A Thermodynamic Analysis of the Copper-Sulfur

System
R. C. SHARMA AND Y. A. C H A N G

An associated solution model is applied to describe the thermodynamic behavior of Cu-S

liquid. This model assumes the existence of 'Cu2S' species in addition to Cu and S in the
liquid. With two solution parameters for each of the binaries Cu-'CuzS' and 'CuzS'-S, this
model accounts for the compositional dependence of the thermodynamic properties of Cu-S
liquid from pure Cu to pure S over a wide range of temperature. The binary Cu-S does not
contribute significantly to the excess Gibbs energy of the liquid due to the rather small
dissociation constant of 'Cu2S' to Cu and S. Using this model for the liquid phase, a
statistical thermodynamic model for the digenite phase, and appropriate thermodynamic
equations for the other phases, the Cu-S phase diagram is calculated. The calculated
diagram is in excellent agreement with the experimental data, accounting for the range of
homogeneity of digenite at all temperatures.
AT high temperatures the Cu-S system is characterized mentioned in the previous section, it is postulated'that
by a congruent-melting phase, digenite. 1,2It exists over a 'Cu2S' exist in the Cu-S liquid in addition to Cu and S.
wide range of homogeneity and is represented by the The relative concentrations of 'Cu2S' (l), Cu (l) and S (l)
chemical formula, Cu2.8S. On either side of digenite, are governed by the following equilibrium
liquid miscibility gaps exist. These features of the Cu-S
system suggest there may be strong association in the 'Cu2S'(l) = 2Cu(/) + S(I) [11
liquid in the vicinity of Cu2S composition but strong with an equilibrium constant given by
repulsion between Cu and Cu2S as well as between Cu2S
and S. The behavior is similar to that of the Fe-S system = [ ' u sl[
with the exception that strong association occurs at 50 [Y'cu2S' J LX'Cuy J
pct S in the Fe-S case as opposed to 33.3 at pct S in the where Ten, Ys and Y,cu=s,are the activity coefficients of
Cu-S case. Furthermore, sulfur activity in the Cu-S Cu, S and 'Cu2S' respectively and Xc,, x s and Xcu2sare
liquid from 1423 to 1623 K shows rapid increase with their mole fractions in the liquid phase. These mole
composition in the vicinity of 33.3 at pct S,3,4 similar to fractions are related to the actual compositions of Cu
the behavior of sulfur activity in the Fe-S liquid at 50 at and S, Ycu and Ys, by the following mass balance
pet S / T h i s is also similar to the activity behavior of a equations
component in an ordered intermediate phase at the
stoichiometric composition? The objectives of the Xc~ = Ycu - [2 - 2ycu] X,cu,s, [31
present study are 1) to describe the thermodynamic
behavior of Cu-S liquid using an associated solution Xs = Ys - [1 - 2ys]X,cu2s, [4]
model as has been done successfully for the Fe-S (Ref. The activity coefficients YCu, Ys and 7,c.~s, are described
5), Co-S (Ref. 7) and Ni-S (Ref. 8) liquids and 2) to in the present study by the so-called three-suffix
calculate the phase diagram of the Cu-S binary using Margules equation as given below 9
this model for the liquid, a statistical thermodynamic
model for digenite and appropriate thermodynamic lnyi = 1/2 ~] (Wis + W j i ) x j
equations for the other phases. The ultimate aim is to j=l
provide a set of thermodynamic values for all the phases
-- 1/2 ~ ~ Wjpxjx,
in this system which would be consistent with the phase j = l p=l
diagram.
1.0 T H E R M O D Y N A M I C MODELS
FOR THE VARIOUS PHASES
[51
1.1 The Liquid Phase j=] p=l
In view of the features of the high-temperature phase where i, j and p refer to Cu, S and 'Cu2S', respectively,
relationships of the Cu-S system and the activity W~i, Wi~ and so forth are the interaction parameters of
behavior of sulfur in the vicinity of 33.3 at pct S as the solution, x~, x i and Xp are the corresponding mole
fractions and W , = Wjj = 0 and so forth. The
R. C. SHARMA, formerly with the Materials Department, Uni-
versity of Wisconsin-Milwaukee, is now Lecturer, Department of
interaction parameters are in general represented as
Metallurgical Engineering, Indian Institute of Technology, Kanpur, A
India. Y. A. CHANG, formerly Professor of Materials Engineering W,j = ~ + B [61
and Associate Dean for Research of the Graduate School of the
University of Wisconsin-Milwaukee, is now Professor of Metallur-
gical and Mineral Engineering, University of Wisconsin-Madison,
with A and B being constants, whose values must be
Madison, WI. determined from experimental data. The corresponding
Manuscript submitted October 29, 1979. expression for the excess Gibbs energy of mixing, AG x',
ISSN 0360-2141/80/1211-0575500.75/0
METALLURGICAL Tt~ANSACTIONS B 9 1980 AMERICAN SOCIETY FOR METALS AND VOLUME 1 IB, DECEMBER 1980--575
THE METALLURGICAL SOCIETY OF AIME
is B' B"
AGxs
l n K l = A ' + - ~ , l n K2 = A " + ~ - [151
-R-T= 1/2 ~ ~ [Wijxix j + (Wii - Wji)xixf]
j = l i=l and
[7] B'" B'"'
where R is the gas constant and T is the absolute u =A'"+ ~-,v = A .... + T [16]
temperature. When W~j = Wj~, Eqs. [5] and [7] reduce
to the appropriate expressions for a regular solution 1.3 The Cu-Phase
with one interaction parameter for each of the three
Since the solubilities of sulfur in Cu are small, the
binaries.
following expressions are used to describe the ther-
The activities of Cu and S in the liquid phase are
m o d y n a m i c properties of this phase
as = ysXs [81 G~ = G~ ~' + R T l n y s [17A]
and
and
a c u = YcuXcu [9]
G ~ = G ~ + R T l n (1 - Ys) [17B]
The standard rates are Cu (l) and S (l).
where Ys is the atom fraction of S in (Cu) and G~v and
G ~ are the Gibbs energies of sulfur and copper in
1.2 The Digenite Phase their standard states. The standard state of sulfur is
purely artificial and corresponds to that of a solution at
Rau m~ and NagamorP 2 attempted to develop ther-
infinite dilution while that of copper is pure copper.
m o d y n a m i c equations to describe the thermodynamic
behavior of digenite as a function of composition and
temperature. Neither the equations of R a u " nor those 1.4 The Hexagonal fl-Cu2S and the Covellite (CuS)
of NagamorF 2 are able to describe the thermodynamic Phases
data of digenite over all the compositions of interest.
The phases, fl-CuzS and CuS are essentially line
More recently, Sharma and Chang L3developed a model
compounds and their Gibbs energies of formation are
which accounts for all the data available for digenite
over a large temperature and compositional range. taken to vary linearly with T.
Their equations are used in the present study.
The assumptions made by Sharma and Chang j3 are 2. E V A L U A T I O N S OF T H E
i) copper interstitials exist in the lattice and ii) CUES and THERMODYNAMIC PROPERTIES
'CuS' (with the structure of digenite) form a subregular OF T H E V A R I O U S PHASES
solution. With these assumptions, the following chem-
ical reactions may be written 2.1 Pure Sulfur and Copper

1/4 S2(g) + 1/2 Cu2S(6 ) = 'CuS'(6)
1/4 S2(g) + Cu, = 1/2Cu2S(6 )
where Cui stands for copper interstitials. The equilib-
rium constants for the above reactions are
[10] Since the activity of sulfur in condensed metal-sulfur
phases is usually determined by equilibration with a
mixture of H2S/H2 gases, activity of sulfur is usually
[11]
referred to a standard state of 1/2 Sz(g). To convert
the standard state from 1/2 S2(g) to S(/), the follow-
ing equation is used in the present study

K1 = 7'Cus'X'c~s' [ 11A] G~ t - 1/2 Gs,Og__

- - 6 5 , 3 5 7 + 165.396 T
r11/4 tl/2 t ' l
F S z /'Cu2S~, ~ X,CuS,) 1/2
- -

- 13.513T In T
1/2 [ 1
K2 = 3'curse- - X,cus,)1/2 in J/mol, 800 K < T < 2000 K
p $2
l/4XCui [ 12A]
[18]

where Ps is the partial pressure of $2, ~',cos,and ~'c~s are
the actwlty coefficients of 'CuS' and Cu2S in the
solution of digenite, and x,c~s, , xc~' are the concentra-
tions of 'CuS' and Cu~ with the constraint that X,c~s,
+ Xcu~S = 1. The concentrations of 'CuS' and Cui are
related by the following mass balance equation
8 = X,cus, - Xcu'
The following relationships are used to represent
In 7c.~s, In 7,c~s,, In K1 and In K2:
In Yc.,s = [u + VX,cus,]X2,c,s,
In Y,c,s, = [u + 3/2 v - v(1 - X,c,s,) ] (1 - X,c,s,)2
Equation [18] is obtained based on the J A N A F data? 4
For the Gibbs energy difference between Cu (s) and
Cu(l), the values of A H " = 13,054 J / g - a t o m and AS"
= 9.623 J / K g - a t o m , are taken to be temperature in-
dependent, is
2.2 The Liquid Phase
[13]
Values of W i f s and K in Eqs. [2] and [5] are obtained
primarily based on the sulfur activity data available in
the literature 3,4,~6,1vsubject to the constraint that the
[14A] calculated phase diagram involving the liquid phase
must agree with the experimental phase diagram data.
A standard nonlinear regressional analysis technique ~8
is used to obtain the values of W i f s and K as well as to
[14B] carry out the phase diagram calculation as given in a

5 7 6 - - V O L U M E 11 B, D E C E M B E R 1980 METALLURGICAL TRANSACTIONS B

later section. The values of W u's and K are concentrations, but more negative at low sulfur con-
10,385.9 centrations. A much better fit may be obtained between
In K -- - 2.6500 [191 the calculated and experimental values using the
T present formalism by optimizing the values of W u's and
K. However, if these values of In a~ are used, the
2,026.1
Wc,,cu s, = 1.8070 + - - [20A1 calculated phase diagram would not be in agreement
- ]
T with the experimentally determined phase diagram.
Since there is no reason to doubt the correctness of the
2,770.4
W,cu#,.c, = 2.1649 + - - [2oB] known phase diagram, the values of K and Wu's as
T given in Eqs. [19] through [20D] are used in the present
study to describe the thermodynamic properties of the
8,812.4
W,c,2s,.s = 0.0268 + ~ [20C] Cu-S liquids.
Since the dissociation constant of 'Cu2S (l) is small,
and the term R T In K corresponds to the Gibbs energy of
1,863.8 formation ~G ~ for 'Cu2S' (l) from Cu (l) and S (l).
Ws,c,2s, = - 1.6366 T [20D] Using Eq. [18], the Gibbs energy of formation of
'Cu2S' (l) from Cu (l) and 1/2 S2(g) is
Equations [19] through [20D] are valid from pure Cu to
35 at pct S over the temperature interval, 1423 to AG,%ms,1 = -151,701 + 143.365 T - 13.513 T i n T
1623 K. However, since the calculated phase diagram
in J/mol, 1423 ~ T ~ 1623 K [21]
from pure Cu to S is in good agreement with the
experimental data, it is reasonable to expect that these As shown in Fig. 3, the values of AG ~ calculated from
equations are valid at much higher sulfur concentra- Eq. [21] agree well with the selected values of King et
tions and outside the temperature interval, 1423 to al. 19 They also agree with the values of Bale and Toguri 3
1623 K. Due to the rather small values of K as and of Kellogg 9 at low temperatures. At 1523 K, the
evidenced from Eq. [19], the product x c , s s is very small. value of AG ~ from Eq. [21] is about 1 kJ more negative
As a result, the contribution of the interaction between or 1 pct.
Cu and S to the excess Gibbs energy must be extremely
small. For practical purposes, the values of Wc,_s and
Ws_c, are taken to be zero.
2.3 The Cu-Phase
Comparisons between the calculated and experimen-
tal sulfur activity data are shown in Figs. 1 and 2. The Using the solubility data 2~of sulfur in (Cu) and the
agreement is reasonable for sulfur composition up to 2 Gibbs energy of formation of digenite, the following
at. pct S, in view of the uncertainties obtained by the Gibbs energy equation for sulfur in (Cu) is obtained,
three groups of investigators. The comparison for sulfur
G~r - 1/2 Gs~ -25,200 + 21.00 T J/g-atom
concentration from 32 to 35 at. pct S is given in Fig. 2.
The calculated values are more positive at high sulfur 850 < T < 1300 K [22]

-9 I I I

/"

• 1523
•

-I0
YJ. o

/A~'/ + +o ..... PHASE BOUNDARY

#
Fig. 1--The activity of sulfur in Cu-S liquid
from Cu to 3 at. pct S; comparison between
the calculated and experimental values. The
d ~o
o # IY ~
~("/o
i
A ,473 ~ BALE end TOGURI
standard state is S (l).
7m ,523 j
-II -- 9 1420 1
9 1 4 7 0 1 SUDO
| 1520
+ 1423 }
9 1473 YAGIHASHI
9 1523

-12 I I I
0 O.Ol 0.02 0.03
Ys
METALLURGICAL TRANSACTIONS B VOLUME 11B, DECEMBER 1980--577
 -3-- I I [ I
01423] • ~x _
-4- : 14737 BALE (Ind TOGUR, / "
1523 j x o-
,~ ,423 ] ~ ~ _
A 1523 ~ SCHUHMANN and MOLES
+ ,62'3J ~/xf _

........ PHASE BOUNDARY f

J03
Fig. 2 The activity of sulfur in Cu-S liquid
r ///P, from 32 to 36 at. pet S; comparison between
the calculated and experimental values. The
+ + x / # / ~ " standard state is S (l).

,523" 9 -

-10 - o
\

-II i I I I
0.32 0.33 0.34 0.35
Ys

2.4 The Digenite Phase, 8-Cuz_sS 3. P H A S E D I A G R A M C A L C U L A T I O N

The following values of A ', A ", A '", A .... , B', B", B ' " The Cu-S phase diagram is calculated using the
and B .... obtained by S h a r m a and C h a n g ~3based on the condition that the chemical potentials of the c o m p o n e n t
data of R a u H are used in the present study
A' = -4.4919 B' = - 1,644.8 -8o I
A" = 0.8274 B" = 13,186.5
A"
A'"'
= 2.6301
= - 1.2877
B'"
B'"'
=
=
-3,576.1
- 8,300.2
1
KING .
789 K ~ T ~ 1321 K -a2 (INCRA)

The selected Gibbs energy of formation equation for .~/ r----THIS WORK
digenite at the stoichiometric composition, 8-Cu~S, is
from A r n d t and Kordes 21 as given below,
-84
AGscu.2s = - 125,194 + 25.828 T J / m o l "~
800 K ( T ( 1300 K [23] ~n
Values of AG ~ reported by other investigators ~z16,n 27 (-9 9 BALE and TOGURI
are also shown in Fig. 4 for comparison. <~ -86 o KELLOGG

2.5 The Hexagonal fi-Cu2S and the Covellite (CuS)

Phases
-88
Both the fl-Cu2S and the CuS phases are essentially
line c o m p o u n d s . Their G i b b s energies of formation
used in the present study are
AG#~ = - 127,027 + 28.417 T J / m o l
-901 f I
400 ~ T ~ 750 [24] r400 1500 t600 1700
TEMPERATURE, K
AG,~ = - 9 8 , 4 5 0 + 52.10 T J / m o l
Fig. 3 ~ T h e Gibbs energy of formation of liquid 'Cu2S' as a function
400 ~ T ~ 800 [25] of temperature; comparison between the values obtained in the
present study (R T In K) and those reported in the literature. The
The standard states are Cu(s) a n d 1/2 S2(g ). standard states are Cu (l) and 1/2 S2(g).

578--VOLUME 11B, DECEMBER 1980 METALLURGICAL TRANSACTIONS 8

 I I I I I I I I
-20
-- -8.5

-21
O
E -90
x.
-~ -22
SUDO
f NAGAMORI ---95
Fig. 4--The Gibbs energy of formation of <~ E
stoichiometric digenite, CuzS, as a function of -23 RICHARDSON ond ANTILL
temperature. " ~ / , ~ i ARNDT ond KORDES
COX ET AL d
qoo <3
--24 YAGIHASH[ and SATO
b g TANAKA ET AL
h KIHIRA
/_dC / / ~ [ BROOKS
-25
-I05

-26 I I [ I I I
800 900 I000 liD0 1200 1300 1400 1500 1600 700
TEMPERATURE, K

elements in a two-phase field are equal, i.e. I i I i ,,,, , I ' '

G? = Gf [26A]
1600

it
or
/ 9 SCHUHMANN end MOLES
o
9 JOHANNSEN end
G~~ + R T l n ~ = G~ + R T l n a ~ [26B]
/
VOLLME"
9 SMITH .-
where i stands for either Cu or S and the superscripts a
o--'.-'B-JUDIN end EEROLA
and fi indicate any phase. The Gibbs energies of all the - - ' - - SELECTED by KELLOGG
relevant phases in the Cu-S binary are given in the 1500 - i
iv
previous section. Using these values for any two phases,
Y o
Eqs. [26] may be used to solve for the compositions of L1 L~ + L 2 ic
Lz
the coexisting phases as a function of temperature.
Numerical evaluation is carried out using a standard
nonlinear regressional analysis techniqueY oi
Figure 5 shows a comparison between the calculated
Q.- t400 Lz, S ----.
and experimental data between pure Cu and the I 378 l/,
&phase. Good agreement is obtained between the Ld
I'--
calculated and experimental diagram, as shown in this LI+ S
1340
figure and in Table I. At temperatures higher than
1450 K, the calculated phase boundaries for the L~
+ L 2 miscibility gap are wider than the data of Bale (Cu)+L~
~3oo~- (Cu) + 3
and Toguri; 3 however, they are in reasonable agreement
with the data of Schuhmann and Moles, 4 Johannsen
and Vollmer, 2s and Smith. 29 Since an evaluation of the
relevant data between Cu and 3 is completed, the data
of Judin and Eerola36become available. Their data on
the L 2 + L 2 phase boundaries from 1423 to 1580 K are
shown in Fig. 5. Also shown in this figure are the phase 0 0.05 0.30 0.35
boundaries selected by Kellogg9 and Judin and Eerola. 36 Ys
It is interesting to note that the L ~ / L t + L 2 phase Fig. 5--The copper-digenite phase diagram; comparison between the
boundaries of Judin and Eerola agree with those calculated phase diagram and the experimental data. See Table I for
the compositions of the coexisting phases at the invariant equilibria.
selected in the present study but their L 2 / L I + L 2
phase boundary values do not. On the other hand, the
L 2 / L 1 + L 2 phase boundaries according to Kellogg data. For instance, if the L 2 / L l + L 2 phase boundary
agree with our values while his L ~ / L I + L 2 values do data of Ref. 36 were extrapolated to lower temperatures,
not. However, when considering the uncertainties of the they will not coincide with the composition of the
experimental data, the discrepancies given by all of us monotectic liquid L 2 at the invariant equilibrium tem-
are unimportant. However, it is important to point out perature. Figure 6 shows in detail the good agreement
that the L~ + L 2 phase boundary values given in the between the calculated and experimental values of
present study are consistent with the thermodynamic Oudar 2~for the solubility of sulfur in (Cu), from 800 to
properties of the liquid phases and other phase diagram 1340 K. Figure 7 shows the calculated Cu-rich phase

METALLURGICAL TRANSACTIONS B VOLUME 1 IB, DECEMBER 1980--579

 1500

1400 L~ L + (Cu)

1300
,3oo-
(Cu) ~ o
~.- 1200
?-- Fig. 6 ~ T h e solubility of sulfur in the copper
~ (Cu)* 8 phase; comparison between the calculated
sulfur-rich phase boundary of (Cu) and the
E I100
experimental data.
f,// o + IN HIGH PURITY COPPER (99.999%) ]

I000 o o IN OFHC COPPER ..~ OUDAR

900

800 I _1 i 1 I l I I I J J i i [ i I i L I i L I J
0.5 I.O 1.5 2.0 2.5
AT FRACTION OF Cu x 10 4

1400
T ~ f ~ composition of 36.29 at pct S for the 6-phase is nearly
the same as that given in the Metals Handbook, i.e.,
36.25 at. pct S. Also shown in Fig. 8 are the experi-
1300
mental values of Roseboom 3~ and Merwin and
- - CALCULATED
Lombard, 32 for the sulfur-rich phase boundary of 6. The
1200 . . . . COOK
calculated phase boundary is in accord with these data.
The calculated peritectic temperature for the formation
of CuS agrees with the literature value as given in Table
W
- II00
j! I. The solidus and liquidus curves of the 6-phase as
reported by Johannsen and Vollmer 28 as shown in Fig. 8

/
(Cu)+ S must be in error. While the calculated liquidus curve is
n" I 0 0 0
UA
in reasonable agreement with the data of Cookp ~ the
EL calculated solidus curve is lower in sulfur concentration.
UJ
i--
However, an extrapolation of Cook's solidus curve to
90O the eutectic temperature would not be in agreement
with the data of Roseboom 31 and Merwin and
Lombar& 2 at lower temperatures. Figure 9 shows in
8 0 0 -- detail the calculated phase relationships involving the
&phase. Figure 10 shows the good agreement between
708 the calculated and experimental phase boundaries of
700
the &phase at temperatures below 708 K. 31,33,34
Figure 11 summarizes the calculated phase diagram
6O0
of the Cu-S binary. Superimposed on this diagram are
0.000 0.001 0.002 the isobar values of S2(g) up to Ps2 (g) = 1 atm.
(3Ys-,/
4. DISCUSSION
Fig. 7--Comparison between the calculated and experimental cop-
per-rich phase boundary of digenite from 700 to 1350 K. Kellogg9 recently analyzed the thermodynamic data
of Cu-S liquid also using an associated solution model,
postulating the existence of 'CuS' and 'Cu2S' species in
boundary of 6 in agreement with the experimental value the liquid phase. One shortcoming of this approach is
of Cook. 3~With increasing temperature, the concen- that the thermodynamic equations developed are not
tration of copper increases, suggesting the model used valid for compositions beyond 50 at. pct S. Moreover,
in the present study for digenite is valid. Figure 8 shows the assumption of 'CuS' species in the liquid phase may
the calculated and experimental phase diagram between not be physically realistic since the intermediate phase
6 and S. The calculated eutectic temperature of 1085 K CuS melts peritectically at a rather low temperature of
is in good agreement with the literature data given in 780 K and a miscibility gap exist at this composition in
Table I. The calculated eutectic composition of 40.09 at the liquids. It is interesting to note that in correlating
pct S is 1 at. pct less in sulfur content than the value the sulfur activity data from 32 to 35 at. pct S in terms
given in the Metals Handbook, ~but the calculated of his equations, Kellogg found that the data of Bale

580--VOLUME 11B, DECEMBER 1980 METALLURGICAL TRANSACTIONS B

 T T
1404 L2
1400[
. . . . J O H A N N S E N and V O L L M E R
..... COOK
1300

L2+ L3

1200

v
Cu)+~ 3+L 2
I100 1085

Fig. 8--The digenite-sulfur phase diagram; I.--

<
comparison between the calculated diagram Q::
L,J + ROSEBOOM
and the experimental data. (3.. I000
o MERWlN and L O M B A R D
Ld
I'--
8+ L3
900

800 780

S +CuS
700~ d _ _ ~ l _ _ L J t .1~
0.. 0.34 0.35 0.36 0.37 0.38 0.39 0.40 OAt 0.42
Ys

1500,

1400 Lz
I I 80o
750
~s
4

708
1300 700 + CuS

~+L 2 ba
* ROSEBOOM
1200 65~

9 WEHEFRITZ
,r 9 (Cu] + S hl o RUHL end SAUR ~CuS
60C
u" I lOC 1085 taJ
rr
55C
< I000
n~ ~,+L 3
W
o_ B 50s
90C
w

45O i I I I . :: L I
0.33 0.34 0.35 0.36 0.37 0.38 0.49 0.50 0.51
800 780 Ys
.~ICu)+/3 / Fig. 10~The phase diagram from 33 to 51 at. pct S at low
temperatures; comparison between the calculated phase diagram and

i- //3 8+ CuS
the experimental data.

and Toguri at 1473 K are higher than the values

L /
6 0 0 --
calculated from his equations at high sulfur concen-
tration. Although he could adjust the values of the
5 0 0 -- parameters to have a near-perfect fit between the
L I I experimental and calculated values at 1473 K, these
0.33 0.34 0.35 0..36 0.37 0.38 parameters would yield poor agreement at other tem-
Ys peratures. This pointed out the difficulty in obtaining
Fig. 9--The calculated phase equilibria involving digenite, the 8- the best possible values of the parameters when exper-
phase. imental data obtained by different investigators are not

METALLU RGICAL TRANSACTIONS B VOLUME 11B, DECEM BER 1980--581

 AT%S
o24 f3, 33 35 37 39 41 43 ,99 99.5 I00

/
1 II I I

1600 - \ ,10-4 2
i L gO i10o
1500 ~- ~ 10-6
li@l . . . . . . . .
[LII L~ + L2 LZ ] L2 + L 3
,4o0 r ~/~-4" ,4o; i
1500- lit\ x
10-8

1200

(Cu) + S
w~ I I 0 0
n~
I0-~~ Fig. 1 l ~ T h e c a l c u l a t e d c o p p e r - s u l f u r ( C u - S )

I000 I
;;~i
I \
1 p h a s e d i a g r a m w i t h i s o b a r v a l u e s of S2(g) u p
to 1 a t m ; 1 a t m = 1.01325 • 105 Pa.
W
~+L 3
Q_
:/ / \
W 900 _ I0 -m2 i0 ~
1--

8OO 78o ~;~ j

~o-~ a\x\
70O
7os Sl-- 10-2 Cu S . ~ " 10 -2

(Cu)+,8 io_~O_L
~
'f., ~ \ \
+ CuS
, , , . . CuS + L3_
60 - #+S ---- -I' \ I 10 - 4

i O-IZ_.~ ~ \ \ ', 10-6

50C - /3----, ~\ ,.\ --

0 I 2 19 20 21 22 23
I I
24
I
25
I
26
I
27
ft
)J
I
33 ' oo
WT% S

Table I. Comparison Between the Calculated and the Literature Values for the Various Invarlant Equilibria in the Cu-S System*

T h e V a l u e s G i v e n in The Values Given by

The Invariant Equilibria The Calculated Values the M e t a l s H a n d b o o k ~ Craig and Scott 2

L] = (Cu) + 8 1340 K 1341 K 1340 K

L I 0.0152 0.0167 --
(Cu) 2.25 x 10 -4 --I" --
0.3335 0.3338 --

L2 = L I + ~ 1378 K 1380K 1378 K

L2 0.3282 0.3285 --
Li 0.0170 0.0187 --
8 0.3335 0.3338 --

L2 = 8 1404 K 1404 K 1402 K

0.3345 0.3345 --

L2 = 8 + t 3 1085 K 1087 K 1086 K

L2 0.4009 0.4105 --
~$ 0.3629 0.3625 --
L3 0.9983 -- --

(Cu) + 8 = fl 708 K 708 K 708 K

8 0.3334 0.3334 --

+ L 3 = CuS 780 K 780 K 780 K

0.3646 0.3652 --
Z3 0.9998 -- --
* C o n c e n t r a t i o n in at. p c t S.
1"See Fig. 6.

582--VOLUME 11B, D E C E M B E R 1980 METALLURGICAL TRANSACTIONS B

in good accord. Kellogg also checked the consistency
between the thermodynamic values of the liquids and
the phase boundaries of the metal-rich miscibility from
1423 to 1523 K. However, he did not attempt to obtain
an internal consistent check between the thermody-
namic values of the liquids obtained by him with the
known phase diagram data and the thermodynamic
values of the other phases in the system.
The associated solution model used in the present
study, on the other hand, assumes the existence of only
'Cu2S' in the liquid. The assumption of strong asso-
ciation at 33.3 at. pct S is consistent with the known
thermodynamic, phase diagram and electrical conduc-
tivity data? 5 Furthermore, the thermodynamic equa-
tions developed in the present study assuming the
presence of Cu, 'Cu2S' and S are valid over all compo-
sitions from pure Cu to pure S. The agreement between
the experimental data of Schuhmann and Moles and the
calculated values as shown in Fig. 2 are fair; however,
the agreement between the data of Bale and Toguri and
the calculated values are less good. As stated in section
2.1, the values of the parameters may be varied in order
to have a better fit between all the experimental and the
calculated values. However, the resulting thermody-
namic equations for the liquid phase are not consistent
with the known phase diagram data and the well-known
thermodynamic data of digenite. The solution param-
eter values as given in Eqs. [19] through [20D] are
obtained based not only on the sulfur activity data but
also on internal consistency between the thermody-
namic values of all the involved phases and the phase
diagram data. Possible reasons for the discrepancies
between the experimental and predicted sulfur activity
data as shown in Fig. 2 are 1) inadequacy of the
associated solution model used in the present study to
describe the thermodynamic properties of the Cu-S
liquid phase, 2) incorrect sulfur activity data for the
liquid phase and 3) incorrect phase diagram data. It
seems the most likely cause is that of 2) and next likely
cause is that of 1). The phase diagram data involving
(Cu), L 1, L 2 and 6 are believed to be correct.
5. ACKNOWLEDGMENT
The authors wish to thank the Division of Materials
Research of the National Science Foundation NSF-
DMR-78-04066 for financial support of this research.
REFERENCES
1. Metals Handbook, 8th Ed., vol. 8, p. 297, ASM, Metals Park, OH,
1973.
2. J. K. Craig and S. D. Scott: Sulfide Mineralogy, P.H. Ribbe, ed.,
voL 1, chapt. 5, Mineralogical Society of America, Washington,
D.C., 1974.
METALLURGICAL TRANSACTIONS B
3. C.W. Bale and J. M. Toguri: J. Thermal. Anal., 1971, vol. 3, pp.
153-67.
4. R. Schuhmarm, Jr. and O. W. Moles: Trans. AIME, 195 I, vol.
191, pp. 235-41.
5. R.C. Sharma and Y. A. Chang: Met. Trans. B, 1979, voL 10B, pp.
103-108.
6. Y. A. Chang: Treatise on Materials Science and Technology, H.
Herman, ed., vol. 4, pp. 137-259, Academic Press, 1974.
7. R.C. Sharma and Y. A. Chang: Z. Metallkd., 1979, vol. 70, pp.
104-108.
8. R. C. Sharma and Y. A. Chang: Met. Trans. B, !,980, vol. l IB, pp.
139-46.
9. H. H. Kellogg: Physical Chemistry in Metallurgy, R. M. Fischer,
R. A. Oriani and E. T. Turkdogan, eds., p. 49, U.S. Steel Research
Lab., MonroeviUe, PA, 1976.
10. H. Rau: J. Phys. Chem. Solids, 1974, vol. 35, pp. 1415-24.
11. H. Rau: J. Phys. Chem. Solids, 1967, vol. 28, pp. 903-16.
12. M. Nagamori: Met, Trans, B, 1976, vol. 7B, pp. 67-80.
13. R. C. Sharma artd Y. A. Chang: Chinese& Mats. Sci., 1979, vol.
1t, pp. 58-62.
14. JANAF Thermochemical Tables, 2nd Ed., NSRDS-NBS 37, 1971.
15. R. R. Hultgren, P. D. Desai, D. T. Hawkins, M. Gleiser, K. K.
Kelley, and D. D. Wagman: Selected Values of the Thermodynamic
Properties of the Elements, ASM, Metals Park, OH, 1973.
16. K. Sudo: Sci. Rep, Res. Inst. Tohoku Univ., 1950, Series A, vol. 2,
pp. 513-18; pp. 519-30.
17. T. Yagihashi: Nippon Kinzoku Gakkai-Si., 1953, vol. 17, pp.
483-87.
18. Non-Linear Regression Routines, Mathematical Routine Series,
Academic Computer Center, University of Wisconsin-Madison,
1972.
19. E. G. King, A. D. Mah, and L. B. Pankratz: INCRA Monograph
II, Thermodynamic Properties of Copper and Its Inorganic Com-
pounds, the International Copper Research Association, Inc., New
York, NY, 1973.
20. J. Oudar: Compt. Rend., 1959, vol. 249, pp, 259-61.
21. D. Arndt and E. Kordes: Z. Anorg. Allg. Chem., 1968, vol. 359,
pp. 1-13.
22. F. D. Richardson and J. E. Antill: Trans. Faraday Soc., 1955, vol.
51, pp. 22-23.
23. E. M. Cox, M. C. Baehelder, N. H. Nachtreib, and A. S. Skapski:
Trans. A1ME, 1949, vol. 185, pp. 27-31.
24. T. Yagihashi and T. Sato: Nippon Kinzoku Gakkai-Si., 1952, vol.
16, pp. 482-86.
25. T. Tanaka, K. Watanabe, and J. Kurihara: Mere. Fac. Eng.
Hokkaido Univ., 1950, vol. 8, pt. 1, pp. 14-22.
26. A. Kihira: Res. Rep. Fac. Eng. Nagoya Univ., 1950, vol. 3, pp.
39-42.
27. A. A. Brooks: J. Am. Chem. Soc., 1953, vol. 75, pp. 2464-67.
28. F. Johannsen and H. Vollmer: Z, Erzbergbau. Metallhf~ettenwes.,
1960, vol. 13, pp. 313-20.
29. C. S. Smith: Trans. AIME, 1938, vol. 128, pp. 325-36.
30. W. R. Cook, Jr.: Nat. Bur. Stand. (U,S.) Spec. Publ., 1972, no.
364, pp. 703-12.
31. E. H. Roseboom, Jr.: Econ. Geol., 1966, vol. 61, pp. 641-72.
32. H. E. Merwin and R. H. Lombard: Econ. Geol., 1937, vol. 32, no.
2 (Suppl.), pp. 203-84.
33. V. Wehefritz: Z. Phys. Chem. (Frankfurt am Main), 1960, vol. 26,
pp. 339-59.
34. W. Ruhl and E. Saur: Ber. Oberhess. Ges. Nat. Heilkd. Giessen,
Naturwiss. Abt., 1957, vol. 28, pp. 35-47.
35. E. A. Dancy and G. Derge: Trans. TMS-AIME, 1963, vol. 227,
pp. 1034-38.
36. V. K. Judin and M. Eerola: Seand.J. Metall., 1979, vol. 8, pp.
128-32.
VOLUME 11B, DECEMBER i980--583