1 and 2 = species 1 and 2.

There are no viscosity terms, only mole fractions of the components and diffusion coefficients of one
component in the other as the concentration of the diffusing component approaches zero.

Procedure
Step 1: If experimental data are not available, estimate the dilute binary coefficients with Procedure 13A1.1
or 13A1.3 (depending on the chemical nature of the system).
Step 2: Calculate the binary diffusion coefficient in the concentrated mixture using equation (13A1.5-1).

© 2013, The American Petroleum Institute and EPCON International - All Rights Reserved 13 - 11
 API TECHNICAL DATA BOOK – 9th Edition

Comments on Procedure 13A1.5

(1980)

Purpose
Equation (13A1.5-1) is a correction applied to dilute solution diffusion coefficients when dealing with
concentrated binary solutions (more than 5 mole percent solute).

Limitations and Reliability

In addition to being entirely empirical, equation (13A1.5-1) is based on very few data. Use of it for solutions
containing polar or associating compounds may result in substantial error. However, when used for hydrocarbon
systems in conjunction with procedure 13A1.1, experimental data are reproduced with an average deviation of 14.0
percent.

Literature Source
Equation (13A1.5-1) was developed by Caldwell and Babb, J. Phys. Chem. 60 51 (1956) and cited by Sanchez
and Clifton, Ind. Eng. Chem. Fundam. 16 318 (1977).

Example
Estimate the diffusion coefficient of methylbenzene (1) in a binary liquid solution composed of 0.40 mole
fraction methylbenzene and 0.60 mole fraction cyclohexane (2) at 104 °F and 14.7 psia.
The dilute binary solution diffusion coefficients are obtained from Procedure 13A1.1 as 7.32×10–5 ft2/hr for
methylbenzene in cyclohexane at 104 °F and 1.14×10–4 ft2/hr for cyclohexane in methylbenzene at 104 °F (see
Procedure 13A1.1).
The concentrated solution diffusion coefficient is then calculated using equation (13A1.5-1):

D1,m = [0.40 (11.4) + 0.60 (7.32)] × 10–5

= 8.95×10–5 ft2/hr

An experimental value of 8.76×10–5 ft2/hr is available. This represents an absolute deviation of 2.2 percent.

13 - 12 © 2013, The American Petroleum Institute and EPCON International - All Rights Reserved
 API TECHNICAL DATA BOOK – 9th Edition

13A2 Multicomponent Liquid Systems

Procedure 13A2.1 – Diffusion Coefficients in Multicomponent Liquid Systems

(1980)

Discussion
The following equation estimates diffusion coefficients in multicomponent liquid systems consisting of
nonpolar components from the binary coefficients of the component pairs (with reference to the dilute solute) in the
mixture:

x2 x3
D1,m m   D1,2

2  D

1,3 3  (13A2.1-1)

Where:

D1,m = diffusion coefficient for the diffusion of solute l with reference to the multicomponent
mixture in ft2/hr

D1, j = diffusion coefficient of the binary pairs (j = 2, 3, … ) at infinite dilution of the solute in
ft2/hr (obtained from Procedure 13A1.1 or 13A1.3)
xj = mole fraction of component j in the mixture
μj = viscosity of component j in centipoise
μm = viscosity of mixture in centipoise

The binary pairs referred to are the solute and individual components of the mixture. This formulation assumes
that the solute is dilute and that there are no concentration gradients for the solvent species. One can, therefore,
speak of a single solute diffusivity D1,m with respect to the mixture.

Procedure
Step 1: If experimental data are not available, estimate the dilute binary diffusion coefficients with
Procedures 13A1.1 or 13A1.3 (depending on the chemical nature of the system).
Step 2: Obtain the pure-component and mixture viscosities from Chapter 11 if no experimental viscosity
values are available.
Step 3: Using equation (13A2.1-1), calculate the mixture diffusion coefficient (i.e., the single solute
diffusivity with respect to the mixture).

© 2013, The American Petroleum Institute and EPCON International - All Rights Reserved 13 - 13
 API TECHNICAL DATA BOOK – 9th Edition

Comments on Procedure 13A2.1

(1980)

Purpose
Equation (13A2.1-1) is recommended for the estimation of diffusion coefficients in multicomponent liquid
systems from the binary coefficients of the component pairs (see Procedure 13A1.1).

Limitations and Reliability

Equation (13A2.1-1) has a semiempirical basis but has been tested with very few data points for nonpolar
systems involving three components. Use of the equation for solutions containing polar or associating compounds
may result in substantial errors.

Literature Source
Equation (13A2.1-1) was developed by Leffler and Cullinan, Ind. Eng. Chem. Fundam. 9 88 (1970).

Example
Estimate the diffusion coefficient of toluene in a mixture consisting of 0.786 mole n-hexane and 0.214 mole
cyclohexane at 77 °F. The viscosity of n-hexane at 77 °F is 0.2985 centipoise and that of cyclohexane at 77 °F is 0.898
centipoise. The mixture viscosity which may be estimated from the recommendations of Chapter 11 is 0.3918
centipoise. The infinite dilute solution binary diffusion coefficients are obtained from Procedure 13A1.1 as 1.78×10–4
ft2/hr for toluene in n-hexane at 77 °F and 5.08×10–5 ft2/hr for toluene in cyclohexane at 77 °F.

The multicomponent diffusion coefficient is then calculated using equation (13A2.1-1):

D1,m 
1.78 10

4
  0.2985
0.786
 5.08 105   0.898 
 
0.214

0.3918
 1.31104 ft 2 /hr

An experimental value of 1.59×10–4 ft2/hr is available.

13 - 14 © 2013, The American Petroleum Institute and EPCON International - All Rights Reserved
 API TECHNICAL DATA BOOK – 9th Edition

13B Diffusivity in Gas Systems

13B1 Binary Gas Systems

Procedure 13B1.1 – Diffusion Coefficients in Binary Gas Systems at Low Pressures -

Hydrocarbon/Hydrocarbon Systems

(1980)

Discussion
The following equation is recommended for estimating diffusion coefficients in binary gas systems of
hydrocarbon/hydrocarbon components at pressures less than 500 psia.

0.5
 1 1 
0.1014 T 1.5   
M
 1 M 2 (13B1.1-1)
D1,2 
p V11/3  V21/3 
2

Where:
M1 = molecular weight of gas component 1

M2 = molecular weight of gas component 2

T = temperature in degrees Rankine

D1‚2 = diffusion coefficient of component 1 in component 2 in ft2/hr

p = pressure in psia

Vi = molar volume of component i at the normal boiling point, given by Vi = 0.285 Vci1.048 (i = 1,2)
in cc/g-mole

Procedure
Step 1: From the Pure Component Properties Program, obtain the molecular weights, M1, M2 of components
1 and 2 and the critical volumes Vc1 and Vc2.
Step 2: Calculate the molar volume at the normal boiling point.
Step 3: With the temperature and pressure known, calculate the diffusion coefficient using equation
(13B1.1-1).

© 2013, The American Petroleum Institute and EPCON International - All Rights Reserved 13 - 15
 API TECHNICAL DATA BOOK – 9th Edition

Comments on Procedure 13B1.1

(1980)

Purpose
Procedure (13B1.l) is presented primarily to estimate diffusion coefficients in binary vapor systems consisting
of hydrocarbon/hydrocarbon components at pressures less than 500 psia.

Limitations
Equation (13B1.1-1) is not accurate for pressures greater than 500 psia. The constant gives the best fit to the
data available and hence is restricted to systems consisting of hydrocarbon/hydrocarbon components.

Reliability
Equation (13B1.1-1) is an empirical equation which does not require use of the collision diameter that is
generally associated with more theoretical methods based on the kinetic theory of gases. Nevertheless, it reproduces
experimental hydrocarbon-hydrocarbon diffusivity data to an average deviation of not more than 4 percent. Also, the
parameters associated with equation (13Bl.l-1) are easier to obtain than those associated with mare theoretical
methods.

Special Comments
If the desired pressure is greater than 500 psia, use a convenient low pressure in equation (13B1.1-1) and apply
the pressure correction (Figure 13B1.4) to the resulting diffusion coefficients. For diffusion in a multicomponent
vapor, combine the binary coefficients using Procedure 13B2.1. For systems involving air, use Procedure 13B1.2.

Literature Sources
Equation (13B1.1-1) was developed by Gilliland, Ind Eng. Chem. 26 681 (1934). The correlation between the
molar volume at the normal boiling point and the critical molar volume was developed by Tyn and Calus,
Hydrocarbon Processing 21 [4] 16 (1975).

Example
Estimate the diffusion coefficient of n-hexane vapor diffusing into methane at 50 °F and 14.7 psia.
From the Pure Component Properties Program, the molecular weights are: n-hexane (M1) = 86.18 and methane
(M2) = 16.04. The critical molar volumes are n-hexane (Vc1) = 370.17 cc/g-mole and methane (Vc2) = 99.23 cc/g-
mole. The molar volumes at the normal boiling points are n-hexane (V1) = 140.13 cc/g-mole and methane (V2) =
35.27 cc/g-mole.
The diffusion coefficient is calculated using equation (13B1.1-1):

0.5
1.5  1 1 
0.1014  50  459.7   
D1,2   86.18 16.04 
1/3 2
14.7 140.13   35.27  
1/3
 
2
 0.301 ft /hr

An experimental value of 0.2914 ft2/hr is available. This estimated value represents a deviation of 3.3 percent
from the experimental value.

13 - 16 © 2013, The American Petroleum Institute and EPCON International - All Rights Reserved
 API TECHNICAL DATA BOOK – 9th Edition

Procedure 13B1.2 – Diffusion Coefficients in Binary Gas Systems at Low Pressures -

Air/Hydrocarbon Systems
(1980)

Discussion
The following equation is recommended for the estimation of diffusion coefficients in binary gas systems of
air/hydrocarbon components at pressures less than 500 psia.

0.5
 1 1 
0.0204 T 1.75    (13B1.2-1)
D1,2   M1 M 2 

   
1/3 2
 1/3

P   vi   vi 
 1 2 

Where:

Σjvi = sum of atomic diffusion volumes of component j (see Table 13B1.3)

T = temperature in degrees Rankine

p = pressure in psia

D1,2 = diffusion coefficient of component 1 in component 2 in ft2/hr

Procedure
Step 1: From the Pure Component Properties Program, obtain the molecular weights, M1, and M2.
Step 2: From Table 13Bl.3, obtain Σ j vi .
Step 3: With the temperature and pressure known, calculate the diffusion coefficient using equation
(13B1.2-1).

© 2013, The American Petroleum Institute and EPCON International - All Rights Reserved 13 - 17
 API TECHNICAL DATA BOOK – 9th Edition
Comments on Procedure 13B1.3
(1980)

Purpose
Procedure 13B1.2 is presented primarily for the estimation of diffusion coefficients in binary vapor systems
consisting of air/hydrocarbon components at pressures less than 500 psia.

Limitations
Equation (13B1.2-1) is not accurate for pressures greater than 500 psia. The constants were chosen to give the
best fit to the data used in the development of the correlation. Thus, the results can be no more reliable than the
diffusion data.

Reliability
Its empirical nature notwithstanding, equation (13B1.2-1) reproduces experimental air/hydrocarbon diffusivity
data with an average deviation of not more than 9 percent. The nature of the atomic and structural diffusion volumes
is such that diffusion coefficients in a wider range of nonhydrocarbon systems can also be estimated.

Special Comments
If the desired pressure is greater than 500 psia, use a convenient low pressure in equation (13B1.2-1) and apply
the pressure correction (Figure 13B1.4) to the resulting diffusion coefficients. For diffusion in a multicomponent
vapor, combine the binary coefficients using Procedure 13B2.1. For hydrocarbon/hydrocarbon systems use
Procedure 13B1.1.

Literature Source
Equation (13B1.2-1) and Table 13B1.3 were developed by Fuller, Schettler, and Giddings, Ind. Eng. Chem. 58
19 (1966).

Example
Estimate the diffusion coefficient of benzene vapor diffusing into air at 86 °F and 14.7 psia.
From the Pure Component Properties Program, the molecular weights are benzene (M1) = 78.11 and air (M2) =
28.86. From Table 13B1.3, the diffusion volumes are:

 1
vi (benzene)  6 16.5  6 1.98   1 20.2   90.68

 2
vi (air )  20.1

The diffusion coefficient of benzene vapor in air at 86 °F and 14.7 psia is calculated using equation (13B1.2):

0.5
 1 1 
86  459.7 
1.75
  
 78.11 28.86 
D1,2  0.0204 2
14.7  90.68    20.1 
1/3 1/3
 

= 0.358 ft2/hr

An experimental value of 0.357 ft2/hr is available. This estimated value represents a deviation of 0.28 percent.

13 - 18 © 2013, The American Petroleum Institute and EPCON International - All Rights Reserved
 API TECHNICAL DATA BOOK – 9th Edition
Table 13B1.3 – Atomic Diffusion Volumes for Use in Procedure 13B1.2
(1980)

Atomic and Structural Diffusion-Volume

Increments v

C………. 16.5 Cl………………… (19.5)

H………… 1.98 S………………….. (17.0)
O………… 5.48 Aromatic ring…….. –20.2
N*………. (5.69) Heterocyclic ring…. –20.2

Diffusion Volumes for Simple Molecules

H2………… 7.07 CO……………... 18.9

D2………… 6.70 CO2…………….. 26.9
He………. 2.88 N2O…………….. 35.9
N2………. 17.9 NH3…………….. 14.9
O2………. 16.6 H2O…………….. 12.7
Air……… 20.1 CC12F2……..…… (114.8)
Ar……….. 16.1 SF6……………... (69.7)
Kr……….. 22.8 Cl2……………... (37.7)
Xe……… (37.9) Br2…………….. (67.2)
SO2…………..... (41.1)

* Parentheses indicate that the value listed is based on only a few data points.

© 2013, The American Petroleum Institute and EPCON International - All Rights Reserved 13 - 19
 API TECHNICAL DATA BOOK – 9th Edition
Figure 13B1.4 – Diffusion Coefficients in Gas Binary Systems at High Pressure
(1980)
( pD )0
pD

FIGURE 13B1.4
DIFFUSION COEFFICIENTS IN
GAS BINARY SYSTEMS
AT HIGH PRESSURES
Created: June 1980 Remastered : 2012

TECHNICAL DATA BOOK VERSION 9

Copyright © 2013. The American Petroleum Institute and EPCON International, Inc.

REDUCED PRESSURE , P/ppc

13 - 20 © 2013, The American Petroleum Institute and EPCON International - All Rights Reserved
 API TECHNICAL DATA BOOK – 9th Edition
Comments on Figure 13B1.4
(1980)

Purpose
Figure 13B1.4 is presented for estimating the effect of pressure on binary gas diffusion coefficients when the
desired pressure is greater than 500 psia. It is applicable to both hydrocarbon/hydrocarbon systems and
air/hydrocarbon systems.

Limitations
This figure is based on very few experimental data.

Reliability
Published experimental data are not available to evaluate the figure. The estimated reliability is 50 percent.

Notation
pD1‚2 = product of the pressure and binary diffusion coefficient at a pressure greater than 500
psia
(pD1‚2) = same product at a convenient pressure below 500 psia in consistent units
p = pressure of the system in psia
ppc = pseudocritical pressure of the mixture in psia
T = temperature of the mixture in degrees Rankine
Tpc = pseudocritical temperature of the mixture in degrees Rankine

Special Comments
In the absence of experimental data, use Procedure 13B1.1 or Procedure 13B1.2 (depending on the type of
system) to determine (pD1‚2)°. For a multicomponent mixture at high pressure, the binary diffusion coefficients
from this figure should be used in conjunction with Procedure 13B2.1.

Literature Sources
This figure was developed by Slattery and Bird, AIChE Journal 4 137 (1958) and reproduced in Bird et al.,
Transport Phenomena, John Wiley and Sons, Inc., New York (1960).

Example
Estimate the diffusion coefficient of n-hexane vapor diffusing into methane at 50 °F and 600 psia.
In the absence of experimental data, calculate the diffusion coefficient for n-hexane diffusing into methane at
50 °F and a convenient pressure less than 500 psia using Procedure 13B1.1. From the example of Procedure
13B1.1, D°1‚2 = 0.301 ft2/hr at a pressure 14.7 psia. The pseudocritical properties of the predominantly methane
mixture are taken as 343 R and 667 psia. The reduced pressure is therefore 0.8995 and the reduced temperature is
1.486. From Figure 13B1.4,


 ( pD1, 2 )  
 0 
 0.86

 ( pD1, 2 ) 


The desired coefficient is:

 0.86   pD1,2 
0

D1,2 
p


0.8614.70.301
600

= 0.0063 ft2/hr.

© 2013, The American Petroleum Institute and EPCON International - All Rights Reserved 13 - 21
 API TECHNICAL DATA BOOK – 9th Edition

13B2 Multicomponent Gas Systems

Procedure 13B2.1 – Diffusion Coefficients in Multicomponent Gas Systems

(1980)

Discussion
The following equation is recommended for the estimation of diffusion coefficients in multicomponent gaseous
systems from the binary coefficients of the component pairs in the mixture:

1  y1
D1, 2, 3,  n  (13B2.1-1)
y2 y3 y
   n
D1,2 D1,3 D1,n

Where:

D1,2‚3…n = diffusion coefficient of component 1 in the n-component mixture, in ft2/hr

D1‚i = binary diffusion coefficient of component 1 in component i, in ft2/hr (i ≠ 1)

yi = mole fraction of component i in the mixture.

Procedure
Step 1: Obtain all the necessary gas binary diffusion coefficients. In the absence of experimental data use
the procedures of section 13B1.
Step 2: Calculate the multicomponent diffusivity using equation (13B2.1-1).

13 - 22 © 2013, The American Petroleum Institute and EPCON International - All Rights Reserved
 API TECHNICAL DATA BOOK – 9th Edition

Comments on Procedure 13B2.1

(1980)

Purpose
This procedure is useful for estimating diffusion coefficients in multicomponent vapor systems from the binary
coefficients of the component pairs in the mixture.

Limitations
Equation (13B2.1-1) is based on very few experimental data for vapor systems most of which are
nonhydrocarbons (usually hydrogen).

Reliability
Few published experimental data are available to evaluate equation (13B2.1-1). It is estimated that
approximately 5 percent uncertainty is added to errors in the binary coefficients when using this equation.

Literature Source
Equation (13B2.1-1) was developed by Wilke, Chem. Eng. Progr. 46 95 (1950).

Example
Estimate the diffusion coefficient of n-hexane at 50 °F and 14.7 psia in a gaseous mixture containing 0.01 mole
fraction n-hexane, 0.03 mole fraction propane and 0.96 mole fraction methane.
The diffusion coefficients of n-hexane in propane and n-hexane in methane are first calculated using Procedure
13B1.1. From the example for Procedure 13B1.1, Dc6,c1 = 0.3010 ft2/hr. Similarly, Dc6,c3 = 0.1662 ft2/hr.
The multicomponent diffusion coefficient is determined using equation (13B2.1-1):

DC6 ,C3 ,C1 

1  0.01
 0.03   0.96 
  
 0.1662   0.3010 
 0.294 ft 2 /hr

© 2013, The American Petroleum Institute and EPCON International - All Rights Reserved 13 - 23
 API TECHNICAL DATA BOOK – 9th Edition

13C Diffusivity in Gas-Liquid Systems

13C1 Gas – Liquid Systems

Procedure 13C1.1 – Diffusion Coefficients of Dissolved Gases in Liquids

(1980)

Discussion
The following equation is recommended for estimating diffusion coefficients of dissolved gases diffusing into
liquids. It is the same equation used in Procedure 13A1.1. It was developed for use in both gas/liquid and dilute
binary liquid systems.

T R2 (13C.1-1)
D1,2  5.922 108 2/3
 R1

Where:

D1,2 = diffusion coefficient of solute (component 1) in solvent (component 2) in ft2/hr

T = temperature in degrees Rankine
μ = viscosity of solution (considered as pure solvent) in centipoise
R2 = radius of gyration of solvent in Ångstrom units
R1 = radius of gyration of solute in Ångstrom units

Procedure
Step 1: Obtain R 1 and R 2 from Table 13A1.2. In the absence of experimental data, estimate μ using the
procedures of Chapter 11.
Step 2: Calculate D1‚2 from equation (13C1.1-1).

13 - 24 © 2013, The American Petroleum Institute and EPCON International - All Rights Reserved
 API TECHNICAL DATA BOOK – 9th Edition

Comments on Procedure 13C1.1

(1980)

Purpose
Procedure 13C1.1 is presented primarily for the estimation of diffusion coefficients of gases diffusing into
nonpolar liquids.

Limitations and Reliability

Equation (13C1.1-1) was developed from hydrocarbon/hydrocarbon and hydrocarbon/carbon tetrachloride data.
It is based on relatively few experimental data points. For the systems studied, an average deviation of 16 percent
was obtained with occasional errors of 25 percent. It is not intended to be used for the diffusion of acid gases into
liquids. The coefficients predicted for these systems are too small, generally by 40 to 50 percent.

Source of Equation
Equation (13C1.1-1) was developed by Umesi (M.S. Thesis, The Pennsylvania State University, 1980).

Example
Estimate the diffusion coefficient of methane diffusing into n-hexane at 86 °F.

From Chapter 11, the viscosity of the solution (considered as the viscosity of pure n-hexane) at 86 °F is 0.2820
centipoise.

From Table 13A1.2, the radius of gyration of n-hexane, R 2 , is 3.812Å and that of methane, R 1 , is 1.1234 Å.
The diffusion coefficient is calculated using equation (13C1.1):

5.922 108  86  459.7  3.812 

D1,2 
 0.282 1.1234 
2/3

 4.04 104 ft 2 /hr

An experimental value of 3.95×10–4 ft2/hr is available. This estimated value represents a deviation of 2.28
percent from the experimental value.

© 2013, The American Petroleum Institute and EPCON International - All Rights Reserved 13 - 25
 API TECHNICAL DATA BOOK – 9th Edition

Bibliography
Chapter 13 – Diffusivity
(1981)

1. Bird, R.B., Stewart, W.E., and Lightfoot, E.N., The Properties of Gases and Liquids, 3rd ed.,
Transport Phenomena, John Wiley and Sons, McGraw-Hill, New York (1977).
Inc., New York (1960). 8. Sanchez, V., and Clifton, M., “An Empirical
2. Caldwell, C.S., and Babb, A.L., “Diffusion in Relationship for Predicting the Variation with
the System Methanol-Benzene,” J. Phys. Chem. Concentration of Diffusion Coefficients in
60 51 (1956). Binary Liquid Mixtures,” Ind. Eng. Chem.
3. Fuller, E.N., Schettler, P.D., and Giddings, J.C., Fundam. 16 318 (1977).
“A New Method for Prediction of Binary Gas- 9. Slattery, J.C., and Bird, R.B., “Calculation of the
Phase Diffusion Coefficients,” Ind. Eng. Chem. Diffusion Coefficient of Dilute Gases and of the
58 [5] 19 (1966). Self-Diffusion Coefficient of Dense Gases,”
4. Gilliland, E.R., “Diffusion Coefficients in AIChE Journal 4 137 (1958).
Gaseous Systems,” Ind. Eng. Chem. 26 681 10. Tyn, M.T., and Calus, W.F., “Estimating Liquid
(1934). Molar Volumes,” Processing 21 [4] 16 (1975).
5. Kuong, J.F., “Nomograph Gives Diffusion Rate 11. Umesi, N.O., M. S. Thesis, The Pennsylvania
in Dilute Solutions,” Chem. Eng. 68 [12] 258 State University, May 1980.
(1961). 12. Wilke, C.R., “Diffusional Properties of
6. Leffler, J., and Cullinan, H.T., “Variations of Multicomponent Gases,” Chem. Eng. Progr. 46
Liquid Diffusion Coefficients with Composition: 95 (1950).
Ternary Systems,” Ind. Eng. Chem. Fundam. 9 13. Wilke, C.R., and Chang, P.,“Correlation of
88 (1970). Diffusion Coefficients in Dilute
7. Reid, R.C., Prausnitz, J.M., and Sherwood, T.K., Solutions,”AIChE Journal 1 264 (1955).

13 - 26 © 2013, The American Petroleum Institute and EPCON International - All Rights Reserved