Thermogravimetric study of GaAs chlorination between -30 and 900 °c

Article abstract of DOI:10.1016/j.tca.2011.05.012

Gallium (as GaAs) is at present an essential part of electronic devices, and the recovery of this element from electronic wastes is fundamental for the metallurgic industry. In this work, with the aim of recovering Ga by chlorination, the following reacti

Full text of DOI:10.1016/j.tca.2011.05.012

Thermochimica Acta 523 (2011) 124–136
Contents lists available at ScienceDirect
Thermochimica Acta
journal homepage: www.elsevier.com/locate/tca
Thermogravimetric study of GaAs chlorination between −30 and 900 ◦C
Fernando M. Tuneza,b,∗, Jorge A. Gonzaleza,b, María del C. Ruiza
a Instituto de Ciencias Básicas, Universidad Nacional de Cuyo, 5500 Mendoza, Argentina
b Instituto de Investigaciones en Tecnología Química (INTEQUI-CONICET), Universidad Nacional de San Luis, Argentina
a r t i c l e i n f o
a b s t r a c t
Article history:
Gallium (as GaAs) is at present an essential part of electronic devices, and the recovery of this element from
electronic wastes is fundamental for the metallurgic industry. In this work, with the aim of recovering
Ga by chlorination, the following reaction was investigated:
Received 30 December 2010
Received in revised form 8 May 2011
Accepted 11 May 2011
Available online 19 May 2011
GaAs(s) + 3Cl2(g) → GaCl3(s, g, l) + AsCl3(s, g, l)
A thermogravimetric study of the previous reaction was carried out, and the following variables were
investigated: chlorine partial pressure, in the range between 0.2 and 1 atm, by dilution with N2 and
reaction temperature in the −29 ◦C to 900 ◦C interval. The thermodynamic calculations predicted that the
reaction was feasible during the entire analyzed temperature range, and the experimental results were
in agreement with the thermodynamic estimations. The results showed that the chlorination rate did not
change significantly with temperature, but it increased with the chlorine partial pressure. It was found
that AsCl3 was partially evaporated at temperatures above 20 ◦C, while the GaCl3 started evaporating at
100 ◦C. The experimental observations showed that the mechanism originating the reaction was different
in each temperature range ranging between −30 and 160 ◦C, 200–400 ◦C, and above 500 ◦C. In the first
case the reaction was caused by adsorption of Cl2 on the solid surface. In the second case, the solid
started reacting when little quantities of GaCl3 and AsCl3 vapor were injected in the entrance of the
system, indicating that the chlorination kinetics responded to an autocatalytic mechanism. Above 500 ◦C,
the reaction was caused without adding chlorination products. The results of the analysis by scanning
electron microscopy and X-ray diffraction showed a preferential attack on one of the crystal planes of
GaAs. Besides, the formation of Ga2O3 and GaAsO4 was observed as a consequence of the presence of
small quantities of oxygen accompanying Cl2.
Keywords:
Gallium
Arsenic
Recovery
Chlorination
Thermogravimetry
The results showed that chlorination is a selective and economic methodology for the recovery of
gallium from electronic wastes.
© 2011 Elsevier B.V. All rights reserved.
1. Introduction
sistors, lasers, conductors coating), vacuum technology, and in the
chemical industry for the catalyzers manufacturing [1].
Gallium is a rare element with a natural abundance of 16 ppm.
Although this mineral is more concentrated in gallite, it is mostly
found in aluminosilicates such as bauxite, clays and zinc ores
for example, sphalerite). Gallium is mainly extracted from wastes
obtained during aluminium processing, and also from electrolytic
zinc condensers. Other extraction sources include ashes released
from coal burning [1].
Gallium metal, alloys and compounds are used in a relatively
reduced scale in industries such as the construction of machines,
the production of electrical and electronic equipment (diodes, tran-
Gallium has important applications in optoelectronics (LED’s),
telecommunications, aerospace technology, and many commercial
and domestic devices such as computers and DVD’s [1].
Gallium has found application in the Gallex Detector Experi-
ment located in the Gran Sasso Underground Laboratory in Italy.
In this experiment, 30.3 tons of gallium in the form of 110 tons of
GaCl3–HCl solution are being used to detect solar neutrinos [2].
When gallium is used for glass painting, it forms a brilliant mir-
ror and also wets the crystal and porcelain. The element adapts
easily to form alloys of low fusion point (for example, eutectic
alloys) with most metals.
Gallium arsenide (GaAs) is a valuable compound widely used
in the manufacturing of advanced semiconductors for microwave
transceivers, transistors, solar cells, lasers, laser diodes, compact
discs and other electronic applications. This compound has several
advantages in relation to other semiconductor materials, such as
∗
Corresponding author at: Instituto de Investigaciones en Tecnología Química
(INTEQUI), Universidad Nacional de San Luis - CONICET, CC: 290, 5700 San Luis,
Argentina. Tel.: +54 2652423789x116; fax: +54 02652426711.
E-mail address: fmtunez@unsl.edu.ar (F.M. Tunez).
0040-6031/$ – see front matter © 2011 Elsevier B.V. All rights reserved.
doi:10.1016/j.tca.2011.05.012

F.M. Tunez et al. / Thermochimica Acta 523 (2011) 124–136
125
Fig. 1. Change of Gibbs free energy versus temperature for the GaAs chlorination reaction.
fast operation with low energy consumption, good radiation resis-
tance and capacity to convert electrical signals into optical ones
[1,3].
The manufacturing of electrical and electronic devices has
rapidly developed, giving place to an increase in the production
of wastes from these devices. The recycling of these wastes is an
important issue not only because their reduction is necessary but
also because valuable materials can be recovered. Besides, these
kinds of wastes constitute a concern for the governments and also
for the population due to their dangerous materials [4].
Electronic wastes are of diverse complexity due to their differ-
ent materials and components and to the different processes used
for the equipment manufacturing. The characterization of these
wastes is of key importance for the development of a recycling sys-
tem with a good cost-benefit relation between the system and the
environment. The selective dismantling used to identify dangerous
substances and/or valuable materials is an essential process in the
recycling of these wastes [5].
mechanism involves adsorption of Cl2 on the surface of GaAs
and subsequent desorption of GaCl. Hung et al. [10] studied the
chlorination of GaAs at low temperatures (110 K) using soft X-ray
photoelectron spectroscopy and found that at low exposures of
Cl2 it dissociates itself and preferentially adsorbs on the atoms
of As inducing the breaking of the bond Ga–As, generating sub-
sequently gallium and arsenic chlorides. Bond et al. [11] using
pulsed supersonic molecular beam scattering studied the reaction
of Cl2 with the surface of GaAs, and they observed that the reaction
products depended on the flow of chlorine and temperature; at
low flows of Cl2 and high temperature, the formation of GaCl and
As predominates, and at high flows of Cl2 and low temperature,
there is a predominance of GaCl3 and AsCl3 formation.
This paper presents the thermogravimetric study of GaAs chlo-
rination with Cl2 between −30 and 900 ◦C. The partial pressure of
Cl2 was varied between 0.2 and 1 atm. The purpose of the study is
to provide data to facilitate elucidating the mechanism and kinetics
of the reaction in working conditions so that these results can then
be applied to the recovery of Ga.
Progress in the development of materials with application in
different fields, ranging from medicine to aerospace industries, has
made the production of high purity metals necessary. Several meth-
ods exist to produce rare and precious metals with a high purity
degree; however, liquid–liquid extraction from metallic halides
in solution and chlorides distillation are the most widely applied
methodologies nowadays. In both cases, obtaining the metallic
halides is the first stage of the process, and one of the most widely
used techniques for this stage is the previous chlorination of mate-
rials such as metals, ores, metallic oxides or industrial wastes [6].
The use of chlorination in the metal extraction processes by
means of pyrometallurgical and hydrometallurgical methods has
received great attention in recent decades, and an increasing use
of the chlorine chemistry is expected due to the advantages of this
process [7]. According to Kubo et al. [8], the main advantages of the
chlorination process in relation to other refinement methods for
gallium recovery from GaAs can be summarized as follows: opera-
tion temperatures are low, most of the impurities can be effectively
eliminated, high-purity metallic gallium can be obtained and the
productivity increase is significant.
2. Thermodynamic analysis
This section presents the thermodynamic calculations for the
GaAs–Cl2 system, which have been done for the temperature range
used for the experimental essays. All the calculations were obtained
using the HSC Chemistry for Windows, version 5.1 software [18].
The reactions that may occur between GaAs and Cl2 have been
theoretically and experimentally studied by several authors [9–17]
under different temperature conditions, partial pressure and Cl2
quantity in order to obtain components for electronic applications.
The results of these investigations showed that the reaction prod-
ucts vary in their composition as well as in their aggregation state
according to the working conditions.
A first analysis involves to study the feasibility of a gallium and
arsenic chlorides formation from GaAs. Fig. 1 shows the results of
the calculation of the Gibbs free energy variation for the following
global reaction:
GaAs + 3Cl2(g) → GaCl3(s, l, g) + AsCl3(s, l, g)
The analysis of Fig. 1 permits to conclude that the reaction
between GaAs and Cl2 is feasible at all the studied temperature
(1)
Several authors have reported the mechanism of the chlorina-
tion reaction of GaAs. Jenichen and Engler [9], using the functional
density method, theoretically found that the most feasible reaction

126
F.M. Tunez et al. / Thermochimica Acta 523 (2011) 124–136
kmol
1.1
1.0
85ºC
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
Cl2(g)
AsCl3(l)
GaCl3
GaAs
0.9
GaCl3(g)
AsCl3(g)
0.8
0.7
0.6
0.5
0.4
0.3
0.2
As
Cl2(g)
GaCl3(s)
GaCl3(l)
AsCl3(l)
Ga2Cl6(g)
0.1
Ga2Cl6(g)
3.5 4.0
AsCl3(g)
2.5 3.0
Cl(g)
As(l)
1.0
0.0
0.0
0.5
1.5
2.0
-100
100
300
500
700
900
kmol Cl2 (g)
Temperature (ºC)
Fig. 3. Effect of Cl2 quantity at 85 ◦C.
Fig. 2. Equilibrium composition in pure chlorine.
The formation of small amounts of oxygenated compounds,
such as Ga2O3 and GaAsO4 was observed (see Section 4.2.1) in
the experimental phase when working at a temperature above
300 ◦C. In order to investigate this phenomenon, the analysis of
the compounds thermodynamic feasibility to be formed from the
chlorination products of the gallium arsenide was carried out.
The results shown in Fig. 5a indicate that the reactions of the
compounds formation are thermodynamically feasible at almost all
the studied temperature range, except for the reaction of the Ga2O3
formation which is spontaneous up to 810 ◦C. Also, the predomi-
nance diagrams (pO2 versus pCl2) presented in Fig. 5b and c for Ga
and As, respectively, show that Ga and As oxygenated compounds
predominate over those chlorinated at 25 ◦C.
range, showing greater spontaneity as temperature decreases. The
changes in the graph slope indicate phase changes of the reaction
products. It is important to emphasize that between 150 and 360 ◦C
the formation of the gaseous gallium trichloride is more feasible as
dimmer Ga2Cl6.
The equilibrium compositions at different temperatures and Cl2
amounts was used to present the thermodynamic analysis for the
reaction between GaAs and Cl2.
The equilibrium compositions were obtained assuming that
from the elements constituting the reagents used, all the species
available in the software database may be formed: As(g),
As2(g), As3(g), As4(g), AsCl3(g), Cl(g), Cl2(g), Cl3(g), Cl4(g), Ga(g),
Ga2(g), GaAs(g), GaCl(g), GaCl2(g), GaCl3(g), Ga2Cl2(g), Ga2Cl4(g),
Ga2Cl6(g), AsCl3(l), Ga(l), As(l), GaCl3(l), As(A) (amorphous arsenic),
As(Y) (yellow arsenic, cubic), Ga(s), As(s), GaAs(s), GaCl3(s).
Fig. 2 shows the equilibrium composition for the GaAs–Cl2 react-
ing system in the temperature range ranging between −30 and
1000 ◦C obtained considering a total pressure of the system of
1 atm. The initial reagents were GaAs(s) and Cl2(g). The quantity of
Cl2(g) was higher than the stoichiometric. Since the experimental
assays were carried out with excess Cl2, the estimates were per-
formed with excess Cl2 in relation to the reaction using the largest
quantity of Cl2 (reaction (1)). The composition was in the same fig-
ure it can be observed that in all the analyzed temperature range,
the reaction products are gallium trichloride and arsenic trichlo-
ride which are formed according to reaction (1), with different
aggregation states depending on the temperature.
Fig. 3 shows the equilibrium composition in relation to Cl2 quan-
tity. This diagram was calculated at a total pressure of 1 atm, with
pure Cl2 at 85 ◦C, starting with one mol of GaAs and increasing
the quantity of Cl2. It can be observed that as GaAs(s) is consumed
and GaCl3(l) and As(s) are produced; when all the GaAs has been
consumed, the formed As starts reacting with the Cl2 producing
AsCl3(l).
While Fig. 2 provides a thermodynamic estimation of the species
that may exist in equilibrium at different temperatures when GaAs
reacts with Cl2 in excess, Fig. 3 shows the thermodynamic evolution
of the reaction in relation to the amount of the present chlorine,
which might be useful to elucidate the most probable reaction
mechanism.
3. Experimental
3.1. Materials
The materials used were GaAs Sigma–Aldrich with a purity of
99.999%, gaseous chlorine INDUPA Argentina with a purity of 99.5%
and N2, as gas for dilution and purge, AGA ARGENTINA with a purity
of 99.9%.
1.00
0.90
0.80
0.70
AsCl3(g)
GaCl3(g)
0.60
0.50
0.40
0.30
0.20
0.10
0.00
Ga2Cl6(g)
Fig. 4 shows the partial pressure variation in relation to tem-
perature for the three products that may be formed by the reaction
between GaAs and Cl2. AsCl3 vapor pressure reaches 10−3 atm at
58 ◦C, where as Ga2Cl6 reaches this pressure at 118 ◦C, and GaCl3,
at 125 ◦C.
0
100
200
300
400
500
Temperature (ºC)
Fig. 4. Partial pressure of equilibrium versus temperature for AsCl3, Ga2Cl6 and
GaCl3.

F.M. Tunez et al. / Thermochimica Acta 523 (2011) 124–136
127
Fig. 5. (a) Change of Gibbs free energy in relation to temperature for the reactions which include the formation of oxygenated compounds. (b) Predominance diagrams (pO2
versus pCl2) for Ga. (c) Predominance diagrams (pO2 versus pCl2) for As.

128
F.M. Tunez et al. / Thermochimica Acta 523 (2011) 124–136
200
150
100
50
-29 ºC
30 ºC
-10 ºC
50 ºC
10 ºC
85 ºC
110 ºC
0
-50
-100
-150
160 ºC
-29 ºC
0
2
4
6
8
10
12
14
16
18
Time (min)
Fig. 6. Isothermal chlorination of GaAs with pure Cl2, in the temperature range between −30 and 200 ◦C.
3.2. Experimental equipment and procedure
times were standardized in an N2 environment. After that, the cru-
cible with the chlorination wastes was removed, and the wastes
were prepared for analysis. The effluent gases from the reactor were
treated by H2SO4 and NaOH traps.
The monitoring of the chlorination reaction was performed
using thermogravimetry. An experimental equipment capable of
working in corrosives and no corrosives atmospheres, developed
in our laboratory [19], was used for this purpose.
The analyses of the chlorination wastes were carried out using
X-ray diffractometry (XRD) in a Rigaku equipment model D-Max
III C; X-ray fluorescence (XRF) using a Philips equipment model
PW 1400; scanning electron microscopy (SEM) and electron probe
microanalysis (EPMA) using a LEO 1450VP microscope and a cou-
pled EDAX Gemini 2000 energy dispersive spectrometer (EDS) and
atomic absorption spectrometry (AAS) using a Shimadzu 6800 AA
equipment.
The experimental procedure for the thermogravimetric study
of GaAs chlorination was as follows: the sample, previously ground
and classified to a particle size of −40 +90 mesh, was placed in a pre-
viously weighed quartz crucible, then the crucible with the sample
was supported in a thermobalance. In each experiment a mass of
solid reagent of approximately 50 mg was used.
After being placed in the thermogravimetric system, the sample
was thermostatized for 30 min in an N2 current of 50 ml/min. Once
the working conditions were stabilized, the N2 flow was turned off
and a Cl2 current of 50 ml/min was introduced in the thermogravi-
metric system at the corresponding partial pressure by dilution
with N2; time was measured from this moment, and data of mass
change were collected. When the experiment was finished the Cl2
gas flow was turned off and the thermogravimetric system was
purged with N2. In all the essays the thermostatization and purge
4. Results and discussion
Isothermal essays at different temperatures and partial pres-
sures of chlorine were carried out until complete conversion of
GaAs in order to study the influence of both variables on the chlo-
rination reaction.
The thermogravimetric analysis of the GaAs–Cl2 system is pre-
sented separately for temperature ranges lower and higher than
200 ◦C.
4.1. Thermogravimetric analysis between −30 and 200 ◦C
4.1.1. Temperature effect
The effect of temperature on the reaction between GaAs and
pure Cl2(g) was analyzed from thermogravimetric measurements.
Fig. 6 shows the results of the time evolution of relative mass
changes. There is an induction time of approximately 1 min at all
the studied temperatures whose origin will be discussed in Section
4.3. The maximum relative mass changes obtained from Fig. 6 are
shown in Table 1 close to the expected values for the temperatures
at which aggregation states of reaction products can be assumed.
For other temperatures the mass change contribution of vaporiza-
tion process of reaction products) cannot be theoretically predicted
Table 1
Mass percentage changes expected in relation to the aggregation state of the products.
Reaction
Maximum theoretical
mass change (5%)
Temperature
(◦C)
Maximum experimental
mass change (%)
x
GaAs + 3Cl2(g) → GaCl3(s) + AsCl3(l)
GaAs + 3Cl2(g) → GaCl3(l) + AsCl3(l)
GaAs + 3Cl2(g) → GaCl3(l) + (1 − x)AsCl3(l) + xAsCl3(g)
GaAs + 3Cl2(g) → GaCl3(l) + (1 − x)AsCl3(l) + xAsCl3(g)
GaAs + 3Cl2(g) → GaCl3(l) + (1 − x)AsCl3(l) + xAsCl3(g)
GaAs + 3Cl2(g) → GaCl3(l) + (1 − x)AsCl3(l) + xAsCl3(g)
(1)
(2)
148.15
148.15
−29
−10
10
154.59
144.81
139.58
118.23
101.89
78.02
0.06
0.23
0.36
0.55
30
50
85
GaAs + 3Cl2(g) → GaCl3(l) + AsCl3(g)
GaAs + 3Cl2(g) → (1 − x)GaCl3(l) + xGaCl3(g) + AsCl3(g)
GaAs + 3Cl2(g) → GaCl3(g) + AsCl3(g)
(3)
21.73
110
160
200
21.96
−82.32
−94.55
0.88
(4)
−100

F.M. Tunez et al. / Thermochimica Acta 523 (2011) 124–136
129
80
70
60
50
40
30
20
10
0
As
Ga
-10
-30
-10
10
30
50
70
90
110
Temperature (ºC)
Fig. 7. Ga and As loss in the chlorination essays at different temperatures.
but empirically quantified through the calculation of the stoichio-
metric coefficient x, as shown in last column of Table 1. Also, at
low temperatures, absorption of Cl2 (b.p. = −34.04 ◦C, [2]) in liquid
chlorides must be taken into account. Finally, chlorides vaporiza-
tion could be enhanced by a local temperature rising due to reaction
exothermic effect. Therefore, chemical equations (1), (2) accounts
well for reaction at −29 ◦C and −10 ◦C, respectively, and reaction at
10 ◦C deviates slightly from mass changes predicted by equation (2)
indicating a low contribution (x ≈ 0) of AsCl3(l) vaporization. As the
temperature rise from 30 ◦C, the increasing x values indicates the
enhancing contribution of AsCl3(l) vaporization. In fact, for experi-
ments carried out at 50 ◦C and 85 ◦C (Fig. 6), once reaction finished,
a continuous mass decreasing corresponds to AsCl3(l) vaporization
as unique phenomenon.
ative mass loss. Fig. 7 quantitatively agree with the increasing
AsCl3(l) vaporization at temperatures greater than 10 ◦C, and partial
GaCl3(l) vaporization for the highest temperatures, as determined
from thermogravimetric measurements.
4.1.2. Effect of the partial pressure of Cl2
Figs. 8–10 the results of the thermogravimetric essays carried
out to investigate the effect of the Chlorine partial pressure, pCl2,
(using N2 as inner diluent gas) at −20, 20 and 85 ◦C, respectively. For
all tested temperatures, the greater pCl2, the greater the reaction
rate (with the only exception shown in Fig. 10 for the two highest
pCl2). However, a detailed analysis is not easy because chlorination
reaction is overlapped with chlorides evaporation. Therefore, any
reaction order calculation, for example, will be falsified by other
processes. Moreover, observed maximum relative mass changes
indicates that contribution of chlorides vaporization seems to be
dependent, not only on temperature (x value in Table 1) but also
on pCl2. The net effect for lower pCl2 is some combination of N2
dilution, that enhances the heat dissipation and reduces the men-
tioned exothermic effect on chlorides vaporization, and a lowering
in chlorination rate enhances the relative contribution of chlorides
vaporization. This complexity is also confirmed by the (unexpected)
Total mass change after chlorination at 110 ◦C is accounted for
total vaporization of AsCl3 (chemical equation (3)) which, despite
temperature is lower to its boiling point (130 ◦C [2]), can be
explained from the exothermic effect above mentioned. The exper-
iments at highest temperatures are interpreted taking into account
GaCl3(g) or Ga2Cl6(g) formation which is partial at 160 ◦C and is
nearly total at 200 ◦C.
Also reaction residues where quantified by atomic absorption
measurements and results are shown in Fig. 7 as Ga and As rel-
160
0.8 atm
-20 ºC
140
120
0.6 atm
0.4 atm
1 atm
100
80
60
40
20
0
0.2 atm
0
5
10
15
20
25
30
35
Time (min)
Fig. 8. Influence of Cl2 partial pressure of at −20 ◦C.

130
F.M. Tunez et al. / Thermochimica Acta 523 (2011) 124–136
160
140
120
100
80
20ºC
0.8 atm
0.6 atm
1 atm
0.2 atm
0.4 atm
60
40
20
0
0
5
10
15
20
25
30
Time (min)
Fig. 9. Influence of Cl2 partial pressure at 20 ◦C.
inverse dependence of reaction rate process with pCl2 observed at
85 ◦C as above mentioned from Fig. 10.
ture and, excepting for the two highest temperatures, solid residues
remain (see Section 4.2.3). After chlorinations, solid residues of
white color were observed deposited, not also in crucible, but also
in reactor walls.
4.2. Thermogravimetric analysis between 300 and 900 ◦C
Fig. 11 also shows that solid residues remain, except in the
essays performed between 800 and 900 ◦C. It was observed in the
experimental study that the residues were deposited in the crucible
and also in the reactor walls. The residues observed in this figure,
according to the analyses carried out using SEM, EPMA and XRD,
correspond to compounds containing Ga, O and As. These data indi-
cate that the GaAs chlorination was complete, and consequently
the masses found at the end of the reaction were due only to these
oxygenated compounds. The formation of these residues, whose
origin, composition and structure will be discussed in detail in Sec-
tion 4.2.3, might be due to the reactions between the products of the
GaAs reaction and small amounts of O2 present in the Cl2 (Fig. 5).
4.2.1. Temperature effect
Chlorination performed at this temperature range exhibits a
particular behavior. For the tested temperatures below 600 ◦C reac-
tion does not start unless small amounts of a GaCl3(g)–AsCl3(g)
mixture is added. Therefore, for the experiments at 300 ◦C and
500 ◦C, those chlorides were injected in Cl2 flow and supply was
finished once reaction was started.
Fig. 11 shows the results of the thermogravimetric study of the
GaAs chlorination with pure chlorine at temperatures between 300
and 900 ◦C. According to the thermodynamic predictions analyzed
in Section 2, all the products of the GaAs chlorination are in gaseous
state in this temperature range, and in a flow system, they may be
carried away by the gaseous current and then condensed outside
the reaction zone. Consequently, the expected relative mass loss
might be 100%.
4.2.2. Effect of the Cl2 partial pressure
Figs. 12 and 13 show the effect of pCl2 on reaction rate at 300 ◦C
and 500 ◦C, respectively. The effect of pCl2 on reaction rate is less
prominent as pCl2 decreases. In all cases, final relative mass change
indicates variable amounts of solid residues. At both temperatures,
Fig. 11 shows that chlorination rate does not practically depend
on temperature; induction time changes inversely with tempera-
100
0.8 atm
85 ºC
90
0.6 atm
80
1 atm
70
0.2 atm
0.4 atm
60
50
40
30
20
10
0
0
5
10
15
20
25
30
35
Time (min)
Fig. 10. Influence of Cl2 partial pressure at 85 ◦C.

F.M. Tunez et al. / Thermochimica Acta 523 (2011) 124–136
131
pCl2 = 1 atm
0
-20
300ºC
350ºC
500ºC
400ºC
700ºC
600ºC
-40
-60
-80
800ºC
900ºC
-100
-120
0
2
4
6
8
10
12
14
16
Time (min)
Fig. 11. GaAs isothermal chlorination with pure Cl2, between 300 and 900 ◦C.
300
C
0
-20
0.2 atm
-40
-60
0.5 atm
-80
1 atm
-100
-120
0
10
20
30
40
50
60
Time (min)
Fig. 12. Influence of Cl2 partial pressure at 300 ◦C.
these solid residues were black or white, respectively, at lower or
higher pCl2 (see Section 4.2.3).
Fig. 14 shows the X-ray diffractograms of residues after reac-
tion with pure Cl2 at 300 ◦C, 350 ◦C and 400 ◦C and it can be clearly
observed that solids in all cases are amorphous. The SEM images
and EPMA analyses on these solids were all similar. Fig. 15 shows
results corresponding to residue obtained at 350 ◦C. Solid particles
exhibits irregular surfaces and are composed by As, Ga and O. This
last result is consistent with the above mentioned white color of
4.2.3. Characterization of the solid residues
Solid residues of chlorinations performed at temperatures
between 300 ◦C and 900 ◦C were analyzed by XRD, SEM and EPMA.
500 o
C
0
-20
0.7 atm
-40
-60
0.5 atm
0.2 atm
1 atm
-80
-100
-120
0
5
10
15
20
25
30
35
40
Time (min)
Fig. 13. Influence of Cl2 partial pressure at 500 ◦C.

132
F.M. Tunez et al. / Thermochimica Acta 523 (2011) 124–136
100
80
60
40
20
(a) 300 ºC
10
20
30
40
50
60
70
2θ (º)
110
100
90
80
70
60
50
40
30
20
(b) 350 ºC
10
20
30
40
50
60
70
2θ (º)
110
100
90
80
70
60
50
40
30
20
(c)
400 ºC
10
20
30
40
50
60
70
2θ (º)
Fig. 14. Diffractograms of the residues from the essays with pure Cl2. (a) 300 ◦C; (b) 350 ◦C; (c) 400 ◦C.
residues observed at high pCl2 if it is assumed residues formed by
As/Ga oxides.
Finally, the residues composition dependence on pCl2 (some
combination of GaAs and As–Ga–O compounds) explain the vari-
able limit for mass loss for experiments carried out with low pCl2
(Figs. 12 and 13) and the incomplete products vaporization for
experiments with pure Cl2 (Fig. 11) due to formation of non reactive
As–Ga–O compounds after complete GaAs chlorination.
Fig. 16 shows the X-ray diffractograms of residues after reaction
with pure Cl2 at 500 ◦C and 700 ◦C. residues are clearly crystalline
solids. Solid obtained at 500 ◦C is a mixture of Ga2O3 [21,22] and
GaAsO4 [23]. At 700 ◦C an additional, and unknown, crystalline
phase also appears.
The formation of oxygenated species during chlorinations is
explained from O2 traces presence in commercial Cl2, originated
at its industrial production. The appearance of these oxides, inside
and outside the crucible, likely indicates they are originated by
the oxidation of the GaCl3(g) and AsCl3(g) which are feasible reac-
tions according to Fig. 5. This assumption is also supported by
the fact GaAs does not react with air below 800 ◦C (results not
shown).
Fig. 17 shows X-ray diffractograms of residues obtained after
chlorinations for pCl2 = 0.2 atm at 300 ◦C and 500 ◦C. XRD peaks
allow to identify both solids as GaAs [24], which explain the above
mentioned black color of residues observed at low pCl2. The diffrac-
togram of the residue obtained at 500 ◦C shows a high preferred
orientation effect of planes (2 2 0). Likely Cl2 through a selective
attack promotes changes in crystallites enhancing morphological
importance of those planes. Fig. 18 shows the SEM observations of
residue obtained at 300 ◦C compared with starting GaAs confirm a
localized reactivity on solid particles.
4.3. Reaction mechanism
GaAs chlorination shows two main behaviors depending on
temperature. As mentioned, chlorination were performed in a sim-
ple experimental procedure at all temperatures, except for the
range between 300 ◦C and 500 ◦C. For this temperatures the reac-
tion does not start when Cl2 reaches GaAs, unless small amounts of
a mixture of reaction products (GaCl3(g)–AsCl3(g)) are added. This
last feature corresponds to an autocatalytic behavior.
Any mechanism proposal must explain autocatalytic and non
autocatalytic behaviors at each temperature range, and the
observed induction time variation with experimental conditions.
Each temperature range will be separately discussed.
4.3.1. Low temperature range (−30 ◦C and 200 ◦C). Non catalytic
behavior
A first step proposal in chlorination process could be supported
on experimental [11,14] and theoretical [15] studies which con-
clude the GaAs(s)–Cl2(g) interaction, at room temperature, yields
the formation of GaCl(g) and As2(g) or As4(g). Therefore, Ga is firstly
The absence of oxygenated compounds is consistent with the
low O2 content for pCl2 = 0.2 atm.

F.M. Tunez et al. / Thermochimica Acta 523 (2011) 124–136
133
Fig. 15. SEM micrographs of the residue from the essay carried out at 350 ◦C with pure Cl2. (a) Residue particle; (b) enlarged surface of the particle; (c) residue analysis of
the essay carried out at 350 ◦C with pure Cl2.
Sample
Ga2O3
GaAsO4
100
a
80
60
40
20
0
10
20
30
40
50
60
70
2θ (º)
Sample
Unidentified phase
100
80
60
40
20
b
Ga2O3
GaAsO4
0
10
20
30
40
50
60
70
2θ (º)
Fig. 16. Diffractograms of the residues from the essays carried out with pure Cl2. (a) 500 ◦C; (b) 700 ◦C.

134
F.M. Tunez et al. / Thermochimica Acta 523 (2011) 124–136
Diffractogram
GaAs
100
80
60
40
20
0
a
10
20
30
40
50
60
70
2θ (º)
Diffractogram
GaAs
100
80
60
40
20
0
b
10
20
30
40
50
60
70
2θ (º)
Fig. 17. Diffractograms of the residues from the essays with a Cl2 partial pressure of 0.2 atm. (a) 300 ◦C; (b) 500 ◦C.
chlorinated instead As is. Then initiation includes adsorption of Cl2
on As atoms of GaAs. Adding possible propagation steps, a mecha-
nism could be expressed as:
a. The addition of reaction products must generate the reaction ini-
tiation. Singh and Krout [20] found that AsCl3 reacts with GaAs
forming GaCl(g), GaCl2(g) and As(g) at this temperature range,
therefore the following steps, can be added:
Stage 1 (initiation)
GaAs + AsCl3(g) ꢀ AsCl2(ads) + Cl(ads) + GaAs
(h)
Cl Cl
Cl
Cl
1
Ga Ga
As As
Ga
Ga
2 GaCl(g) + As2 or /2As4(g)
AsCl2(ads) + Cl(ads) + GaAs → AsCl2(g)
+ GaCl(g) + As2(ads) + GaAs
As As
(i)
(j)
(a)
As2(ads) → As2(g) or (1/2)As4(g)
Stage 2 (propagation)
where Cl radicals are directly formed the initiation step, h. This
accounts for decreasing induction time with increasing temper-
ature (Fig. 11) at 300–500 ◦C range due to acceleration of Cl
radicals concentration. Besides, the increase of induction time
with decrease of pCl2 (Figs. 12 and 13) suggests the simultane-
ous participation of step d whose rate will proportional to Cl2(g)
concentration. Therefore, once reaction starts catalyzed by gaseous
products supply, also propagation steps associated the non auto-
catalytic regime also operate.
GaCl(g) + Cl2(g) → GaCl2(g) + Cl(g) or GaCl3(l, g)
2GaAs + 2Cl(g) → GaCl(g) + As2 or (1/2)As4(g)
GaCl2(g) + Cl2(g) → GaCl3(l, g) + Cl(g)
As2 o (1/2)As4(g) + 2Cl(g) → 2AsCl(g)
AsCl(g) + Cl2(g) → AsCl2(g) + Cl(g) or AsCl3(l, g)
AsCl2(g) + Cl2(g) → AsCl3(l, g) + Cl(g)
(b)
(c)
(d)
(e)
(f)
(g)
4.3.3. High temperature range (600 ◦C and 900 ◦C). Non catalytic
behavior
Assuming control by stage 1, the decrease of induction time as
temperature increases (Fig. 6) can be explained based on an increas-
ing rate on GaCl(g) formation, due to the break of Ga–As is favored,
which faster increase Cl radicals concentration through the first
propagation step, b.
Finally, reaction is sustained due to formation of the very active
Cl(g) radicals at several of propagation steps, two of them (d, g) also
accounts for GaCl3 and AsCl3 products formation.
For this temperature range again direct contact between GaAs
with Cl2(g) is enough to produce chlorination without needing to
add reaction products. However for this temperature range step a
cannot be considered as for autocatalytic regime. For highest tem-
perature dissociation of Cl2(g) can occur. Fig. 19 shows a calculation
of amounts of equilibrium species in a Cl2(g)–N2(g) mixture at sev-
eral temperatures. This diagram shows that the concentration of
atomic chlorine Cl(g) is almost null at low temperatures, and as
the temperature increases, its molar fraction also increases, and
above 500 ◦C, it is higher than 10−5. This concentration could be
enough to make important step b as a first propagation one. The
4.3.2. Medium temperature range (300 ◦C and 500 ◦C). Catalytic
behavior
As temperature increases Cl2 adsorption on GaAs could be
enoughly unfavorable to avoid the effective participation of step

F.M. Tunez et al. / Thermochimica Acta 523 (2011) 124–136
135
Fig. 18. SEM micrographs of GaAs particles partially attacked at 300 ◦C and a Cl2 partial pressure of 0.2 atm. (a) Original particle of GaAs; (b) residue particles; (c) enlarged
zone of one of the particles; (d) surface of an attacked particle.
5
0
the system was exposed. The variation of Gibbs standard free
energy versus temperature between −30 and 1000 ◦C shows, for all
the proposed reactions that they are thermodynamically feasible in
the established direction.
Cl2(g)
N2(g)
-5
GaAs chlorination with Cl2 is quite favorable in the tempera-
ture range between −30 and 900 ◦C, and it occurs at a high reaction
rate, even at low temperatures. The system, at the studied tem-
perature range, presents three intervals according to the way the
reaction starts. Between −30 and 200 ◦C, the reaction starts by chlo-
rine adsorption on the GaAs surface and the subsequent breakdown
of the bond between Ga and As. Between 300 and 500 ◦C, the behav-
ior is autocatalytic, requiring the presence of a reaction product
such as AsCl3 for the reaction to begin. Above 500 ◦C, there is a direct
reaction between Cl2 and GaAs due to the formation of atomic chlo-
rine. The reaction temperature has little effect on the chlorination
rate although it influences notably the induction times and deter-
mines if the reactions products volatilize or remain as solid or liquid
residues into the crucible.
Cl(g)
-10
-15
-20
-25
-30
-35
Cl3(g)
Cl4(g)
0
200
400
600
800
1000
Temperature (ºC)
Fig. 19. Equilibrium diagram of the chlorine species at different temperatures and
a Cl2 pressure of 0.9 atm.
In the temperature range between −30 and 200 ◦C, the reaction
is completed in approximately 8 or 9 min. Between −30 and 110 ◦C,
the reaction causes a mass gain, whose magnitude varies according
to the working temperature. These mass gains are due to different
factors: adsorption of Cl2 on the sample at −30 ◦C; at a higher
temperature solid and/or liquid products are formed, and a partial
evaporation of some species occurs, which is produced by the
exothermic effect of the GaAs chlorination reaction. The Cl2 partial
pressure for the different studied temperatures has a marked
decreasing on induction time observed for increasing temperature
for highest range (Fig. 11) can be explained from Fig. 19 because of
concentration of Cl radicals increases with temperature.
5. Summary
GaAs chlorination with Cl2 is feasible under several conditions,
forming different products according to the environment to which

136
F.M. Tunez et al. / Thermochimica Acta 523 (2011) 124–136
influence on the chlorination rate, varying the required time to
complete the reaction between about 8 min at higher partial pres-
sures and 25 min for the lowest. Also, as in the isothermic essays at
constant pressure, the influence of the exothermic effect produced
by the chlorination reaction is observed, which causes a partial
evaporation of some reaction products, producing a difference
between the mass changes expected and those observed.
The chlorination rate of GaAs with pure Cl2 (1 atm) in the tem-
perature range between 300 and 900 ◦C is slightly affected by the
reaction temperature. Between 300 and 700 ◦C, and after the reac-
tion is finished, solid residues constituted by Ga2O3 and GaOAs4
remain. The formation of these compounds is due to the inter-
action in gaseous phase of the trichloride of Ga and As from the
chlorination of GaAs and the O2 contained in the used Cl2.
The effect of the partial pressure of Cl2 on the chlorination rate
at 300 and 500 ◦C is noticeable. The chlorination at 300 ◦C is com-
pleted in approximately 8 and 42 min, working with 1 and 0.2 atm
of Cl2, respectively. The residues of the chlorination at 0.2 atm con-
tain only GaAs because the reaction stops due to a lack of products
in the vapor phase which may continue catalyzing the chlorination
reaction.
References
[1] R.R. Moskalyk, Gallium the backbone of the electronics industry, Miner. Eng.
16 (2003) 921–929.
[2] D.R. Lide, Handbook of Chemistry and Physics, CRC Press Inc., Florida,
2006–2007.
[3] W.S. Lim, S.D. Park, B.J. Park, G.Y. Yeom, Atomic layer etching of (1 0 0/1 1 1)
GaAs with chlorine and low angle forward reflected Ne neutral beam, Surf.
Coat. Technol. 202 (2008) 5701–5704.
[4] J. Cui, L. Zhang, Metallurgical recovery of metals from electronic waste: a
review, J. Hazard. Mater. 158 (2008) 228–256.
[5] J. Cui, E. Forssberg, Mechanical recycling of waste electric and electronic equip-
ment: a review, J. Hazard. Mater. B99 (2003) 243–263.
[6] J.E. Hoffmann, Advances in the extractive metallurgy of selected rare and pre-
cious metals, JOM 43 (1991) 18.
[7] P.K. Jena, E.A. Brocchi, Metal extraction through chlorine metallurgy, Mineral
Process Extractive Metallurgy Review 16 (1997) 211.
[8] S. Kubo, M. Yukinobu, O. Yamamoto, G. Nabeshima, J. Okajima, Recovery of
high purity gallium metal from gallium-arsenide scrap, in: J.H.L. Van Linden,
D.L. Stward Jr., Y. Sahai (Eds.), In the Second International Symposium-recycling
of Metals and Engineered Materials, TMS, 1990, pp. 505–513.
[9] A. Jenichen, C. Engler, Etching of GaAs(1 0 0) surfaces by HCl: density functional
calculations to the mechanisms, Surf. Sci. 475 (2001) 131–139.
[10] W. Hung, S. Wu, C. Chang, Low-temperature chlorination of GaAs(1 0 0), J. Phys.
Chem. 102 (1998) 1141–1148.
[11] P. Bond, P.N. Brier, J. Fletcher, W.J. Jia, H. Price, P.A. Gorry, Reactive scat-
tering study of etching dynamics: Cl2 on GaAs(1 0 0), Surf. Sci. 418 (1998)
181–209.
[12] L. Jalabert, P. Dubreuil, F. Carcenac, S. Pinaud, L. Salvagnac, H. Granier, C.
Fontaine, High aspect ratio GaAs nanowires made by ICP-RIE etching using
Cl2/N2 chemistry, Microelectron. Eng. 85 (2008) 1173–1178.
[13] W.S. Lim, S.D. Park, B.J. Park, G.Y. Yeom, Atomic layer etching of (1 0 0)/(1 1 1)
GaAs with chlorine and low angle forward reflected Ne neutral beam, Surf. Coat.
Technol. 202 (2008) 5701–5704.
The reaction mechanism depends on temperature range. At all
temperatures steps b–j operates. At temperatures between −30 ◦C
and 200 ◦C, step a must be added as the initiation one. At tem-
peratures between 300 ◦C and 500 ◦C, the initiation step involves
the step h. Finally, at temperature between 600 ◦C and 900 ◦C, the
initiation step is Cl2(g) thermal dissociation.
[14] F. Stietz, J.A. Schaefer, A. Goldmann, Chemisorption of chlorine on GaAs(1 0 0)
surfaces studied by high-resolution electron energy-loss spectroscopy, Surf.
Sci. 383 (1997) 123–129.
[15] T. Ohno, Theory of adsorption of Cl2 molecules on GaAs(0 0 1) surfaces, Surf.
Sci. 357 (1996) 322–326.
6. Conclusions
[16] Y.N. Park, J.K. Kim, J.H. Lee, Y.W. Joo, H.S. Noh, J.W. Lee, S.J. Pearton, N2 effect on
GaAs etching at 150 mTorr capacitively-couped Cl2/N2 plasma, Microelectron.
Eng. 87 (2010) 548–552.
[17] F.M. Tunez, M. del C. Ruiz, J.A. González, Kinetics of chlorination of GaAs, in:
2nd Mercosur Congress on Chemical Engineering EMPROMER, 2005.
[18] HSC Chemistry for Windows, versión 5.1, Outokumpu Research, Finland, 2002.
[19] F.M. Tunez, J. González, M. del C. Ruiz, Equipo experimental de laborato-
rio para realizar termogravimetrías en atmósferas corrosivas y no corrosivas,
P060100450 (2007).
The results obtained permit to state that GaAs chlorination is
a viable methodology to recover Ga and As from GaAs. At low
temperature, Ga and As chlorides are in solid and/or liquid state,
and the pressure vapor diagrams versus temperature indicate that
separation by distillation is feasible.
Acknowledgment
[20] N.K. Singh, R. Krout, Thermal and electron-induced reactions of AsCl3 on
GaAs(1 0 0), Surf. Sci. 442 (1999) 442–454.
The authors wish to thank the National University of San Luis,
CONICET (National Council of Scientific and Technical Research)
and FONCyT (Fund for Scientific and Technological Research) for
financial support.
[21] Card N◦ 20-0426, JCPDS Powder Diffraction File (2003).
[22] Card N◦ 76-0573, JCPDS Powder Diffraction File (2003).
[23] Card N◦ 89-4432, JCPDS Powder Diffraction File (2003).
[24] Card N◦ 32-0389, JCPDS Powder Diffraction File (2003).