Kagata et al.: Reduction of the Oxide Layer on a Lead-Free Solder (1/4)
Reduction of the Oxide Layer on a Lead-Free Solder Material
by Hydrogen Radicals
S. Kagata, T. Nakashima and A. Izumi
Kyushu Institute of Technology, 1-1 Sensui, Tobata, Kitakyushu Fukuoka 804-8550, Japan
(Received January 30, 2009; accepted October 20, 2009)
Abstract
Recently, there has been a demand for lead (Pb)-free solders for use in the electronics industry. However, this poses a
challenge in their practical applications because the main component of typical Pb-free solders is tin (Sn) and therefore
these solders are prone to oxidize easily. In this paper, we propose a novel technique for the reduction of oxides of the
powder which is the raw material of Pb-free solders, and which has Sn as its main component using NH3 decomposed
species generated using a hot-wire (HW) method. It is confirmed that the tin oxide can be reduced by this treatment
and the re-oxidation of the solder is also suppressed. Moreover, it is also confirmed that this treatment is effective in
controlling the particle size of the Pb-free solder. Further, we clarify that the activation energy of the reduction of the
tin oxide is almost zero by this process is almost zero, which indicates that the temperature dependence of the reduction
is low.
Keywords: Lead-Free Solder, Oxide, Ammonia, Radical, Hot-Wire Method, Raw Material of Solder
1.
Introduction
related contaminations from the surface of the raw mate-
Lead (Pb)-containing solder alloys have long been used
rial powder of Pb-free solders by using hydrogen radicals.
as solder materials for the interconnection of microelec-
The hydrogen radicals are generated by a hot-wire (HW)
tronic devices. The popularity of these solder alloys is pri-
decomposition of NH3 or H2. This method can effectively
marily due to their low cost, good wettability, and high
reduce the amount of oxides of various metals,[7] the
strength, which ensures a high interconnection reliability
carbon contaminations on metals,[8] and photoresist
during long-term use. However, due to the inherent tox-
removal,[9] even at low substrate temperatures.
icity of Pb and Pb-containing compounds, the reduction of
the Pb-containing compounds is strongly demanded in the
2.
Experimental
electronics industry.[1] Therefore, serious efforts are
Figure 1 shows a schematic diagram of the HW appara-
being made to develop a suitable Pb-free alternative that
tus used for the generation of ammonium or hydrogen rad-
can replace the conventional Sn–Pb eutectic solder.[2]
icals. NH3 or H2 gas was used as the reduction gas and
However, typical Pb-free solders are tin (Sn) alloys and
introduced into the system at a flow rate of 50 sccm. A Sn
these solders are prone to oxidize easily. Furthermore, the
substrate and Pb-free solder with the composition of Sn–
resultant oxide causes problems such as a decrease in wet-
96.5 mass% Cu–3.0 mass% Ag–0.5 mass% were placed in a
tability and the generation of whiskers.[3, 4] Therefore, a
vacuum chamber; the back pressure inside the chamber
flux is used to remove the oxides and improve the wetabil-
was below 2.0 × 10–5 Torr. The Pb-free solder powder was
ity of the solder.
adhered to the substrate (10 mm × 10 mm) using carbon
The application of the flux, however, may lead to the
tape. The typical conditions under which the reduction was
generation of its residue. This residue adds to the global
carried out are shown in Table 1. A zigzag-shaped tungsten
environmental load.[5] Therefore, the use of a fluxless sol-
wire with a diameter of 0.5 mm and a length of 800 mm
dering process that have been subjected to plasma treat-
was used as the catalyst. It was placed parallel to the sub-
ment has been proposed.[6]
strate holder at a distance of 50 mm. The temperature of
In this paper, we propose the elimination of carbon-
the tungsten wire was maintained at approximately
1
Transactions of The Japan Institute of Electronics Packaging
Fig. 1
Vol. 2, No. 1, 2009
Schematic diagram of the HW apparatus used for the
Fig. 2
XPS spectrum of Sn(3d5/2) before and after NH3
treatment for 5 min at various Sn substrate temperatures.
generation of ammonium or hydrogen radicals.
Table 1
Processing conditions.
Cleaning gas
NH3, H2
Substrate temperature (°C)
25–150 (Sn), 25–50 (Pb free solder)
W wire temperature (°C)
1200
W wire length, diameter (mm)
800, 0.50
Distance between wire and substrate (mm)
50
NH3 or H2 flow rate (sccm)
50
Gas pressure (Torr)
0.03
1200°C. The temperature of the substrate holder was mea-
reduced the oxide layer present on the Sn surface. From
sured using a thermocouple. The treatment time for the Sn
these spectra, the thickness of the oxide layer was calcu-
substrate was 5 min, while that for the Pb-free solder was
lated using the following equation[7]:
20–180 min.
The characterization of the surface conditions of the Sn
substrate and the Pb-free solder was carried out using X-
⎡⎛ δ M λ ⎞ I
⎤
d = λO sinθ ln ⎢⎜ Sn Sn Sn ⎟ O + 1⎥
⎣⎝ δ O MO λO ⎠ I S
⎦
rays photoelectron spectroscopy (XPS) measurements
Here, d is the thickness of the oxide layer, λ Sn is the
using monochromatic Mg-Kα radiation with emission
photoelectron escape depth of the Sn surface, λ O is the
angle of 90°. In addition, the Pb-free solder powder was
photoelectron escape depth of the Sn oxide layer, θ is the
observed at various points in the chamber using an optical
photoelectron take off angle, δ Sn is the photoelectron emis-
microscope (HIROX, KH-3000).
sion cross-section of the Sn substrate, δ O is the photoelectron emission cross-section of the Sn oxide, MSn is the
3.
Results and Discussion
3.1 Reduction of the oxide layer on the Sn surface
molecular weight of the Sn, MO is the molecular weight of
the Sn oxide, IS is the photoelectron intensity from Sn and
Figure 2 shows the XPS spectra of Sn(3d5/2) before and
IO is the photoelectron intensity from Sn oxide. For the
after NH3 treatment for 5 min at various temperatures of
calculation of d, the following values were substituted into
the Sn substrate. Here, we used Sn substrates instead of
the above equation: λ Sn = 1.87 nm, λ O = 2.20 nm, θ = 90°,
the Pb-free solder powder for the experiment to make han-
δ Sn ≅ δ O, MSn = 118.7, MO = 150.7.
dling of the sample and control of the substrate tempera-
The reduction rate was defined as the difference in the
ture better. After the oxide reduction process, the intensity
thickness of the oxide layer before and after treatment per
of the Sn–O peak (486.7 eV) decreased at all substrate tem-
unit of treatment time. The Sn oxide reduction rate was
peratures, and simultaneously, the Sn–Sn (484.9 eV) peak
found to be approximately 0.85 nm/min at all substrate
appeared. This result implies that the ammonia species
temperatures. The activation energy was obtained from
2
Kagata et al.: Reduction of the Oxide Layer on a Lead-Free Solder (3/4)
Fig. 3
Relationship between the treatment time and the
thickness of the oxide layer in the b-free solder powder. The
substrate temperature was varied between 25 and 60°C.
Fig. 5
Oxide thickness of the Sn oxide layer in the Pb-free
solder powder that was left in air after the NH3 treatment. The
thickness of the oxide layer in the Pb-free solder powder
before treatment is shown by the broken line in the figure.
ance, σ 2, was defined by the following expression.
n
σ2 =
∑ ( x − xi )
i =1
n
Here, x is the average diameter of the particle size, xi is
the diameter of the individual particle size and n is the
number of samples.
We confirmed that, with appropriate time, this treatment
was effective in maintaining the powder size of Pb-free
Fig. 4
Variance of the particle size of the powder of Pb-free
solder with NH3 treatment time.
solder at a particular value. The substrate temperature was
varied between 25 and 60°C.
Figure 5 shows the oxide thickness of the Sn oxide layer
in the powder of Pb-free solder that was left in air after the
the relationship between the oxide reduction rate and the
NH3 treatment. The thickness of the oxide layer in the the
reciprocal of the substrate temperature (Arrhenius plot).
Pb-free solder powder before treatment is shown by the
10–2
broken line in this figure. We confirmed that more than 20
eV. This result implies that the Sn–O reduction rate is
days after treatment, the thickness of the oxide layer in the
independent of the substrate temperature. In addition, we
Pb-free solder powder is lower than that before treatment.
obtained almost the same results using H2 instead of NH3.
Therefore, we confirmed that the re-oxidation of the Pb-
Therefore, it may be possible that hydrogen radicals in the
free solder is suppressed by our proposed cleaning pro-
NH3-decomposed species reduce the oxide.
cess. We think that the hydrogen radicals cover with the
3.2 Reduction of the oxide layer in the powder of
solder after treatment, which prevents the oxidation of the
Pb-free solder
solder.
The activation energy was found to be as small as 3.0 ×
Figure 3 shows the relationship between the treatment
time and the thickness of the oxide layer in the Pb-free
4.
Conclusion
solder powder. The substrate temperature was varied
The reduction of oxides of Sn in an Sn substrate and Pb-
between 25 and 60°C. The thickness of the Sn oxide layer
free solder by NH3-decomposed species, that is, by hydro-
in the Pb-free solder powder decreased after the solder
gen radicals, was investigated.
was subjected to the cleaning treatment. This result
The results of this study can be summarized as follows.
implies that hydrogen radicals can reduce Sn oxide pres-
(1)
tion Sn oxide.
ent in the Pb-free solder.
Figure 4 shows the variance of the particle size of the
Pb-free solder powder with NH3 treatment time. The vari-
The hydrogen radicals are effective for the reduc-
(2)
Sn oxide in the Sn substrate and Pb free solder are
reduced, and their re-oxidation was also sup-
3
Transactions of The Japan Institute of Electronics Packaging
Vol. 2, No. 1, 2009
pressed by this treatment.
(3)
The effect of the substrate temperature on the activation energy is found to be small.
(4)
The particle size of the Pb-free solder material can
be controlled by this process.
[3] R. K. Shiue, L. W. Tsay, C. L. Lin and J. L. Ou, Microelectronics Reliability, 43, pp. 453–463 (2003).
[4] Kyung-Seob Kim, Chung-Hee Yu and Jun-Mo Yang,
Thin Solid Films, 504, pp. 350–354 (2006).
[5] Naoe Hosoda and Tadatomo Suga, Applied Surface
Science, 227, pp. 81–86 (2004).
[6] R. Deltschew, D. Hirsch, H. Neumann, T. Herzog, K.
Acknowledgements
We would like to thank Harima Chemicals, Inc. for providing the Pb-free solder powder.
J. Wolter, M. Nowottnic and K. Wittke, Surface and
Coatings Technology, 142–144, pp. 803–807 (2001).
[7] Akira Izumi, Tomoya Ueno, Yasuo Miyazaki, Hiroaki
References
[1] Ning Bai, Xu Chen and Hong Gao, Materials and
Design, 30, pp. 122–128 (2009).
[2] Tetsu Ichitsubo, Eiichiro Matsubara, Kozo Fujiwara,
Masahiko Yamaguchi, Hisao Irie, Seishi Kumamoto
and Takaaki Anada, Journal of Alloys and Compounds,
392, pp. 200–205 (2005).
4
Oizumi and Iwao Nishiyama, Thin Solid Films, 516,
pp. 853–855 (2008).
[8] I. Nishiyama, H. Oizumi, K. Motai, A. Izumi, T. Ueno,
H. Akiyama and A. Namiki, J. Vacuum Sci. & Tech B,
23, pp. 3129–3131 (2005).
[9] A. Izumi and H. Matsumura, Jpn. J. Appl. Phys., 41,
pp. 4639–4641 (2002).