Materials Transactions, Vol. 57, No. 5 (2016) pp. 624 to 626
©2016 The Japan Institute of Metals and Materials
Effect of Aging and Stress on Stability of β Phase in Au-Cd-Ag
Yuki Matsuoka1,*1 and Mao Fujita2,*2
1
Faculty, Nara Women s University, Nara 630–8506, Japan
Graduate School of Humanities and Science, Nara Women s University, Nara 630–8506, Japan
2
Au52.5-xCd47.5Agx has a single β phase in the composition range of x ≤ 41 and a wide temperature range extending to room temperature or
below. Coexistence of the β and α phases has been observed for 41 < x < 47, and the coexistence of the β, α, and ζ phases has been observed for
x ≥ 47, as thermal equilibrium states. We found that the α phase appears and coexists with the β phase for x ≤ 41 in the aged alloy at room temperature. The mole fraction and the stability of the α phase depend on the atomic order of the β phase. Moreover, it was found that the ζ phase
is stress induced in the composition range of x ≤ 41. The ζ phase disappears with heat treatment to release strain, and so this ζ phase is not stable.
[doi:10.2320/matertrans.MB201514]
(Received October 1, 2015; Accepted February 5, 2016; Published April 25, 2016)
Keywords: beta phase, alpha phase, zeta phase, aging, stress induction
1. Introduction
Au52.5-xCd47.5Agx is a well-known martensitic alloy.1–4)
The β phase of this alloy exists as a single phase over a wide
composition range. Although it shows a complex composition dependence, the martensitic transformation temperature
T0 decreases as the Ag composition increases. The atomic order of the β phase changes as the Ag concentration changes:
as the Ag content increases, the atomic order of the β phase
changes as B2 (CsCl) → L21 (Heusler) → B2. The β phase
transforms to M2H martensite.4) We have studied the atomic
order and stability of the β phase, which exists in the range of
x ≤ 41, of this noble metal alloy.4) Furthermore, T0 for the
M2H transformation decreases and shows a complicated Ag
composition dependence; the temperature range gradually
narrows as the Ag content increases. It is supposed that the β
phase has good stability at room temperature because T0 is
lower than 240 K in the region of 15 ≤ x ≤ 41.
For x ≥ 43, the α phase coexists with the β phase, and the
temperature range of the coexistence phase extends to room
temperature or lower.5,6)
For x ≥ 47, another phase appears in which the β, α, and ζ
coexist, and in this region, T0 increases.7) There is some kind
of relationship between the phase stability and the single
phase/triple phase.
It is well known that aging stabilizes the β phase of the AgCd alloy, and so it is expected to destabilize the coexisting
phases.
In this work, we studied the effects of aging and stress on
the stability of the β phase and other phases that are observed
in the region of β single phase to define the phase diagram
of Au52.5-xCd47.5Agx.
Table 1 Composition of sample alloys (at %) and observed crystal structure
at room temperature before aging.4)
Sample name
Au
Cd
Ag
16Ag
36.5
47.5
16.0
18Ag
34.5
47.5
18.0
23Ag
29.5
47.5
23.0
26Ag
26.5
47.5
26.0
38Ag
14.5
47.5
38.0
41Ag
11.5
47.5
41.0
Crystal structure of β-phase
B2
L21
B2
773 K for 86400 s, the obtained ingots were cut into pieces.
Some pieces were used for electric resistivity measurements
to determine T0, while other pieces were filed and sieved to
less than 53-μm-diameter in size for X-ray diffraction measurements to check the crystal structure and atomic order. The
composition of the sample alloys and crystal structure at
room temperature are shown in Table 1.4) The remaining
pieces were aged for 3 × 108 s at room temperature and then
three kinds of samples were prepared: a) bulk samples (18Ag,
23Ag, 26Ag, 38Ag), b) powder samples with no heat treatment, c) powder samples that were heat treated at 773 K for
600 s in an evacuated quartz tube and slowly cooled to release
strain.
All X-ray diffraction measurements were performed at
room temperature over angles of 20 to 100 in 2θ space with
Cu Kα1,2 (λ = 0.1542 nm) radiation using a PANalytical EXpert PRO. Profiles of the powder diffraction patterns were
precisely analyzed with Rietan to determine the atomic sites
and mole fraction of the phases.8)
3. Results
2. Experimental
Samples with 6 different compositions were prepared. The
sample alloys were prepared by melting the component elements in evacuated quartz tubes. After homogenizing at
*1
Corresponding author, E-mail: [email protected]
Graduate student, Nara Women s University. Present address: Himeji Higashi Senior High School, Hyogo 670–0012, Japan
*2
Profiles of the X-ray diffraction patterns of the a), b), and
c) samples are shown in Fig. 1, 2, and 3, respectively. Coexistence of the β and the α phases was observed for all compositions for samples a) and c). All three phases (α,β,δ)were observed for all compositions of sample b). The experimental
values of the lattice parameter, atomic order, and atomic site
of the β phase in the heat-treated powder samples obtained
from the Rietan analysis were consistent with the former re-
625
Effect of Aging and Stress on Stability of β Phase in Au-Cd-Ag
Fig. 1 Profiles of the X-ray diffraction patterns of the aged bulk samples.
The peak indicated by the solid circle is from the sample holder.
Fig. 4 Lattice constants of the ζ phase of the non-heat-treated powder samples.
Table 2 Mole fraction of each phase with and without heat treatment after
aging.
Sample name
Without heat treatment
(No stress released)
With heat treatment
(Stress released)
16Ag
β:α:ζ = 53:30:24
β:α = 97:3
18Ag
β:α:ζ = 85:6:9
β (L21), α*
23Ag
β:α:ζ = 86:6:8
β (L21), α*
26Ag
β:α:ζ = 75:13:13
β (L21), α*
38Ag
β:α:ζ = 54:25:21
β (B2)
β:α = 85:15
41Ag
β:α:ζ = 39:33:28
*
in these cases, the amount is too small to allow estimation of the mole
fraction
Fig. 2 Profiles of the X-ray powder diffraction patterns of the aged samples. These samples were not heat treated after filing.
Fig. 5 Mole fraction of the β, α, and ζ phases in the non-heat-treated and
heat-treated powder samples.
Fig. 3 Profiles of the X-ray powder diffraction patterns of the aged samples. These samples were heat treated for stress release after filing.
port.5) The crystal structure of the α phase was A2. The atomic order and lattice parameters of the α phase were also consistent with the former report.5) The crystal structure of the ζ
phase was hcp. The atomic order of the ζ phase in the nonheat-treated samples was not determined because the peak of
the ζ phase was widened by the stress, and a good enough
value of R factor could not be obtained. The ζ phase of the
near equi-atomic Ag-Cd alloy is A3. The ζ phase observed in
Au52.5-xCd47.5Agx is estimated to be on the same atomic order
as near equi-atomic Ag-Cd.9) The obtained lattice parameters
of the ζ phase are shown in Fig. 4.
The mole fraction of the three phases obtained by the present analysis are listed in Table 2 and shown in Fig. 5.
626
Y. Matsuoka and M. Fujita
4. Discussion
Both the α and ζ phases were observed in the aged Au-CdAg alloys in the composition range that has been regarded as
the β-single-phase (Table 1). These two phases do not exist as
the thermal equilibrium state in the composition range of x ≤ 41.7) The atomic order of the β phase in the aged samples was
the same as that before aging as shown in Table 1 for all compositions. The β phase of the 18Ag, 23Ag, and 26Ag samples
has a highly ordered L21 structure. Considering the results in
Table 2 and Fig. 5, it is estimated that the higher atomic order
makes the β phase more stable against aging and stress. The
ratio of Au to Ag is close to 1:1 in the 26Ag sample compared
with that in the 18Ag and 23Ag samples. Therefore, the 26Ag
sample has the highest ordered structure of all the alloy samples. The stability of the β phase is sensitive to the Au - Ag
equality; compositional differences between Au and Ag make
the phase transition easier and recovery to the β phase harder.
In all compositions of the aged powder samples without
heat treatment after filing, three phases were observed. The
mole fraction of the α phase, which was estimated from the
main peak intensity (α111), is different from that of the bulk
samples. The ζ phase appears in the powder samples. Stress
induces the phase transition from the β phase to the α or ζ
phase.
In the case of the heat-treated aged powder samples, the α
phase was observed. The peak intensities of the α phase in the
18Ag, 23Ag, and 26Ag samples are too weak to allow estimation of its mole fraction. The ζ phase was not observed in the
samples prepared under these conditions. The ζ phase is easily induced by stress and easily reduced by heating. Therefore,
this phase is unstable in the range of x ≤ 41. The α phase is
more stable than the ζ phase after heat treatment at 773 K.
The ζ phase exists in the thermal equilibrium state in the composition range of x ≥ 47, which is further from the composition of the equivalent Au-Ag ratio than that in the range over
which the coexisting phase of the β and the α exists (41 < x < 47).
A similar stress-induced phase transition from the β phase
to the ζ phase has also been observed in Ag-Zn alloy.10,11)
Therefore, Ag is supposed to play an important role in determining the transition phase induced by stress.
5. Conclusions
The α phase coexists with the β phase in aged
Au52.5-xCu47.5Agx alloys for x ≤ 41 as the thermal equilibrium
state. The α and ζ phases appear on adding stress. The mole
fractions of these two phases increase with an increase in the
deviation from equivalent Au and Ag. The ζ phase disappears
when stress is released, while the α phase remains. The ζ
phase was unstable in the range of x ≤ 41. On the other hand,
the α phase is stable in this region. It is speculated that Ag
plays an important role in the different phase transition processes in the case of heating or stress addition because a similar phenomenon was also observed in Ag-Zn.
REFERENCES
1) F. Rothwarf and L. Muldawer: J. Appl. Phys. 33 (1962) 2531–2538.
2) N. Nakanishi, M. Takano, H. Morimoto, S. Miura and F. Hori: Scr.
Metall. 12 (1978) 79–83.
3) T. Suzuki and S. Nakamoto: Mater. Sci. Eng. A 273–275 (1999) 549–
553.
4) Y. Matsuoka, A. Nagahara and T. Suzuki: Proc. Symposia of IUMRSICAM, 2003 (Trans. MRJ-J 29 2004) pp. 3025–3028.
5) Y. Matsuoka, M. Fujita and A. Nagahara: Mater. Today Proc. 2S (2015)
S573–S576.
6) Y. Matsuoka and N. Kudo: E-MRS 2005 Fall Meeting-Book of Abstracts, (European Materials Research Society, 2005) pp. 67.
7) K. Sakamoto and Y. Matsuoka: Abstr. Int. Conf. on Martensitic Transformations 2014, (ICOMAT 2014) p. 103.
8) F. Izumi and K. Momma: Solid State Phenom. 130 (2007) 15–20.
9) A. Tonejc, A.M. Tonejc and A. Bonefačič: Phys. Lett. A 49 (1974)
145–146.
10) B. Magyari-Köpe, L. Vitos and G. Grimvall: Phys. Rev. B 70 (2004)
052102.
11) Y. Matsuo, K. Ohshima, H. Iwasaki, Y. Kuroiwa, H. Maeta and K. Haruna: J. Phys. F Met. Phys. 18 (1988) 2505–2512.