Chemical interaction of Cu-In, Cu-Sn, and Cu-Bi solid solutions with liquid Ga-In and Ga-Sn eutectics

Article abstract of DOI:10.1134/S0020168506100025

The reactions of copper-based Cu-In, Cu-Sn, and Cu-Bi solid solutions with liquid Ga-In and Ga-Sn eutectics have been studied in situ by synchrotron x-ray diffraction. The results indicate that the dynamics of the process and the phase composition, grain

Full text of DOI:10.1134/S0020168506100025

ISSN 0020-1685, Inorganic Materials, 2006, Vol. 42, No. 10, pp. 1058–1064. © Pleiades Publishing, Inc., 2006.
Original Russian Text © A.I. Ancharov, T.F. Grigorieva, S.V. Tsybulya, V.V. Boldyrev, 2006, published in Neorganicheskie Materialy, 2006, Vol. 42, No. 10, pp. 1164–1170.
Chemical Interaction of Cu–In, Cu–Sn,
and Cu–Bi Solid Solutions with Liquid Ga–In
and Ga–Sn Eutectics
a
a
b
b
A. I. Ancharov , T. F. Grigorieva , S. V. Tsybulya , and V. V. Boldyrev
a
Institute of Solid-State Chemistry and Mechanochemistry, Siberian Division, Russian Academy of Sciences,
ul. Kutateladze 18, Novosibirsk, 630128 Russia
b
Novosibirsk State University, ul. Pirogova 2, Novosibirsk, 630090 Russia
e-mail: ancharov@mail.ru
Received August 10, 2005; in final form, March 22, 2006
Abstract—The reactions of copper-based Cu–In, Cu–Sn, and Cu–Bi solid solutions with liquid Ga–In and
Ga−Sn eutectics have been studied in situ by synchrotron x-ray diffraction. The results indicate that the dynam-
ics of the process and the phase composition, grain size, and microstructure of the resulting materials depend
on the components of the solid solution and eutectic.
DOI: 10.1134/S0020168506100025
INTRODUCTION
compounds formed? Will the components liberated in
this process influence the formation of the intermetallic
compounds that appear first?
Chemical interaction of powdered metals and alloys
with liquid gallium eutectics is basic to diffusion-hard-
ening solders. In such multicomponent systems, several
intermetallic phases form in parallel or in sequence,
influencing the phase formation process and the prop-
erties of the resulting material. For example, in the
reaction of gallium metal (tm = 29.9°C) or gallium-con-
taining eutectics with metal (copper, nickel, and other)
powders at ꢀ35°ë, the first to form is an intermetallic
phase with the highest possible gallium content. The reac-
tion proceeds until this phase is fully consumed [1–4]. If
a gallium eutectic is used as the liquid metallic phase,
the transfer of gallium from the eutectic to the forming
intermetallic phase is accompanied by the liberation of
the other component of the eutectic. Also, if the solid
phase is an alloy rather than a pure metal, one compo-
nent of the alloy reacts with gallium, while the other is
liberated. As a result, there are free metals from the
eutectic and alloy, which may also react to form an
intermetallic compound.
In this paper, we report in situ studies of reactions
between copper-based solid solutions and liquid gal-
lium eutectics and analyze the effects of the compo-
nents present on the dynamics of the process and the
phase composition, grain size, and microstructure of
the resulting materials.
EXPERIMENTAL
In our preparations, we used copper powder (PMS-
1), tin powder (POE), gallium (RF Standard GOST
12797-77), indium (RF Standard GOST 10297-94),
bismuth (Technical Specifications TU 6-09-3616-82),
80 wt % Cu + 20 wt % Sn solid solution (in what fol-
lows, Cu〈Sn〉), 80 wt % Cu + 20 wt % In solid solution
(Cu〈In〉), 90 wt % Cu + 10 wt % Bi solid solution
(Cu〈Bi〉), 88 wt % Ga + 12 wt % Sn eutectic
(LE(Ga−Sn)), and 75.5 wt % Ga + 24.5 wt % In eutectic
(LE(Ga–In)). The solid solutions were prepared by
mechanical alloying in a water-cooled AGO-2 high-
energy planetary ball mill under an argon atmosphere,
using 250-cm3 grinding vials and 5-mm-diameter balls.
The ball load was 200 g, and the sample weight was
10 g. The rotation rate of the supporting disk was
ꢀ1000 rpm.
Analysis of the literature indicates that this issue has
not yet been addressed in sufficient detail. The forma-
tion sequence of intermetallic compounds in systems
containing several intermetallics is unclear, and the fac-
tors governing this process are as yet not understood. It
is unclear whether the metals liberated from the eutec-
tic and alloy will crystallize or will react directly. Will
the resulting phases undergo phase transitions? How
will those components noninteracting with the other
Reactions between powder solid solutions and liq-
uid eutectics were studied by synchrotron x-ray diffrac-
tion (SXRD) using radiation from the fourth beamline
elements crystallize in the presence of the intermetallic at the VEPP-3 storage ring at the Siberian Synchrotron
1058

CHEMICAL INTERACTION OF Cu–In, Cu–Sn, AND Cu–Bi SOLID SOLUTIONS
1059
(a)
(b)
3000
2700
2400
2100
1800
1500
1200
900
2398
2356
2314
2272
2230
2188
600
300
0
1642 1684 1720 1768 1810 1852
Fig. 1. Portions of the SXRD pattern from the alloy prepared by reacting solid copper and liquid gallium: (a) 2D diffraction pattern,
(b) 3D representation of the diffraction pattern; X and Y are pixel coordinates in the detector system, and Z is the observed intensity
in arbitrary units.
Radiation Centre (Budker Institute of Nuclear Physics, the material obviously consisted of both fine, poorly
crystallized particles and large crystals (Fig. 1b).
Siberian Division, Russian Academy of Sciences). At a
wavelength λ = 0.368 Å, most diffraction peaks are sit-
uated at small diffraction angles and can be recorded
using a 2D x-ray detector array. We used a Marresearch
mar345 image plate detector and 0.4 × 0.4 mm x-ray
beam. The sample–detector distance was 400 mm, the
pixel size 0.1 × 0.1 mm, and the image plate diameter
345 mm. The read-out time at the smallest pixel size
and largest scanned plate diameter was within 2.5 min,
and the exposure time was 7.5 min. In this way, the for-
mation of intermetallic compounds was monitored at
10-min intervals for a total of 24 h. The intersection of
diffraction cones with the detector plate produces a set
of concentric circles. In contrast to conventional
diffractometers, which record only a narrow strip of
the entire diffraction pattern, a plate detector records
the entire diffraction pattern. Keeping the sample
immobile, one can follow reflections from individual
crystallites and evaluate their size and preferential
alignment.
In the reactions between the copper-based Cu–Sn
solid solution and Ga–Sn eutectic and between the cop-
per-based Cu–In solid solution and Ga–In eutectic, the
liberated component is the same element.
In the former system, the reaction follows the
scheme
Cu〈Sn〉 + LE(Ga–Sn)
CuGa2 + Sn.
The intermetallic compound CuGa2 forms immedi-
ately after mixing the components, while reflections
from Sn emerge 4 h after the beginning of the process,
when the copper and gallium are, for the most part,
bound. The intensity of these reflections then rises very
rapidly (Fig. 2). After 24 h, the product consists of
rather fine intermetallic particles (arcs in Fig. 3a) and
large Sn crystals (spots in Fig. 3a). The corresponding
diffraction pattern shows rather strong reflections from
Sn and substantially weaker reflections from the inter-
metallic phase (Fig. 3b).
RESULTS AND DISCUSSION
Our results indicate that the formation of the inter-
metallic compound is influenced by the liberated com-
ponent. The crystallite size of the first-formed interme-
tallic compound (CuGa2) is substantially smaller than
that of Sn, with insignificant scatter. Since the reflec-
tions from Sn are far stronger than those from the start-
ing powder, it is reasonable to assume that Sn crystal-
First, we studied the reaction between copper pow-
der and liquid gallium. As shown earlier [1–7], the
product of this reaction is the intermetallic phase
CuGa2. SXRD examination showed that the forming
CuGa2 ranged very widely in crystallite size, as evi-
denced by the spots and arcs in Fig. 1a. The average
crystallite size is difficult to evaluate from our data, but lizes from a liquid phase.
INORGANIC MATERIALS Vol. 42 No. 10 2006

1060
ANCHAROV et al.
(a)
(b)
10000
8994
7997
7000
6003
5006
4009
3012
2015
1018
21
10000
8994
7997
7000
6003
5006
4009
1126
1124
1082
1040
998
956
914
3012
2015
1018
21
1084
1042
1000
958
916
Sn
Sn
1387 1429 1471 1513 1555 1597
1387 1429 1471 1513 1555 1597
(c)
(d)
10000
10000
8994
7997
7000
6003
5006
4009
3012
2015
1018
21
8994
7997
7000
6003
5006
4009
3012
2015
1018
21
1124
1082
1040
998
956
914
1124
1082
1040
998
956
914
Sn
Sn
1387 1429 1471 1513 1555 1597
1387 1429 1471 1513 1555 1597
Fig. 2. Portions of SXRD patterns illustrating the time evolution of the reaction between a copper-based Cu–Sn solid solution and
the liquid Ga–Sn eutectic; the time after mixing the components is (a) 250, (b) 260, (c) 270, and (d) 500 min; the same designations
as in Fig. 1.
A similar effect was observed in the latter system,
where the reaction followed the scheme
Thus, we obtained similar results in the two solid
solution–liquid eutectic systems: the first to form is a
microcrystalline intermetallic compound; then, after an
induction period, the metal resulting from the reaction
between copper and gallium appears. The metallic
phase consists of much coarser grains in comparison
with the starting powder and primary (intermetallic)
phase.
Cu〈In〉 + LE(Ga–In)
CuGa2 + In.
The intermetallic compound CuGa2 was detected by
SXRD immediately after mixing the mechanochemi-
cally synthesized solid solution Cu〈In〉 with the Ga−In
eutectic and remained the only detectable phase during
the first 230 min of reaction. Next, the diffraction pat-
tern showed a reflection from indium (Fig. 4a), whose
intensity then increased rapidly (Fig. 4). After 11 h, the
reaction product consisted of microcrystalline CuGa2
and coarse indium crystallites.
To analyze reactions between a solid solution and
liquid eutectic in systems where two different compo-
nents are liberated, the same component being liberated
from the liquid eutectic in one system and from the
solid solution in the other system, we examined the
reactions between the copper-based Cu–Sn solid solu-
INORGANIC MATERIALS Vol. 42 No. 10 2006

CHEMICAL INTERACTION OF Cu–In, Cu–Sn, AND Cu–Bi SOLID SOLUTIONS
(a) (b)
1061
10000
8996
7998
7000
6002
5004
4006
3008
2010
1012
14
Sn
Sn
Sn
Sn
2247
2205
2163
2121
2079
2037
Sn
1308 1350 1392 1434 1476 1518
Fig. 3. Portions of the SXRD pattern from the end product of the reaction between a copper-based Cu–Sn solid solution and the
liquid Ga–Sn eutectic; the same designations as in Fig. 1.
(a)
(b)
10000
9000
8000
7000
6000
5000
4000
3000
2000
1000
0
10000
9000
8000
7000
6000
5000
4000
1317
1312
1270
1228
1186
1144
1102
3000
2000
1000
0
1275
1233
1191
1149
1107
In
In
2940 2982 3024 3066 3108 3150
2940 2982 3024 3066 3108 3150
Fig. 4. Portions of SXRD patterns illustrating the time evolution of the reaction between a copper-based Cu–In solid solution and
the liquid Ga–In eutectic; the time after mixing the components is (a) 210 and (b) 230 min; the same designations as in Fig. 1.
tion and Ga–In eutectic and between the copper-based systems:
Cu–In solid solution and Ga–Sn eutectic.
⎫
⎪
⎬
⎪
⎭
Cu 〈Sn〉 + LE(Gu–In) (1)
In both systems, the first to form, immediately after
mixing, is CuGa2. The liberated indium and tin may
react with one another since, according to the equilib-
rium In–Sn phase diagram, this system contains two
intermetallic compounds: In3Sn and InSn4. Indium
and/or tin may also crystallize with no reaction. Thus,
the following chemical reactions are possible in these
Cu 〈In〉 + L (Ga–Sn) (2)
E
CuGa + In Sn + Sn
⎧
⎪
⎨
⎪
⎩
2
3
CaGa2 + InSn4 + In
CuGa2 + Sn + In.
INORGANIC MATERIALS Vol. 42 No. 10 2006

1062
ANCHAROV et al.
began to drop (Fig. 5b), and a reflection from the inter-
(a)
metallic compound In3Sn emerged. Subsequently, its
intensity increased, while the reflection from Sn van-
ished altogether (Fig. 5c). The disappearance of the
reflection from Sn may be due to a tilt of the crystallite,
resulting in a deviation from the Bragg angle. To rule
out this effect, we took a diffraction pattern while rock-
ing the sample by 15°. The diffraction pattern showed
reflections from the intermetallic compounds CuGa2
and In3Sn but not from Sn.
10000
9000
8000
7000
6000
5000
4000
3000
2000
1000
0
Sn
1769
1727
1685
1643
1601
1559
Calculations indicate that the Sn could not be
entirely bound in the intermetallic compound In3Sn:
since the reaction mixture contains roughly equal
amounts of indium and tin, only part of the Sn could
participate in the formation of the intermetallic com-
pound. Therefore, most of the Sn was undetectable by
SXRD.
2622 2664 2706 2748 2780 2832
(b)
10000
Sn is liberated from the solid solution concurrently
with the formation of the intermetallic compound
CuGa2 and can be detected by SXRD after 2 h of the
reaction between the solid solution and eutectic, i.e.,
after most of the copper from the solid solutions has
reacted with gallium to form CuGa2. It is reasonable to
assume that the Sn dissolved in the liquid gallium
eutectic and that the rapid increase in the intensity of
the reflections from Sn was due to its crystallization
from the melt. The same is suggested by the large size
of the Sn crystallites.
9000
8000
7000
6000
5000
4000
3000
2000
1000
0
Sn
1763
1721
1679
In Sn
3
1637
1595
1553
2621 2663 2705 2747 2789 2831
Free indium must appear in the system after all of
the gallium is bound in CuGa2. The absence of reflec-
tions from indium even after all of the gallium is com-
bined with copper may be due to indium adsorption in
the form of a thin, “liquidlike” layer on the surface of
the CuGa2 formed.
(c)
10000
9000
8000
7000
6000
5000
4000
3000
2000
1000
0
It is well known that the reaction between solid and
liquid metals leads to the formation of an intermetallic
phase which is richer in the component that is in the liq-
uid (or liquidlike) state.
In Sn
3
1763
1721
1679
1637
1595
1553
Therefore, in the case of liquid (or liquidlike) Sn, the
forming intermetallic phase must be richer in Sn. To
validate this assumption, we investigated the reaction
between the copper-based Cu–In solid solution and liq-
uid Ga–Sn eutectic. In this reaction, the liberated ele-
ments are the same as above, but indium is liberated
from a solid phase, while tin, from the liquid eutectic as
copper and gallium react to form the intermetallic com-
pound CuGa2.
2621 2663 2705 2747 2789 2831
Fig. 5. Time evolution of the SXRD pattern after mixing
powder of the mechanochemically synthesized copper-
based Cu–Sn solid solution with the liquid Ga–In eutectic;
reaction time of (a) 150, (b) 330, and (c) 350 min; the same
designations as in Fig. 1.
SXRD examination showed that the intermetallic
phase CuGa2 appeared immediately after mixing the
copper-based Cu–In solid solution with the liquid Ga–
Sn eutectic. The intensity of the reflections from CuGa2
increased very rapidly, and neither indium nor tin were
In the former system, SXRD showed only grain
growth of the intermetallic compound CuGa2 during
the first 140 min of reaction. After 140 min, a reflection
from Sn emerged, and its intensity then increased rap-
idly (Fig. 5a). After 330 min, however, its intensity detected during the first several (>8) hours of reaction.
INORGANIC MATERIALS Vol. 42 No. 10 2006

CHEMICAL INTERACTION OF Cu–In, Cu–Sn, AND Cu–Bi SOLID SOLUTIONS
1063
Bi + BiIn2
BiIn2
BiIn2
BiIn2
BiIn2
100
201
111
110
200
210
300
211
Fig. 6. SXRD pattern of the sample prepared by reacting a
copper-based Cu–In solid solution with the liquid Ga–Sn
eutectic for 24 h; the indexed rings are from the intermetal-
lic phase InSn ; the others are from CuGa .
Fig. 7. SXRD pattern of the sample prepared by reacting a
copper-based Cu–Bi solid solution with the liquid Ga–In
eutectic.
4
2
After 8 h, the intermetallic phase InSn4 appeared
The formation of CuGa2 in this system must be
(Fig. 6), while crystalline indium was still missing. The accompanied by Bi and Sn liberation, since the Bi–Sn
formation of the intermetallic phase InSn4 lends sup-
port to the above assumption.
phase diagram has no intermetallic compounds. Indeed,
SXRD results show that the reaction products in this
system are CuGa2, Bi, and Sn (Fig. 8).
Similar results were obtained for a mixture of the
copper-based Cu–Bi solid solution and liquid Ga–Sn
eutectic. In this system, several reactions are also pos-
sible:
CuGa + Bi + In,
⎧
⎪
⎨
⎪
⎩
2
Cu 〈Bi〉 + LE(Ga–In)
CuGa2 + BiIn2,
CuGa2 + BiIn + In.
Bi
Bi
According to SXRD data, the reaction products in
this system are bismuth metal and the intermetallic
phases CuGa2 and BiIn2 (Fig. 7). This phase composi-
tion suggests that, like in the above system, indium is
present in the form of a surface layer on intermetallic
particles and may react with crystalline bismuth to form
an indium-rich phase.
Sn
Sn
We also investigated a system in which the metals
being liberated during CuGa2 formation do not form
intermetallic phases. The metals must then crystallize
with no reaction:
Fig. 8. SXRD pattern of the sample prepared by reacting a
copper-based Cu–Bi solid solution with the liquid Ga–Sn
eutectic.
Cu〈Bi〉 + LE(Ga–Sn)
CuGa2 + Bi + Sn.
INORGANIC MATERIALS Vol. 42 No. 10 2006

1064
ANCHAROV et al.
Alloys during Solidification, Fiz.-Khim. Mekh. Mater.,
1969, vol. 5, no. 4, pp. 455–458.
CONCLUSIONS
Our results demonstrate that the intermetallic phase
CuGa2 resulting from the reaction between copper and
liquid gallium ranges widely in crystallite size. In all of
the reactions between a binary solid alloy and liquid
eutectic, the first to form is fine-particle CuGa2. The
intermetallic phase forming at a later stage of the reac-
tion consists of much coarser grains in comparison with
the starting powder and primary intermetallic phase.
The present results on the reactions between the
copper-based Cu–Sn solid solution and Ga–In eutectic,
copper-based Cu–In solid solution and Ga–Sn eutectic,
and copper-based Cu–Bi solid solution and Ga–In
eutectic indicate that the second-formed intermetallic
phase is richer in the element that was in the liquid
phase (in the eutectic) in the starting mixture, indepen-
dent of the melting points of the elements present.
2. Glushkova, L.I. and Petrov, G.L., Copper–Gallium and
Gallium–Nickel Solder Pastes, Svar. Proizvod., 1968,
no. 11, pp. 36–37.
3. Ivanov, E. and Grigorieva, T., Room Temperature
Solder Material Based on the Reaction of Nanocrystal-
line Mechanochemically Alloyed Cu–Sn Alloy with
Liquid Eutectic Alloys, Advances in Brazing and Sol-
dering Technology, Conf. Proc. New-Mexico, 2000,
pp. 587−594.
4. Ancharov, A.I., Grigorieva, T.F., Nikitenko, S.G., and
Sharafutdinov, M.R., X-ray Diffraction and EXAFS
Studies of Reactions in Diffusion-Hardening Cu–Ga
Solders, Poverkhnost, 2002, no. 7, pp. 23–28.
5. Grigorieva, T.F., Barinova, A.P., Ivanov, E.Yu., and
Boldyrev, V.V., Stage Sequence in Mechanochemical
Synthesis of Nanometric Solid Solutions in Metal Sys-
tems, J. Metastable Nanocryst. Mater., 2003, vol. 15/16,
pp. 553–556.
ACKNOWLEDGMENTS
This work was supported by the RF Ministry of Edu-
cation through the Universities of Russia Program (Basic
Research in Natural and Human Sciences at Higher Edu-
cation Institutions, grant no. UR-06.01.209).
6. Savitskii, A.P., Grigorieva, T.F., Barinova, A.P., et al.,
Mechanical Alloying Mixtures of Solid and Liquid Met-
als, Proc. PM2004 Powder Metallurgy World Congr.,
Vienna: EPMA, 2004, vol. 1, pp. 171–176.
7. Ancharov, A.I. and Grigorieva, T.F., Investigation of the
Mechanism of Interaction between Reagents in Alloys
Based on Cu–Sn System, Nucl. Instrum. Methods Phys.
Res., Sect. A, 2005, vol. 543, pp. 139–142.
REFERENCES
1. Tikhomirova, O.I., Pikunov, M.V., Marchukova, I.D.,
et al., Structural Transformations of Copper–Gallium
INORGANIC MATERIALS Vol. 42 No. 10 2006