Thermodynamic properties of alloys and phase equilibria in the In-Sb system

Article abstract of DOI:10.1023/B:INMA.0000027589.05401.9c

New thermodynamic data for the In-Sb system are obtained and compared with calculation results available in the literature. An isothermal vacuum cell and a procedure for electrolyte preparation are described which ensure high measurement accuracy.

Full text of DOI:10.1023/B:INMA.0000027589.05401.9c

Inorganic Materials, Vol. 40, No. 5, 2004, pp. 445–450. Translated from Neorganicheskie Materialy, Vol. 40, No. 5, 2004, pp. 524–529.
Original Russian Text Copyright © 2004 by Vasil’ev.
Thermodynamic Properties of Alloys and Phase Equilibria
in the In–Sb System
V. P. Vasil’ev
Moscow State University, Moscow, 119992 Russia
e-mail: vvassilev@veernet.iol.ru
Received November 4, 2003
Abstract—New thermodynamic data for the In–Sb system are obtained and compared with calculation results
available in the literature. An isothermal vacuum cell and a procedure for electrolyte preparation are described
which ensure high measurement accuracy.
INTRODUCTION
to 500°C in an electrical furnace. To remove oxychlo-
rides, the molten salts are treated with HCl and Cl
2
III–V semiconductors and their solid solutions find
wide application in optoelectronics and high-speed
electronic devices.
The In–Sb system is often used to check the accu-
racy of measurements with liquid-electrolyte electro-
chemical cells. In addition, In–Sb alloys are used as
internal standards in studies of multicomponent In-
based systems.
gases. After drying under optimal conditions, the melt
contains insignificant amounts, if any, of oxychlorides
(
dark gray flakes). In addition, the melt may contain
carbon particles resulting from the carbonization of
organics. The melt temperature is then increased to
6
00–650°C, which ensures rapid burnout of the carbon.
The presence of carbon particles in the salt melt may
lead to the shorting of the cell electrodes.
The phase relations in the In–Sb system are very
simple (Fig. 1) [1]. It contains only one intermediate
phase with a nearly perfect stoichiometry (zinc-blende
structure, a = 0.6479 nm, congruent melting at 800 K).
In removing oxychlorides, one can use commer-
cially available or laboratory-produced HCl and Cl2
gases. Hydrogen chloride can easily be synthesized by
reacting KCl or NaCl with concentrated H SO . The
InSb and Sb form a eutectic at x = 0.688 with a melt-
Sb
2
4
ing point T = 767.3 K. The In-rich eutectic is nearly
e1
appropriate apparatus and the preparation of gaseous
compounds of HCl, HBr, and HI are described in [2].
The apparatus should contain no rubber or plastic tub-
degenerate, with a melting point (T = 427.5 K) close
e2
to that of pure indium (429.75 K).
This paper reports on the thermodynamic properties ing. All its parts must be made of glass or quartz and
of In–Sb alloys and describes in detail the emf measure-
ment procedure developed earlier, which was repeat-
edly tested in studies of various metal and semiconduc-
tor systems.
T, K
9
8
7
6
5
00
00
00
00
00
EXPERIMENTAL
8
00.0
Electrolyte preparation. This step is of key impor-
tance in emf studies. The optimal process for salt dehy-
dration, developed earlier in our laboratory, is as fol-
lows: Salts are first dried in vacuum (0.1–1 Pa) for 30–
0 h without heating. Next, the temperature is gradually
3–5 days) raised to 150–200°C, and the hot salts are
7
67.3
4
(
transferred to gas-tight containers. The salts are
weighed and mixed in a dry box and then again closed
in the containers. If there is no dry box, the most hygro-
scopic salt is weighed in a gas-tight container of known
weight, and less hygroscopic salts are weighed by a
standard procedure. The salt mixture is dehydrated in
vacuum for about 1 day at a temperature slightly above
4
27.5
4
3
00
00
In
0.2
0.4
0.6
0.8
Sb
xSb
1
00°C and then transferred to a silica beaker preheated
Fig. 1. Phase diagram of the In–Sb system [1].
0
020-1685/04/4005-0445 © 2004 MAIK “Nauka/Interperiodica”

4
46
VASIL’EV
salts (InCl and LuCl ), it is possible to dispense with
3
salt additions, since the salts form when hydrogen chlo-
ride or chlorine (dissolved in small amounts in the elec-
trolyte) is brought into contact with the metal:
To vacuum
pump
In(l) + HCl(g)
Lu(s) + HCl(g)
InCl + H (g),
(1)
(2)
2
Container for
the electrolyte
LuCl + H (g).
3
2
Electrolyte
ingot
We performed experiments with In-, Sn-, Zn-, Pb-, and
Lu-based alloys with (0.05 wt %) and without additions
of a salt of the potential-forming ion [6–10]. The addi-
tions were found to have an insignificant effect on the
emf, causing a slight decrease in the emf equilibration
time (≤2 days).
Tungsten
current leads
Isothermal electrochemical cell. Figure 2 shows a
schematic of the isothermal Pyrex cell used in this
study. The lower part of the cell (below the dashed line)
is 54–58 mm in diameter and ~90 mm in height. The
tungsten current leads and the electrodes attached to
them are soldered in inlet tubes 8 mm in diameter. The
bottom of the cell has cruciblelike holes, which enables
studies of both solid and liquid alloys, with no risk of
accidental mixing.
Thermocouple
casing
Alloy
Reference
electrode
The upper part of the cell, ~400 mm in length and
2
5 mm in diameter, fitted with a ground-glass joint,
serves as a container for the electrolyte. The time
needed to withdraw the ingot from the tube, introduce
it into the container, and connect the container to the
vacuum system does not exceed 10 s.
After pumping the cell (10 to 10 Pa) for a day,
followed by flushing with purified argon, the ingot is
melted under dynamic vacuum using a portable gas
torch. The melt drains down into the lower part of the
cell, which is introduced into a microfurnace heated to
Fig. 2. Schematic of the isothermal cell.
must be fitted with ground-glass joints. A proven lubri-
cant for such joints is concentrated H SO .
–
3
–4
2
4
Hydrogen chloride must be dried using zeolites
loaded into a U-tube. (Phosphoric anhydride is unsuit-
able because phosphoric acid vapor is transported by
the gas stream to the electrolyte.) The zeolite is regen-
erated by calcination at 300°C immediately before
assembling the apparatus. The gas is bubbled through
the melt until there are no suspended particles (~1 h).
The electrolyte must be transparent, with a light yellow
color owing to the dissolved hydrogen chloride. After
the hydrogen chloride flow is turned off, the coloration
disappears rapidly because of the HCl volatilization.
The melt thus prepared is poured into Pyrex tubes with
a neck, which are then sealed. The electrolyte can be
stored indefinitely in sealed tubes and used as required.
Mechanical purification of molten salts is impermissi-
5
0–100°C above the melting point of the eutectic mix-
ture. Next, the cell is sealed off at the neck under vac-
uum and transferred to a preheated working furnace. A
calibrated Pt/Pt–10% Rh thermocouple is introduced
into the casing, which is soldered in the center of the
cell and is level with the electrodes and electrolyte.
Such cells can operate indefinitely between the
solidification temperature of the eutectic melt and the
onset of softening (870 K). In the latter stages of exper-
iments, the temperature can be raised to 900 K.
Synthesis of alloys. The alloys were prepared from
ble. The addition of ammonium chloride to the melt, weighed mixtures (~1.6 g) of high-purity (5N) indium
proposed by Wagner and Werner [3], is inefficient.
and antimony, which were sealed in Pyrex tubes under
–
4
a vacuum of 10 Pa or better. The mixtures were
reacted at 800 K for 48 h in resistance furnaces. The
resultant ingots were up to 5 mm in diameter and
Amount of potential-forming ions in the electro-
lyte. In spite of the large number of studies dealing with
emf measurements in molten salts, there is no agree-
ment as to the content of the salt of the potential-form-
ing ion, which was varied from 0.05 to 5–7 wt % [4, 5].
According to Wagner and Werner [3], the content of the
salt of the potential-forming metal must be 1–3 wt %.
Our experiments indicate that the optimal content of
1
0 mm in length. In each experiment, a whole ingot
was used. The difference in weight between the starting
mixture and ingot was within 0.05%. The compositions
of the alloys studied are listed in Table 1.
Equipment and emf measurements. The tempera-
such a salt is 0.05 to 0.1 wt %. In the case of extremely ture in the resistance furnaces was controlled to an
hygroscopic salts (e.g., ZnCl and AlCl ) or rare-metal accuracy of ±0.2°ë. In emf measurements, we used a
2
3
INORGANIC MATERIALS Vol. 40 No. 5 2004

THERMODYNAMIC PROPERTIES OF ALLOYS AND PHASE EQUILIBRIA
447
E, mV
2
1
1
00
60
20
InSb(s) + Sb(s)
xSb = 0.688
InSb(s) + L1
L1
1
2
3
4
5
6
7
xSb = 0.600
xSb = 0.572
xSb = 0.500
8
0
6
50
700
750
800
850 T, K
Fig. 3. Temperature-dependent emf data obtained in this work for In–Sb alloys with xSb = (1) 0.5520, (2) 0.5717, (3) 0.6002, and
4) 0.5002 in comparison with earlier results: (5) [13], (6) [12], (7) [1].
(
Keithley Model 193 digital voltmeter with an internal ∆H (In) is the partial enthalpy of indium, and ∆S (In) is
1
4
resistance of 10 Ω and accuracy of ±2 µV. The elec- the partial entropy of indium.
trolyte used in concentration cells
The Ö(í) data for single-phase liquids and
+
InSb(s) + Sb(s) mixtures (Fig. 3) were fitted to equa-
(
–) W, In(l) | In in electrolyte | In Sb1 – x(s), W (+)
x
tions of the form [11]
was a eutectic mixture of LiCl and RbCl (46 wt % LiCl,
T = 625 K). The In ions in the cell were formed by
e
+
Ö = ‡ + bT,
reaction (1).
The system was considered to be in equilibrium if
the measured emf E varied by no more than ±5 µV over
three or four measurement cycles performed at 30-min
intervals. The E(T) data obtained during the first week
where ‡ = –∆H (In)/F and b = ∆S (In)/F. The least
squares fitting results are presented in Table 2.
The Ö(í) values for 87 points in the InSb(s) + L1
of measurements were thought of as corresponding to a field at temperatures from 767 to 800 K are listed in
quasi-equilibrium state and were left out of consider- Table 3.
ation. The reproducibility in each series of measure-
To compare the present thermodynamic data for In–
Sb melts with the reference partial functions of mixing
ments (two or three heating cycles) was ±0.7 mV or
better. Measurements with each cell took 2–3 months.
RESULTS AND DISCUSSION
Table 1. Compositions of In–Sb alloys for emf studies
The emf is related to thermodynamic quantities by
xIn
xSb
∆
µ = –FE = RTlna ,
(3)
(4)
(5)
In
In
0
0
.4998
.4480
0.5002
0.5520
0.5717
0.6002
∆
µIn = ∆H (In) – T∆S (In),
∆
H (In) = ∆µ – T (∂∆µ /∂Τ) ,
In
In
p
0.4283
.3998
where F = 96485.31 C/mol, R = 8314.17 J/(mol K),
µIn is the change in the chemical potential of indium,
0
∆
INORGANIC MATERIALS Vol. 40 No. 5 2004

4
48
VASIL’EV
Table 2. E(T) fitting results for In–Sb alloys
2
0
2
2
Phase
region
2
S × 10 ,
mV
(T – T) , Tmin–Tmax
K
,
∑
i
No.
1
xSb
a, mV b × 10 , mV/K
l
T , K
E, mV
2
2
K
–
26.63
InSb + Sb 0.5520 360.82
186
710.29
171.68
52.1
168973
657–766
0
0
.5717
.6002
2
3
4
L1
L1
L1
0.5002
0.5717
0.6002
61.90
64.43
67.46
4.68
7.04
7.83
104
69
815.17
807.48
795.31
100.06
121.30
129.75
88.8
7.2
34601
17890
3079
773–862
762–835
768–819
14
2.0
2
Note: l is the number of data points, and S0 is the mean-square deviation.
Table 3. E(T) data for the InSb(s) + L region
1
T, K
E, mV
T, K
E, mV
T, K
xSb = 0.5717
71.4 153.97
E, mV
T, K
E, mV
xSb = 0.5002
xSb = 0.5520
xSb = 0.6002
770.1
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
74
70
75
78
85
95
75
78
92
97
95
90
80
82
87
92
94
96
98
69
84
89
93
95
93
95
99
152.5
156.0
150.4
146.0
136.9
117.3
150.2
146.2
125.4
111.6
118.2
128.0
144.6
140.6
134.5
125.1
119.2
115.5
108.0
155.8
137.6
128.0
119.5
114.0
122.0
116.8
101.3
778.9
777.0
775.0
773.1
770.7
771.4
769.3
767.5
769.8
771.6
774.4
776.2
778.3
780.4
781.8
784.1
788.2
789.8
141.2
144.25
146.6
154.26
152.10
149.55
147.03
144.72
142.18
139.35
135.97
140.44
143.46
145.75
148.37
151.03
153.33
155.69
139.30
136.04
774
150.97
153.35
149.86
147.5
772.0
774.2
776.1
778.0
779.7
782.2
784.0
780.8
779.1
777.6
775.2
771.5
771.0
768.4
782.0
784.1
772
149.3
774.7
776.7
779.5
781.4
783.3
785.3
783.1
786.8
788.8
769.6
771.5
773.7
775.6
777.7
779.9
782.0
784.0
786.0
787.8
152.46
151.7
144.8
153.85
154.9
142.16
139.07
136.2
153.38
150.92
148.05
145.85
143.2
139.2
133.28
129.49
155.73
153.5
140.5
138.1
151.24
148.75
146.35
143.58
140.92
137.85
134.86
131.69
135.38
128.7
124.75
INORGANIC MATERIALS Vol. 40 No. 5 2004

THERMODYNAMIC PROPERTIES OF ALLOYS AND PHASE EQUILIBRIA
Table 4. Partial thermodynamic functions of formation of In–Sb alloys
449
–
∆G(In),
–∆H(In),
∆S(In),
No.
1
Phase region
xSb
T, K
Method
Source
kJ/mol
kJ/mol
J/(mol K)
InSb(s) + Sb(s)
0.5520
16.83 ± 0.1
700
34.81 ± 0.24
–25.69 ± 0.4 EMF
This work
0
0
.5717
.6002
2
InSb(s) + Sb(s)
0.5990
.6720
17.06 ± 0.1
700
33.34 ± 1.7
–23.25 ± 2.1 EMF
[13]
0
3
4
5
6
7
InSb(s)
700
700
800
800
800
35.4 ± 1.6
35.1 ± 0.3
5.97 ± 0.4
6.22 ± 0.3
6.51 ± 0.3
Calorimetry
[12]
InSb(s) + Sb(s)
17.0 ± 0.3
9.58 ± 0.02
11.65 ± 0.01
12.55 ± 0.01
–25.8 ± 0.3 Calculation
4.52 ± 1.0 EMF
[14]
L1
L1
L1
0.5002
0.5717
0.6002
This work
This work
This work
6.79 ± 0.4 EMF
7.55 ± 0.4 EMF
∆
H (In) and ∆S (In) [14], the latter were represented in seems likely that the optimized data in [1] need cor-
rection.
the form
2
Sb
CONCLUSIONS
∆
µIn = x (2317.8–30577x + 19301x )
Sb
Sb
(
6)
The present results demonstrate that In–Sb alloys
are ideally suited for checking the accuracy of measure-
ments with liquid-electrolyte electrochemical cells and
for use as internal standards in studies of multicompo-
nent In-based systems.
+
9.4T ln(1 – x ),
Sb
with an accuracy of ±100 J/mol in the range 0.1 < x <
.5 and ±250 J/mol in the range 0.6 < x < 0.9.
Sb
0
Sb
The thermodynamic functions of formation of solid
In–Sb samples [12, 14] (Table 4) are given for the In(l)
and Sb(s) states at 700 K. The data necessary for calcu-
lations were taken from [12, 15].
ACKNOWLEDGMENTS
I am grateful to Prof. B. Legendre (Laboratoire de
Chimie Minérale II, Faculté de Pharmacie, Châtenay
Malabry, Paris-Sud, France) for the opportunity to per-
form some of the experimental work in his laboratory.
The enthalpy of fusion of InSb was evaluated using
the present data (Table 4, nos. 1, 5) and the enthalpy of
fusion of antimony reported in [15]. The results are pre-
sented in Table 5.
REFERENCES
In the proper temperature and composition ranges,
the present emf results for In–Sb alloys are in excellent
agreement with the optimized data reported by Ansara
et al. [1] for melts. In the two-phase regions InSb + Sb
1. Ansara, I., Chatillon, C., Lukas, H.L., et al., A Binary
Database for III–V Compounds Semiconductor Sys-
tems, CALPHAD: Comput. Coupling Phase Diagrams
Thermochem., 1994, vol. 18, no. 2, pp. 177–222.
2
. Handbuch der präparativen anorganischen Chemie in
drei Bänden, von Brauer, G., Ed., Stuttgart: Ferdinand
Enke, 1951. Translated under the title Rukovodstvo po
preparativnoi neorganicheskoi khimii, Moscow: Inos-
trannaya Literatura, 1956, pp. 151–156.
and InSb + L , the data differ somewhat. At the same
1
time, the results of earlier calculations by Degtyarev
and Voronin [14] agree well with the present thermody-
namic functions of formation of solid InSb (Table 4). It
3
4
. Wagner, C. and Werner, A., The Role of Displacement
Reactions in the Determination of Activities in Alloys
with the Aid of Galvanic Cells, J. Electrochem. Soc.,
Table 5. Heat of InSb fusion
1
963, vol. 110, no. 4, pp. 326–332.
Source
[15]
[12]
This work
. Yamshchikov, L.F., Lebedev, I.A., Nichkov, I.F., et al.,
Thermodynamic Properties of Liquid Erbium Alloys
with Low-Melting Metals, Materialy vsesoyuznogo
soveshchaniya po termodinamike metallicheskikh sistem
∆mH(InSb),
kJ/mol
48.5 ± 2
49.4 ± 2.5
48.66 ± 0.5
INORGANIC MATERIALS Vol. 40 No. 5 2004

4
50
VASIL’EV
(
Proc. All-Union Conf. on the Thermodynamics of
Indium–Tin, Thermochim. Acta, 1998, vol. 315,
pp. 129–134.
Metallic Systems), Almaty: Nauka, 1979, pp. 181–185.
5
. Somov, A.P., Nikol’skaya, A.V., and Gerasimov, Ya.I.,
Thermodynamic Properties of Higher Lanthanum Tellu-
rides, Izv. Akad. Nauk SSSR, Neorg. Mater., 1973, vol. 9,
no. 4, pp. 575–579.
. Vassiliev, V., Azzaoui, M., and Hertz, J., EMF Study of
Ternary (Pb, Sn, Sb) Liquid Phase, Z. Metallkd., 1995,
vol. 86, pp. 545–551.
. Borzone, G., Parody, N., Ferro, R., et al., Thermody-
namic Investigation of the Lu–Pb System, J. Alloys
Compd., 1995, vol. 220, pp. 111–116.
1
1
1
1. Kornilov, A.N., Matematicheskie metody khimicheskoi
termodinamiki (Mathematical Methods in Chemical
Thermodynamics),
pp. 157−164.
Novosibirsk:
Nauka,
1982,
6
7
8
2. Hultgren, R., Desai, P.R., Hawkins, D.T., et al., Selected
Values of the Thermodynamic Properties of Binary
Alloys, Metals Park: Am. Soc. Met., 1973,
pp. 1024−1027.
3. Nikol’skaya, A.V., Geiderikh, V.A., and Gerasimov, Ya.I.,
Thermodynamic Properties of Indium Antimonide,
Dokl. Akad. Nauk SSSR, 1960, vol. 130, pp. 1074–1077.
. Vassiliev, V., Lelaurain, M., and Hertz, J., A New Pro-
posal for Binary (Sn, Sb) Phase Diagram and Its Ther-
modynamic Properties Based on a New e.m.f. Study, 14. Degtyarev, S.A. and Voronin, G.F., Calculation of Ther-
J. Alloys Compd., 1997, vol. 247, pp. 223–233.
modynamic Properties of Alloys from Calorimetric and
Phase-Diagram Data: II. Indium–Antimony Alloys, Zh.
Fiz. Khim., 1981, vol. 55, no. 5, pp. 1136–1140.
9
. Vassiliev, V., Voronin, G.F., Borzone, G., et al., Thermo-
dynamics of the Pb–Pd System, J. Alloys Compd., 1998,
vol. 269, pp. 123–132.
1
5. Garbato, L. and Ledda, F., Evaluation of Heats and
Entropies of Fusion by Quantitative Thermal Analysis
Method, Thermochim. Acta, 1977, vol. 19, pp. 267–273.
1
0. Vassiliev, V., Feutelais, Y., Sghaier, M., and Legendre, B.,
Liquid State Electrochemical Study of the System
INORGANIC MATERIALS Vol. 40 No. 5 2004