Materials Transactions, Vol. 48, No. 4 (2007) pp.

787 to 792
#2007 The Japan Institute of Metals

Computation of Equilibrium-State in Gas/Solid Materials Systems

Yalamanchili Krishna Rao
Department of Materials Science and Engineering, University of Washington,
Roberts Hall, Box 352120, Seattle, WA 98195 USA
The determination of equilibrium-state in gas-solid materials systems is of practical interest; this consists of nding the values of intensive
variables such as mole-fractions, partial pressures, activities or chemical potentials of constituents that occur in the gaseous and solid phases for
stipulated temperature, pressure and initial reactant-gas composition. The mass-action iterative equilibrium constant method, shown to be a
direct derivative of the minimization of Gibbs free energy, was used to compute the equilibrium-state in the carburizing and decarburizing ironcarbon heat treatment systems. By suitably combining the equilibrium partial pressure data with the extent of reaction formalism, the optimum
feed-gas mixtures that ensured neutral atmospheres were determined. [doi:10.2320/matertrans.48.787]
(Received July 24, 2006; Accepted January 25, 2007; Published March 25, 2007)
Keywords: equilibrium, heat treatment, carburizing, decarburizing, iron-carbon alloys, activity, mole-fraction, pressure, chemical potential,
extent of reaction, solid solution

1.

Introduction

The determination of equilibrium-state in high temperature

systems is undertaken for practical considerations. At
elevated temperatures, chemical processes (reactions) and
transport phenomena occur at rapid rates and thus the
equilibrium model tends to give a good representation of the
actual state of the system. In general, for a system in stable
equilibrium, the entropy is at maximum whereas the Gibbs
free energy is at a minimum. That is, for chemical reaction
equilibrium at constant temperature and pressure, the rstderivative of Gibbs free energy with respect to extent of
reaction (") is nil.
@G=@"T,P 0

where the variable " is indicative of the progress of the

reaction.1) For the purpose of illustration, let us consider
the dissociation of one mol of copper iodide trimer into
monomeric species at T 1600 K and P 0:5 atm (50.663
kPa) total pressure.
Cu3 I3 (g) 3CuI(g);
(1
") mol 3" mol;

": extent of reaction

(1 2"): molar amount
of gas-phase:

The total Gibbs free energy Gt of the monomer-trimer system

can be readily expressed in terms of chemical potentials and
the respective molar amounts as follows:
Gt 3"
1 1
"
3
3"
o 1 RT ln P1 1
"
o 3 RT ln P3

2)

Using the data of Knacke and co-workers for standard

chemical potentials, in Jmol
1 ,

1
318551 RT ln P1
with P1 3"P=1 2" for CuI(g);

3
942411 RT ln P3
with P3 1
"P=1 2" for Cu3 I3 (g):
At " 0:05 mol, P1 6908 Pa (0.06818 atm) and P3
43754 Pa (0.43182 atm) with the result the chemical potentials become
1
354277 Jmol
1 and
3
953582
Jmol
1 ; and eq. (3) yields Gt
959045 J. It is advanta-

Fig. 1 Gibbs free energy (Gsys ) of the trimer-monomer system shown as a

function of the extent (") of reaction Cu3 I3 = 3CuI. The free energy
minimum Gsys
34:73 kJ occurs at "e 0:8078 mol, equilibrium-state
for the system, at 1600 K and 50.663 kPa (0.5 atm).

geous to select as reference state 1.0 mol of trimer at 1600 K

and 50.663 kPa (0.5 atm), with reference chemical potential

3 (r)
o 3 RT ln P
951632 Jmol
1 . Thus, Gsys , the
relative free energy of the system becomes:
Gsys Gt

3 (r) 3"
1 1
"
3 951632 4
At " 0:05 mol, eq. (4) together with the above values of

1 and
3 yields Gsys
7413 J. The results of similar
calculations for Gsys at several values of the variable " are
presented in Fig. 1. The free energy plot is seen to exhibit a
minimum at " "e , the equilibrium extent of reaction. From
eqs. (1), (3) and (4), it is clear that for the system at
equilibrium,
@Gsys =@"T,P 3
o 1 3RT ln P1e

o 3
RT ln P3e 0
and

Go 3
o 1

o 3
13242 J

RT lnP1e 3 =P3e 
RT ln Ke

788

Y. K. Rao

On expressing the equilibrium partial pressures of the

monomer (P1e ) and the trimer (P3e ) in terms of the equilibrium extent of reaction ("e ), eq. (5) provides,
27P2 4Ke "e 3
3Ke "e
Ke 0;
total pressure P 0:5 atm (50.663 kPa)
Using Ke exp13242=1600R 2:706 for the equilibrium
constant at 1600 K, the above equation was solved by trial
and the equilibrium extent of reaction was determined:
"e 0:8078 mol. This value corresponds to equilibrium
partial pressures of P1e 46:944 kPa for CuI(g) and P3e
3719 Pa for Cu3 I3 (g), respectively. The saturation vapor
pressure of CuI(liq) is 55.151 kPa at 1600 K; and the system
remains all-gas at the 50.663 kPa (0.5 atm) total pressure.
The free energy minimum in Fig. 1 is found: since
1e

328786 J and
3e
986377 J, it can be readily shown
that Gsys
34:73 kJ. A small shift from the equilibrium-state, say, to " "e
, where  0:0122 mol gives
Gsys(
)
34:723 kJ; in a similar manner, " "e  provides Gsys(+)
34:722 kJ. The net positive change in Gsys
due to small displacement  in either direction can be
construed as conrmation of the minimum in the Gsys -" curve
for the system initially composed of 1 mol Cu3 I3 (g) trimer.
In the computation of equilibrium-state by minimizing
Gibbs free energy (Gt ), subject to the (elemental) atombalance constraints, the linear programming techniques
pioneered by White, Johnson and Dantzig3) continue to
occupy a favored position. An excellent review of the
literature on equilibrium computation is oered by Smith and
Missen.4) It is well to note that the free energy minimization
procedure does not make use of stoichiometric equations that
represent chemical reactions occurring in the system. In
contrast, the latter form the main basis for law of mass-action
calculations: Damkohler and Edse5) employed the non-linear
equilibrium constant expressions formulated in terms of
species partial pressures and a converging successive
approximations scheme to compute the equilibrium compositions in the rocket-fuel-combustion systems; the CruiseVillars method6,7) likewise makes use of reaction equilibria,
and equilibrium constants expressed in terms of molefractions of species present in the system, to determine the
equilibrium-state. The accelerated or the brute-force convergence technique reported by Bahn,8) and later developed
by Kellogg,9) bears a close resemblance to the earlier work of
Damkohler and Edse.5) The selection of prime state variables
(partial pressures of major species) and the direct use of
equilibrium constants to calculate the remaining quantities
are the main characteristics of this unusually-resilient computational method, especially suited for application to the
determination of steady-states in closed-systems.10,11)
The free energy minimization algorithms such as the
Rand,3) the NASA12) and the Solgas13) retain a large
following in the scientic community. At this juncture, it is
well to note that there is no real dichotomy between the two
approaches seeking to solve the problem of equilibrium; it
remains to be shown that by minimizing the total free energy
of the system in accordance with appropriate analogs of
eq. (1), one can obtain the entire framework of the iterative
equilibrium constant method,14) the ingenious computational
scheme that originates with Damkohler and Edse.5)

2.

Fe-C-H-O Gas-Solid System

Iron-carbon alloys are heat treated at high temperatures,

usually in the range of 9001050 C, under controlled
atmospheres that consist of carbon monoxide, hydrogen,
methane and water vapor. Such atmospheres can produce
carbon-transfer between the gas-phase and the solid alloy.
Carburizing occurs under CO-rich atmosphere with a net
addition of carbon to the hot steel surface. When the gasphase is carbon-decient and contains reactants carbon
dioxide and water vapor in sucient concentrations, rapid
decarburization of the Fe-C(s) alloy takes place. During heat
treatment under a neutral C-H-O atmosphere, there is neither
the addition nor the removal of carbon from the surface of the
hot steel; the composition of such an atmosphere would be
regarded as equilibrium-state for the two-phase Fe-C-H-O
system at the stipulated temperature and total pressure. Under
heat-treating conditions that do not produce oxidation of the
iron-constituent, the following species are normally present
in the high-temperature system:
CO, CO2 , CH4 , H2 , H2 O: gas-phase; C, Fe: solid-alloy
The atom-matrix constructed of these seven species has a
rank of four (that is, four kinds of atoms Fe, C, H, and O
underlie the species); according to Gibbs stoichiometric rule,
at the most three independent reaction equilibria are possible
in this system: rm 7
4 3. The total Gibbs free energy
of the system can be expressed in terms of molar amounts ni
and chemical potentials
i of the species.
Gt nCO
CO nCO2
CO2 nCH4
CH4
nH2
H2 nH2 O
H2 O nS
C nI
Fe

where nS and nI denote the respective molar amounts of

carbon and iron in the solid alloy.
The elemental mass balances for the solid/gas system can be
formulated as follows:
no C nCO nCO2 nCH4 nS constant
7
no H 2nH2 nH2 O 4nCH4 constant
8
o
n O nCO 2nCO2 nH2 O constant
9
no Fe nI constant
10
Dierentiation of these mass-balance equations with respect
to nCO yields
@nCO2 =@nCO 1 @nCH4 =@nCO @nS =@nCO 0
2@nH2 =@nCO 2@nH2 O =@nCO 4@nCH4 =@nCO 0
1 2@nCO2 =@nCO @nH2 O =@nCO 0
Since the iron constituent is not partaking in any reactions,
eq. (10) is disregarded. Upon rearrangement of terms in the
foregoing relations one obtains,
@nH2 O =@nCO
1
2@nCO2 =@nCO

11

@nCH4 =@nCO 0:5
0:5@nH2 =@nCO

@nCO2 =@nCO

12

@nS =@nCO
1:5
2@nCO2 =@nCO

0:5@nH2 =@nCO

13

The equilibrium-state for the heterogeneous system at

specic temperature and pressure corresponds to a minimum
in the Gibbs free energy. Thus, the rst-dierential of Gt with

Computation of Equilibrium-State in Gas/Solid Materials Systems

respect to molar amount nCO of carbon monoxide must equal

zero. From eq. (6),
0
CO @nCO2 =@nCO
CO2 @nCH4 =@nCO
CH4
@nH2 =@nCO
H2 @nH2 O =@nCO
H2 O
@nS =@nCO
C 
Substituting for the respective partial derivatives from
eqs. (11), (12) and (13), it is found
f
CO

H2 O 0:5
CH4
1:5
C 
@nCO2 =@nCO
CO2
2
H2 O
CH4
2
C 
@nH2 =@nCO
H2
0:5
CH4 0:5
C g 0

14

It is readily seen that each group of terms enclosed in the

square brackets must vanish in order to satisfy eq. (14).
Accordingly, the relations in terms of
i s and the corresponding reaction equilibria are as follows:

CO

H2 O 0:5
CH4
1:5
C 0

15

1.5C(s) H2 O(g) 0.5CH4 (g) CO(g); "1 mol R1

CO2
2
H2 O
CH4
2
C 0
16
2C(s) 2H2 O(g) CH4 (g) CO2 (g); "2 mol

H2
0:5
CH4 0:5
C 0
0.5CH4 (g) 0.5C(s) H2 (g); "3 mol

R2
17
R3

It will be noted that the extensive quantities "1 , "2 and "3
represent the extents of reaction for the equilibria (R1), (R2)
and (R3), respectively. Thus, the minimization of free energy
subject elemental mass balances shows that there exist three
independent reactions or chemical potential constraints in
this system.
3.

Equilibrium-state in Steel Heat-Treating Systems

The iron-carbon phase diagram shows that austenite is an

important phase, especially in steels at high temperatures; it
is a homogeneous solid solution of carbon in face-centeredcubic -iron with a solubility limit of 2.11 w=o carbon
at 1153 C. The activity (aC ) of carbon in austenite was
investigated as a function of temperature and mole-fraction
(XC ). The following relationship was developed by Elliott
and Chipman15) using the extensive experimental data of
Ban-ya et al.16) on carbon activities in austenite:
Log aC LogXC =1
2XC  2300=T

0:92 3860=TXC =1
XC 

18

where the logarithms are for base ten and the temperature (T)
is expressed in Kelvins.
The unknown intensive quantities are the chemical
potentials
CO ,
CO2 ,
CH4 ,
H2 ,
H2 O ,
C and
Fe seven
altogether; furthermore, for the gas-phase,
i
o i
RT ln Pi where Pi is the i-th species partial pressure; in the
Fe-C(s),
C
o C RT ln aC and the non-independent

Fe
o Fe RT ln aFe can be readily found by performing
Gibbs-Duhem integration of eq. (18). The application of
Gibbs phase rule shows the system has four degrees of
freedom; hence four state properties must be assigned
specic values in order to fully dene its thermodynamic
state: (i) temperature (T) (ii) total pressure (P) (iii) inlet
gas-composition, that is, (H=O) atom-ratio of feed-gas (iv)
mole-fraction (XC ) of carbon in steel. For the purpose

789

of illustration, the heat treatment of Fe-C(s) solid solution

with XC 0:02 (or 0.44 w=o C) at T 1200 K and P
101:325 kPa (1 atm) is considered; for the solid-phase,
eq. (18) gives aC 0:24058. The inlet gas contains 1 mol
each of CH4 , CO and H2 O and 0.5 mol H2 : thus no H
7 mols, no C 2 mols and no O 2 mols, giving a (H=O)
atom-ratio of 3.5; a feed-gas having the same atom-ratio can
also be constituted by mixing 3.5 mols H2 and 2 mols CO.
Whether such a gas-mixture transfers carbon to the Fe-C
solid-solution at the specied temperature and pressure
remains to be established. Furthermore, the composition of
the feed-gas that ensures a neutral atmosphereone that
entails zero carbon-transfer between phasesis of practical
interest; it should be determined as part of the equilibriumstate calculation. While changes may occur in nC , the molar
amount of carbon, and nC =nO ( C=O) atom-ratio of the gasphase, as it reaches a state of equilibrium with the Fe-C(s),
the atom-ratio H=O remains the same. Similar behaviour is
indicated when the feed-gas is relatively lean with respect to
carbon; in this instance, however, there will be a net transfer
of carbon from the solid-alloy to the gas-phase resulting in a
decarburized surface-layer on the steel specimen.
4.

Method of Calculation

The details of the iterative solution that makes use of

equilibrium constants for the reactions (R1), (R2) and (R3)
are presented. Data on standard chemical potentials (
o i ) for
the species CO, CO2 , CH4 , H2 , H2 O and C(s) are drawn from
Knacke, Kubaschewski and Hesselmann.2) At T 1200 K,
in kJmol
1 ,

o CO
371:442;
o CO2
685:576;

o CH4
335:186;
o H2
179:347;

o H2 O
496:461;
o C
18:019:
Substituting
CO
o CO RT ln PCO , . . . . . ., and
C

o C RT ln aC in eq. (15), and rearranging terms, it is found
for reaction (R1),

Go R1
o CO 0:5
o CH4
1:5
o C

o H2 O

RT lnfPCO PCH4 0:5 =PH2 O aC 1:5 g
Thus, the equilibrium constant K1 of reaction (R1) at 1200 K
is given by
K1 PCO PCH4 0:5 =PH2 O aC 1:5 
expf

Go R1 =1200Rg 4:75

19

Similarly, one nds for reaction equilibria (R2) and (R3),

from eqs. (16) and (17),

Go R2
o CH4
o CO2
2
o C
2
o H2 O
RT ln K2
K2 PCH4 PCO2 =aC PH2 O 2  0:44

20

Furthermore,

Go R3 0:5
o C
o H2
0:5
o CH4
RT ln K3
K3 aC 0:5 PH2 =PCH4 0:5  8:0133

21

The molar amounts of the species can be expressed in terms

of extents of reactions:1)
nCO no CO "1 ; nCH4 no CH4 0:5"1 "2
0:5"3 ;
nCO2 no CO2 "2

22a-c

790

Y. K. Rao
Table 1 C-O-H-Fe Equilibrium system at 1200 K and 101.325 kPa (1 atm).
Quantity
o

Units

High C

Low C

Equilibrium

nCH4

mol

1.0000

1.2368

1.1126

nCO o

mol

1.0000

1.0000

1.0000

nH2 O o

mol

1.0000

1.6316

1.3002

nH2 o
nH o

mol
mol

0.5000
7.0000

0.5000
9.2104

0.5000
8.0508

RH H=Oo

3.5000

3.5000

3.5000

RC C=Oo

1.0000

0.8500

0.91842

aC

0.24058

0.24058

PCH4 eq

Pa

PCO eq

kPa

PCO2 eq

Pa

919.6

919.6

919.6

PH2 O eq
PH2 eq

Pa
kPa

2338.6
64.0

2338.6
64.0

2338.6
64.0

Htot eq

kPa

RHE H=Oeq

3.5000

3.5000

RCE C=Oeq

0.91842

0.91842

0.91842

"1

mol

0.78063

1.34300

1.04792

"2

mol

0.04830

0.06355

0.05555

"3

mol

2.86127

3.92266

C

mol

0.16310

0:18027

151.5
33.904

nC n

1:5"1
2"2 0:5"3

33.904

133.285

133.285
3.5000

3.36585
0

PCH4 new PCH4 old P=PE0:04 RH=RHE0:06

PCO new PCO old P=PE

22d- f

The net carbon-transfer,
C , from the gas-phase to Fe-C(s) is

readily determined using,

C nC
no C 0:5"3
1:5"1
2"2 ; mol

151.5

33.904

133.285

nH2 no H2 "3 ; nH2 O no H2 O
"1
2"2 ;

o

0.24058

151.5

23

The iterative method described by Rao17) is similar in

many respects to the earlier works5,8,9) that employ successive approximations scheme to secure a converged solution.
Of the ve gaseous species listed, the two species CH4 and
CO are selected as majors and reasonable estimates are
made for their partial pressures: PCH4 0:2P and PCO
0:4P. The partial pressures of species H2 O, CO2 and H2 are
computed using the eqs. (19), (20), and (21) for the equilibrium constants K1 , K2 and K3 . At 101.325 kPa and aC
0:24058,
PH2 O PCO PCH4 0:5 =K1 aC 1:5  32:333 kPa
PCO2 K2 aC PH2 O 2 =PCH4 1314 Pa;
PH2 K3 PCH4 =aC 0:5 740:31 kPa
The new values for PE, the total pressure, and RHE, the H=O
atom-ratio, are computed:
PE PCH4 PCO PH2 O PCO2 PH2 834:756 kPa
RHE Htot =Otot eq
4PCH4 2PH2 O 2PH2 =PCO 2PCO2 PH2 O 
21:542
Both of these exhibit large departures from the constraining
values: P 101:325 kPa and RH 3:5; thus, the rst
iteration values for the ve Pi s obtained above do not
represent equilibrium data. The following convergence
equations are introduced to modify the two estimates
admitted for the majors CH4 and CO at rst iteration:

0:05

24
25

With these modied values, the next iteration is completed

and new values computed for the Pi s, PE and RHE. The
computation is continued until the absolute departure of PE
from P and that of RHE from RH, match the limits dened
below:
j1:0
P=PEj 0:0001;
and

j1:0
RH=RHEj 0:0001:

For the particular equilibrium between Fe-C(s) alloy containing 0.44 w=o C and a feed-gas with the H=O atom-ratio
(RH) of 3.5, converged solution required 341 iterations.
The data on equilibrium partial pressures are summarized
in Table 1; the high C feed-gas consists of 1.0 mol each of
CH4 , CO and H2 O and 0.5 mol H2 , with an atom-ratio RH
H=Oo equal to 3.5. The iron-carbon solid solution
contained 0.44 w=o C which corresponds to a carbon-molefraction of 0.02 and an activity of aC 0:24058. It will be
noted that the equilibrium gas-phase is relatively lean in CH4
and CO2 ; in lieu of CH4 , if one were to choose H2 as major
species, converged solution is obtained in fewer iterations
(184 to be exact). The carbon-to-oxygen atom-ratio decreases
from RC 1:0 in the feed-gas to RCE 0:91842 of the
equilibrium gas-phase; thus, there is a net transfer of carbon
to the Fe-C(s) alloy. The amount of carbon transferred can be
determined by eq. (23) using the extents of reactions
evaluated from the data on equilibrium pressures.
nH2 PH2 =Htot eq nH o nH2 o "3 ;
and
and

"3 2:86127 mol : High C (Table 1)

nCO2 PCO2 =Htot eq nH o "2 ;
"2 0:0483 mol : High C (column 3, Table 1)

Computation of Equilibrium-State in Gas/Solid Materials Systems

791

Table 2 Eect of feed-gas composition on equilibrium in the C-O-H-Fe system at 1200 K and 101.325 kPa (1 atm) for XC 0:02 (0.44
w=o C).
Units

RC:

Units

1.0

1.0

1.0

1.0

1.0

1.25

2.5

1.0

nCH4 o
nCO o

mol
mol

0.11111
1.0

0.42857
1.0

1.0
1.0

2.33334
1.0

0.9
0.1

1.15
0.10

2.125
1.00

1.0
1.0

nH2 O o

mol

0.11111

0.42857

1.0

2.33334

0.9

0.90

0.25

1.0

nH2 o

mol

0.5

0.5

0.5

0.5

0.05

0.05

0.50

PCH4 eq

Pa

68.02

115.05

151.46

179.62

201.83

219.68

240.63

PCO eq

kPa

PCO2 eq

Pa

2297.5

1380.7

PH2 O eq

Pa

2477.1

2497.4

PH2 eq
RCE

kPa

42.889
0.92241

55.778
0.91960

64.001
0.91842

69.696
0.91815

73.881
0.91843

77.077
0.91907

80.670
0.92049

63.711
0.98582

C

mol

0.08637

0.11498

0.16310

0.27237

0.08141

0.33082

1.97418

0.02850

1.5

53.591

2.5

3.5

41.543

4.5

33.904

28.633

919.62

655.85

2338.6

2150.7

24.778
491.13
1972.9

6.5

21.835
381.41
1813.8

8.0

18.532
274.74
1611.2

0.5
508.15
36.058
307.21
731.28

This run is for an iron-carbon alloy with XC 0:05 (1.12 w=o C).

nCO PCO =Htot eq nH o nCO o "1 ;

and

"1 0:78063 mol : High C (Table 1)

Substitutions into eq. (23) yield for net carbon transfer (
C )
a value of 0.1631 mol; it is clear that prolonged heat
treatment under High C atmosphere promotes carburizing
of steel. In contrast, the Low C feed-gas (column 4,
Table 1) induces a net transfer of C from the solid-alloy
phase to the gaseous mixture; in this case,
C

0:18027 mol. The equilibrium (C=O) atom-ratio, RCE
0:91842, is of signicance in that it allows an exact
prediction of the desired feed-gas composition to maintain
a neutral atmosphere. In a four species feed-gas, with 1 mol
CO and 0.5 mol H2 , the molar amounts of CH4 and H2 O are:
no H2 O no CO 0:25RH
RCE 1:0
0:5no H2 
=RCE
0:25RH 0:5
26
no CH4 RCEno H2 O no CO RCE
1:0

27

The equilibrium feed-gas thus contains 1.1126 mols CH4 ,

1.3002 mols H2 O, 1 mol CO and 0.5 mol H2 ; under this
atmosphere (RH 3:5; RC 0:91842) there occurs zero
carbon-transfer, that is
C 0. Alternatively, a feed-gas
containing 28.43 mol% CH4 , 33.23 mol% H2 O, 25.56 mol%
CO and 12.78 mol% H2 ensures that the carbon-content of the
Fe-C(s) alloy does not change during heat treatment at
1200 K.
5.

5.5

3.5

RH:

Discussion of Results

The eect of varying the H=O atom-ratio of the feed-gas

from RH 1:5 to RH 8 was investigated for an Fe-C(s)
alloy with XC 0:02 (or 0.44 w=o C) at 101.325 kPa total
pressure and 1200 K; the inlet gas mixtures were usually rich
in carbon (or High C). The equilibrium gas composition
Pi s of species CH4 , CO, CO2 , H2 O, H2 and the corresponding carbon-to-oxygen atom-ratio (RCE) were computed. The results are listed in Table 2; it will be noted that
RCE PCH4 eq PCO eq PCO2 eq 
=PCO eq 2PCO2 eq PH2 O eq 

28

A careful study of the data reveals that the inlet-gas C=O

atom-ratio (RC) is consistently larger than the corresponding

equilibrium value (RCE); thus carbon-addition to the solidphase is indicated for each of the seven feed-gas compositions selected. The inuence of H=O (or RH) atom-ratio on
the equilibrium C=O (or RCE) atom-ratio appears to be
small; for the most part, RCE  0:92 (Table 2). However, as
the last column shows, for an iron-carbon alloy with XC
0:05 (or 1.12 w=o C) at 1200 K and 101.325 kPa total
pressure, the equilibrium carbon-to-oxygen ratio (RCE) of
the gas-phase is decidedly larger. One can use the data given
in Table 2 in combination with eqs. (26) and (27) to
constitute a feed-gas-mixture that serves as a neutral atmosphere for the heat treatment of steels.
As the hydrogen-to-oxygen atom-ratio (RH) of the inlet
reactant gas-mixture is increased (Table 2), it is seen that the
H2 -concentration of the equilibrium gas-phase rises from
about 42 mol% at RH 1:5 to about 80 mol% at RH 8
while that of the carbon monoxide declines from about 53
mol% to 18 mol%. Similar trends were noted in the composition of the gas-phase in equilibrium with an iron-carbon
alloy containing 1.12 w=o C (or XC 0:05). The eect of
temperature on the equilibrium-state in the two-phase (C-OH-Fe) system was investigated: at T 1300 K and 101.325
kPa total pressure, the equilibrium partial pressures were
calculated for the iron-carbon alloys with XC 0:05. The
data on the standard chemical potentials (
o i ) drawn from
Knacke and co-authors2) were used to compute the equilibrium constants for the reaction equilibria (R1), (R2)
and (R3):
o CO
395:728,
o CO2
713:919,
o CH4

361:63,
o H2
196:664,
o C
20:968 and
o H2 O

520:792 kJmol
1 ; furthermore, K1 9:47, K2 0:47832
and K3 11:4307, respectively. The equilibrium calculations were made with the feed-gas H=O atom-ratio (RH)
varied over a wide range; and the results are presented in
Table 3.
The chosen inlet-gas mixtures are expected to add carbon
to the solid-alloy during the heat treatment; the net carbontransfer (
C ) was computed from the equilibrium gas-phase
partial pressures for each of the seven H=O atom-ratios. The
zero-addition neutral atmosphere that maintains an unchanging carbon-content of the Fe-C(s) alloy is easily established
by means of eqs. (26) and (27) using the data on RH and RCE
of Table 3. For instance, a feed-gas consisting of 1.0108 mols

792

Y. K. Rao
Table 3 Eect of feed-gas composition on equilibrium in the C-O-H-Fe system at 1300 K and 101.325 kPa (1 atm) for XC 0:05 (or 1.12
w=o C).

RH:

Units

RC:

Units

1.5

2.5

3.5

4.5

5.5

6.5

8.0

3.5

1.0

1.0

1.0

1.0

1.0

1.25

2.5

1.0

nCH4 o
nCO o

mol
mol

0.11111
1.0

0.42857
1.0

1.0
1.0

2.33334
1.0

0.9
0.1

1.15
0.10

2.125
1.00

1.0
1.0

nH2 O o

mol

0.11111

0.42857

1.0

2.33334

0.9

0.90

0.25

1.0

nH2 o

mol

0.5

0.5

0.5

0.5

0.05

0.05

0.50

PCH4 eq

Pa

79.36

133.42

175.10

207.28

232.60

252.96

276.90

PCO eq

kPa

30.846

26.731

23.583

20.044

PCO2 eq

Pa

307.38

185.80

124.33

88.98

66.82

52.02

37.57

PH2 O eq

Pa

401.21

404.45

379.01

348.86

320.25

294.65

262.02

PH2 eq
RCE

kPa

43.205
0.98922

56.021
0.98993

64.176
0.99115

69.823
0.99265

73.965
0.99432

77.135
0.99609

80.702
0.99889

64.204
0.95826

C

mol

0.01211

0.01453

0.01772

0.02435

0.00560

0.25382

1.87628

0.08339

57.331

44.573

36.461

35.444
390.51
1225.2

This run is for an iron-carbon alloy with XC 0:02 (0.44 w=o C).

CH4 , 1 mol CO, 1.0287 mols H2 O and 0.5 mol H2 , with atomratios H=O (or RH) = 3.5 and C=O (or RCE) = 0.99115
provides the desired neutral atmosphere (
C  0) at 1300 K
and 101.325 kPa pressure. Similar feed-gas compositions can
be selected for the other H=O (or RH) atom-ratios listed in
Table 3.
In the present work, only ve gaseous species were
considered signicant; the species OH(g), H(g) and O2 (g)
were not included in the iterative computation as their
concentrations are regarded insignicant. Whether this
assumption is valid can be determined by considering the
three related reaction equilibria (R4), (R5) and (R6); at
1300 K, the standard chemical potentials (
o i ) for OH(g),
H(g) and O2 (g) obtained from published sources2,18) are

227:013, 49.867 and
295:418 kJmol
1 , respectively.
H2 O OH 0.5H2 ; "4 mol; K4 1:402  10
8 R4
0.5H2 H; "5 mol; K5 1:110  10
6
R5

8
H2 O H2 0.5O2 ; "6 mol; K6 8:155  10
R6
Selecting the results of equilibrium calculation for the FeC(s) alloy containing 1.12 w=o C (or XC 0:05) and a feedgas with a H=O atom-ratio of RH 3:5 (column 5, Table 3),
the equilibrium partial pressures of the minor species are
determined as follows:
POH K4 PH2 O =PH2 0:5 6:69  10
6 Pa;
PH K5 PH2 0:5 8:95  10
2 Pa;
PO2 K6 PH2 O =PH2 2 2:35  10
14 Pa
Thus, neglecting the presence of species OH(g), H(g) and
O2 (g) in the C-O-H-Fe system does not have any signicant
eect on the results of the equilibrium calculations.
6.

0.5
52.72

Conclusions

The mass-action equilibrium constant method, derived by

minimizing Gibbs free energy, determines the equilibriumstate in terms of intensive state variables (Pi s); these data are
then combined with extent of reaction formalism1) to deduce
extensive quantities such as the molar amounts of species at
equilibrium.

Acknowledgments
The valuable assistance of Mr. Laszlo Szeleczki, Senior
Computer Specialist of the Department of Materials Science
and Engineering, is duly acknowledged. Thanks are also due
to the sta of the Academic Computer Center, Mary Gates
Hall, University of Washington, Seattle, Washington.
REFERENCES
1) I. Prigogine: Introduction to Thermodynamics of Irreversible Processes, Second Edn., (John Wiley & Sons, New York, 1961) pp. 47.
2) O. Knacke, O. Kubaschewski and K. Hesselmann (Ed): Thermochemical Properties of Inorganic Substances, Second Edn., (SpringerVerlag, Berlin and New York, 1991) pp. 2412.
3) W. B. White, S. M. Johnson and G. B. Dantzig: J. Chem. Phys. 28
(1958) 751755.
4) W. R. Smith and R. W. Missen: Chemical Reaction Equilibrium
Analysis: Theory and Algorithms, (John Wiley & Sons, New York,
1982) pp. 364.
5) G. Damkohler and R. Edse: Z. Elektrochem. 49 (1943) 178186.
6) D. R. Cruise: J. Phys. Chem. 68 (1964) 37973802.
7) D. S. Villars: J. Phys. Chem. 63 (1959) 521525.
8) G. S. Bahn: Kinetics, Equilibria and Performance of High Temperature
Systems. Proceedings of the First Conference, Ed. G. S. Bahn and
E. E. Zukoski, (Butterworth, Washington, D.C., 1960) pp. 137140.
9) H. H. Kellogg: Metall. Trans. 2 (1971) 21612169.
10) Y. K. Rao and C. H. Raeder: J. Mater. Synth. Process. 3 (1995) 4968.
11) Y. K. Rao: High Temp. Mater. Sci. 34 (1995) 293342.
12) S. Gordon, F. J. Zeleznik and V. N. Hu: NASA Technical Note,
(National Aeronautics and Space Adminstration, Washington, D.C.,
1959) D-132.
13) G. Eriksson: Acta Chem. Scand. 25 (1971) 26512658.
14) Y. K. Rao: High Temp. Mater. Sci. 36 (1996) 173183.
15) J. F. Elliott and J. Chipman: Chemical Metallurgy of Iron and Steel,
(Iron and Steel Institute, London, U.K., 1973) 348.
16) S. Ban-ya, J. F. Elliott and J. Chipman: Metall. Trans. 1 (1970) 1313
1320.
17) Y. K. Rao: Stoichiometry and Thermodynamics of Metallurgical
Processes, (Cambridge University Press, Cambridge and New York,
1985), Ch. 11.
18) M. W. Chase, Jr. (Ed): NIST-JANAF Thermochemical Tables, Fourth
Edn., (American Institute of Physics, Woodbury, New York, 1998)
1298.