Non-thermal reduction of indium oxide and indium tin oxide by mechanochemical method

Article abstract of DOI:10.1016/j.jallcom.2009.04.114

A non-thermal process for reducing indium(III) oxide (In₂O₃) and/or indium tin oxide (ITO) into indium-metal by milling with lithium nitride (Li₃N) under (NH₃) or nitrogen (N₂) gas environment is prop

Full text of DOI:10.1016/j.jallcom.2009.04.114

Journal of Alloys and Compounds 484 (2009) 422–425
Contents lists available at ScienceDirect
Journal of Alloys and Compounds
journal homepage: www.elsevier.com/locate/jallcom
Non-thermal reduction of indium oxide and indium tin oxide by
mechanochemical method
a,∗
a
b
a
a
Junya Kano , Eiko Kobayashi , William Tongamp , Shoko Miyagi , Fumio Saito
a
Institute of Multidisciplinary Research for Advanced Materials, Tohoku University, 2-1-1, Katahira, Aoba-ku, Sendai 980-8577, Japan
Faculty of Engineering and Resource Science, Akita University, 1-1 Tegata-Gakuen cho, Akita 010-8502, Japan
b
a r t i c l e i n f o
a b s t r a c t
Article history:
2 3
A non-thermal process for reducing indium(III) oxide (In O ) and/or indium tin oxide (ITO) into
Received 6 November 2008
Accepted 26 April 2009
Available online 3 May 2009
indium–metal by milling with lithium nitride (Li3N) under (NH3) or nitrogen (N2) gas environment is
proposed in this paper. Milling operation causes mechanochemical reaction in the systems of In2O3/ITO
and Li3N, to form In and LiOH. The latter is soluble in water, so that the milled sample was subjected
to washing with water, enabling us to recover indium–metal. According to the characterization of the
milled products by X-ray diffraction (XRD), the reduction of In2O3/ITO can be achieved in a short period
of time. Analytical data by inductively coupled plasma (ICP) from dissolution of the pellets in acidic solu-
tion clearly shows that indium–metal concentration is over 95% and the yield of In from the starting
oxide sample is more than 97%, depending on the milling condition. The reaction mechanism between
In2O3/Li3N and NH3/N2 is also discussed in the paper, and this could be applied to recover indium–metal
from electric device wastes containing ITO.
Keywords:
Indium oxide
Indium tin oxide
Lithium nitride
Milling
Mechanochemical reaction
Planetary ball mill
©
2009 Elsevier B.V. All rights reserved.
1. Introduction
oxide (Ga O ) and Li N under NH , preparing GaN as a main prod-
2 3 3 3
uct. This reaction seems to be non-thermal reduction of oxide
through mechanochemical reaction with an aid of reducing agents.
Indium (In), a rare metal, is generally found in low concen-
trations in sulphide ores, and most commonly associated with
zinc-bearing materials and less in copper and lead from which
indium is obtained in by-products such as residues, flue dusts, and
slags. It has many important applications; mainly as thin films of
indium tin oxide (ITO) for liquid crystal displays (LCDs). It also
finds application as constituent of fusible alloys with precious and
base metals, lowering their melting points, in electrical compo-
nents and semiconductors in the form of indium phosphide (InP),
among others [1–4]. Demand for indium in Japan reached 888 t
in 2006, of which about half of the material was recovered from
scrap and a large amount could be recovered from weld metal
alloys, scraped ITO and indium phosphide in LCDs [1]. Recovery
of indium from sulfate or chloride leach liquors by hydrometallur-
gical processing routes followed by solvent extraction, cementation
etc. is widely reported in literature [5–9]. However, there has still
been a strong demand to recover indium (metal) from a waste
material containing In. Recently, Zhang et al. have reported on
application of mechanochemical phenomena to recover useful
materials from wastes [10–13]. Similarly, Kano et al. have reported
mechanochemical reaction between gallium (Ga) and/or gallium
The reducing agents are Li N and NH3 in this case, and they play
3
a significant role in the reduction of the oxides. Kano et al. have
extended their reduction process to In O to form In–metal [14–16].
2
3
The main purpose of this paper is to provide informa-
tion on a non-thermal process for recovering In–metal through
mechanochemical reduction of In O /ITO by its milling with Li N
2
3
3
under non-oxidation state of NH and/or N gas environments. The
3
2
milled sample was subjected to washing with water to recover
In–metal as pellets. The purity of the recovered In–metal is over
95%.
2.
Experimental
In2O3 and Li3N used as starting materials were supplied by Wako Pure Chemical
Industries, Ltd., Japan and ITO sample was supplied by Aldrich.
A planetary ball mill (P-7, Fritcsh, Germany), having a pair of ZrO2 mill pots,
charged with 24 × 10 mm diameter ZrO2 balls each was used for the milling of In2O3
and Li3N, and was conducted under either NH3 or N2 gas atmospheres, respectively
to induce mechanochemical reaction between the starting materials. The diameter
and length of the mill pots are the same values as 40 mm and inner pot volume of ca.
3
45 cm . 2.76 g sample mixture of In2O3 (2.0 g) and Li3N (0.76 g) (1:3 (mol/mol) ratio)
was carefully mixed inside a glove box under argon gas environment, and charged
into the mill pot and the pot was set in a container made of stainless steel (overpot).
The inner air in the mill pot was degassed with a vacuum pump, and NH3 and/or N2
gases were charged at 0.8 MPa, respectively. Both charged mill pots were set on mill
device and rotated at 300 rpm for different times ranging from 30 to 180 min.
After the milling operation, a small amount of the product was removed for
characterization and the milled product was further washed in the same mill with
∗ Corresponding author. Tel.: +81 22 217 5136; fax: +81 22 217 5136.
E-mail address: kano@tagen.tohoku.ac.jp (J. Kano).
0
925-8388/$ – see front matter © 2009 Elsevier B.V. All rights reserved.
doi:10.1016/j.jallcom.2009.04.114

J. Kano et al. / Journal of Alloys and Compounds 484 (2009) 422–425
423
Fig. 2. X-ray diffraction patterns showing effects of NH3 and N2 gases during MC
reduction of In2O3 by milling in the presence of Li3N. Mixing ratio of In2O3/Li3N was
fixed at (1:3) moles and milling conducted for 120 min at 300 rpm.
Fig. 1. Schematic illustration of the experimental procedure used to effect
mechanochemical reduction of In2O3/ITO.
Fig. 4 shows XRD patterns of a mixture of In O and Li N at
2
3
3
(1:3) molar ratio milled for different periods of time under NH3
gas atmosphere. Peaks of In–metal clearly appear for the sample
mixtures milled for 30 min and as milling is progressed, the In2O3
peaks completely disappear from the patterns. It is seen that inten-
sity of In–metal peaks significantly increase in the patterns of the
mixtures as milling progresses. However, further milling for more
water for 10 min to remove the by-product. A schematic flow of the operation to
form In from the oxides is given in Fig. 1.
The milled samples were characterized by XRD analysis using Rigaku, RINT-
2
200/PC system with a CuK␣ irradiation source (ꢀ = 1.5405 Å) at 40 kV and 20 mA
◦
◦
in a continuous scan mode between 10 and 60 in 2ꢁ. 0.1 g of the solid pellets
obtained after washing was dissolved in nitric acid for analysis by ICP (Optima 3300,
PerkinElmer) to evaluate purity and recovery of indium.
3. Results and discussion
3.1. Milling under NH3 and N2 gas environments
Fig. 2 shows XRD patterns of the samples of In O , and the mix-
2
3
ture of In O and Li N at (1:3) molar ratio each milled for 120 min
2
3
3
under NH and N gas environments. The XRD patterns of the milled
3
2
mixtures indicate that complete reduction of In O to In–metal has
2
3
been achieved under both gas environments, with no significant
effect, except for the peak height. The height of the main peak for
NH is slightly higher than that for N , suggesting that NH is rather
3
2
3
effective reductant than N . All the same, peaks of the starting
2
materials in both products are not seen in the patterns, indicating
complete reduction of oxide is achieved within 120 min of milling
under this condition.
Fig. 3 shows XRD patterns of In O sample and other three kinds
2
3
of mixture of In O3 and Li N at different molar ratios of (1:1) to
2
3
(
1:3), each milled for 120 min under NH gas atmosphere. It is seen
3
that the sample of In O3 without Li N (1:0) shows no peaks of
2
3
In–metalin the pattern. As for the milledmixtures, significantpeaks
of In are seen in the patterns, and the peaks of In for the (1:3) mix-
ture is more sharp than those for the (1:2) mixture. This means that
the amount of Li N added to In O is important and more amount
3
2
3
of Li N is better for the reduction of In O . In other words, excess
3
2
3
Fig. 3. X-ray diffraction patterns showing effect of Li3N addition from 0 to 3 mols
Li N is appropriate to cause the reduction of In O into In in the
3
2
3
(In O /Li N = 1:0, 1:1, 1:2, 1:3) on MC reduction of In O by milling in NH3 gas for
2
3
3
2
3
milling condition.
120 min at 300 rpm.

4
24
J. Kano et al. / Journal of Alloys and Compounds 484 (2009) 422–425
3.2. Washing the milled sample to collect In–metal
Washing the mixtures of In O and Li N prepared at (1:3) molar
2
3
3
ratio, milled for different periods of time was performed to collect
In–metal.
Fig. 5 shows photographs of In–metal pellets recovered from the
sample mixtures milled for different periods of time. It is seen that
In–metal can be recovered from the mixture milled for only 30 min,
and it looks almost the same figures in all the milling periods of
time.
The induced solid state reaction between In O3 and Li N under
2
3
NH3 and N2 gas atmospheres could be given by the following Eqs.
1) and (2):
(
In O + Li N + NH → 2In + 3LiOH + N
(1)
(2)
2
3
3
3
2
In O + 2Li N + N → 2In + 3Li O + 2N
2
3
3
2
2
2
In the preliminary experiment, milling the staring mixture
In O and Li N) for 120 min in the absence of NH3 and N2 gases
(
2
3
3
gave us the result of the formation of indium–metal in the product,
although, peaks of In O remained as a dominant component in the
2
3
product. Thus, a partial reaction between In O3 and Li N would be
2
3
shown as below, when NH3 and N2 gases are not present;
In O + 2Li N → 2In + 3Li O + N
(3)
2
3
3
2
2
Chemical constituents of the In–metal pellets recovered are
shown in Table 1. The purity of indium in the pellets is around 95%
in spite of the milling periods of time. As for the lithium as impurity
in the pellet, it shows around 0.5% or less. Total percent is accounted
as a sum of In purity (%) and Li impurity (%), and these values are
ranged from 94.5% to 97.1% for all milling conditions. These data are
slightly lower than those in In recovery calculated from the weight
of the starting sample and In purity. This may be due to experi-
mental loss detected and the main loss would be caused from the
Fig. 4. X-ray diffraction patterns of In2O3/Li3N (1:3) mixture milled in NH3 gas
environment for different milling times at 300 rpm.
than 120 min does not make sense, and no significant change in the
pattern for the samples milled for 120 and 180 min. Consequently,
prolonged milling is preferable for the reduction of In O , but too
2
3
much milling is not effective at all.
un-reacted In O3 during the milling operation.
2
Fig. 5. Shows photographs of In–metal pellets obtained after milling In2O3/Li3N (1:3) mixture in NH3 gas atmosphere for different times between 30 and 180 min and washed
in water for 10 min resulting in formation of pellets.

J. Kano et al. / Journal of Alloys and Compounds 484 (2009) 422–425
425
Table 1
4. Conclusion
Evaluation of purity and recovery for In–metal as a function of grinding time from
data obtained by ICP analysis.
This work was done to find out the possibility that the
mechanochemical reduction of In O3 by milling with Li N under
Grinding time (min) In purity (%) Li impurity (%) Total (%) In recovery (%)
2
3
3
6
0
0
95.73
96.76
94.21
94.25
0.41
0.37
0.39
0.29
96.14
97.13
94.61
94.54
98.12
97.25
97.11
97.39
NH3 and N2 gas atmosphere can be achieved. Followings are the
summary of this work:
120
180
1
. A non-thermal reduction of In O to In–metal by mechanochem-
2 3
ical route to induce solid state reaction conducted under NH and
3
N2 gas environments in the presence of Li N was confirmed.
3
2
3
. NH3 is more effective than N2 in terms of the reduction of In O3
2
. In the milling operation (planetary mill used in the experiment),
prolonged milling is more effective than short milling times, but,
much longer milling times for more than 120 min does not make
sense. Moderate mill speed at 300 rpm is enough to cause the
reduction.
4
. Washing the milled sample enables us to collect high purity
In–metal from the milled sample. The purity of In–metal col-
lected is around 95% in all experimental condition.
5. This process can be extended to reduce ITO for In and is con-
firmed by its reductive reaction and experimental results.
References
[
1] J.D. Jorgenson, M.W. George, U.S. Geological Survey, Open-File Report 2004-
300 (2005) URL: http://pubs.usgs.gov/of/2004/1300/index.html.
1
[
[
2] A.M. Alfantazi, R.R. Moskalyk, Miner. Eng. 16 (2003) 687–694.
3] C.D. Gribble, Rutley’s Elements of Mineralogy, 27th ed., Unwin Hyman Publica-
tion, London, 1988.
[4] H. Yamaura, T. Jinkawa, J. Tamaki, K. Moriya, N. Muira, N. Yamazoe, Sens. Actu-
ators B35–36 (1996) 325–332.
[
[
5] M.A. Barakat, Hydrometallurgy 49 (1998) 63–73.
6] X. Zhang, G. Yin, Z. Hu, Talanta 59 (2003) 905–912.
Fig. 6. X-ray diffraction patterns of indium tin oxide (ITO) and Li3N samples
1:3 = ITO/Li3N) milled in NH3 gas atmosphere for 120 min at 300 rpm.
[7] S.V. Vartak, V.M. Shinde, Talanta 45 (1998) 925–930.
[8] M.C.B. Fortes, A.H. Martins, J.S. Benedetto, Miner. Eng. 16 (2003) 659–663.
(
[9] B. Gupta, N. Mudhar, I. Singh, Sep. Purif. Technol. 57 (2007) 294–303.
[
10] E. Ivanov, C. Suryanarayana, J. Mater. Synth. Proc. 8 (3/4) (2000) 235–244.
Substituting In O3 with indium tin oxide (ITO) and milling the
2
[11] F.J. Fernández-Bertran, Pure Appl. Chem. 71 (4) (1999) 581–586.
[12] P. Butyagin, A: Physicochem. Eng. Aspects 160 (1999) 107–115.
sample with Li N under NH3 gas atmosphere was made to confirm
3
[
13] Q. Zhang, J. Kano, F. Saito, KONA 19 (2001) 7–14.
the reduction of ITO by similar mechanochemical reaction. We have
found that the similar reaction has been taken place by the milling
condition, and Fig. 6 is as evidence showing XRD patterns of the
[
14] J. Kano, L. Jenfeng, I. Kang, W. Tongamp, E. Kobayashi, F. Saito, Chem. Lett. 36 (7)
(
2007) 900–901.
[15] J. Kano, E. Kobayashi, W. Tongamp, F. Saito, J. Alloys Compd. 464 (2008) 337–339.
16] J. Kano, E. Kobayashi, W. Tongamp, M. Shoko, F. Saito, Chem. Lett. 37 (2) (2008)
04–205.
[
sample mixture milled for 120 min at 300 rpm under NH gas atmo-
3
2
sphere. It is seen that ITO is reduced to In–metal formed in the prod-
uct, and no significant peaks of the starting samples in the pattern.