Journal of Alloys and Compounds 417 (2006) 63–68
Phase transformations in a Cu–35 at.% Mn–25 at.% Al alloy
S.Y. Yang, T.F. Liu
∗Department of Materials Science and Engineering, National Chiao Tung University, Hsinchu 300, Taiwan, ROC Received 10 August 2005; accepted 1 September 2005
Available online 13 October 2005
Abstract
The phase transformations in the Cu–35 at.% Mn–25 at.% Al alloy have been investigated by means of transmission electron microscopy (TEM) and energy-dispersive X-ray spectrometry (EDS). In as-quenched condition, the microstructure of the alloy was a mixture of (L21+ B2 + L-J)
phases. This is different from that observed by previous workers in the Cu3−xMnxAl alloys with x≤ 1.0. When the as-quenched alloy was aged at 460◦C for short times,␥-brass precipitates started to occur at anti-phase boundaries (APBs). After prolonged aging at 460◦C, the␥-brass precipitates grew and-Mn precipitates were formed at the regions contiguous to the ␥-brass precipitates. The coexistence of (␥-brass + -Mn) has never been observed by previous workers in Cu–Mn–Al alloy systems before.
© 2005 Elsevier B.V. All rights reserved.
Keywords: Cu–Mn–Al alloy;␥-Brass; Anti-phase boundary; -Mn
1. Introduction
In previous studies, it is seen that when the Cu3−xMnxAl
alloys with 0.5 x 0.8 were solution-treated in single  phase (disordered body-centered cubic) region and then quenched rapidly, a → B2 → D03+ L21phase transition occurred
dur-ing quenchdur-ing[1]; as the Mn content in the Cu3−xMnxAl alloy
was increased to 25 at.% (x = 1), the as-quenched microstructure of the Cu2MnAl alloy became a single L21phase[1–4]. When
the Cu3−xMnxAl alloys with 0.5 x 0.8 were aged at 300◦C
or below for longer times, fine precipitates were observed to appear within the (D03+ L21) matrix[1]. The crystal structure
of the fine precipitates was determined to be of L10having
lat-tice parameters a = 0.424 nm, b = 0.297 nm and c = 0.424 nm[1]. In addition, three kinds of precipitates, namely,␥-brass (D83),
-Mn (A13) and T-Cu3Mn2Al (C15) were reported to form in
the Cu2MnAl alloy after being aged at temperatures ranging
from 350 to 650◦C[2–4]. It is interesting to note that although
the-Mn precipitate was always found in the aged Cu2MnAl
alloy, we are aware of only one article concerning the orien-tation relationship between the-Mn and matrix[4]. In 1987,
Kozubski et al. reported that both the morphology of the
-Mn precipitates and the orientation relationship between the
∗Corresponding author. Tel.: +886 3 573 1675; fax: +886 3 572 8504.
E-mail address: tfliu@cc.nctu.edu.tw (T.F. Liu).
-Mn and L21matrix would vary with the aging temperature
[4].
Recently, we have performed TEM investigations on the phase transformations of Cu2.2Mn0.8Al and Cu2MnAl alloys
[5,6]. Consequently, we found that the fine precipitates formed in the Cu2.2Mn0.8Al alloy aged at 300◦C should belong to
the L-J phase, rather than L10 phase [5]. The L-J phase has
an orthorhombic structure with lattice parameters a = 0.413 nm,
b = 0.254 nm and c = 0.728 nm, which was firstly identified by
the present workers. In addition, TEM examinations indicated that when the Cu2MnAl alloy was aged at temperatures ranging
from 460 to 560◦C, the morphology of the-Mn precipitates
would change with the different aging temperature; however, in spite of the morphology change the same orientation
rela-tionship between the-Mn and the L21matrix was maintained.
This result is different from that reported by Kozubski et al.
[4]. However, to date, all of the examinations were focused
on the Cu3−xMnxAl alloys with x 1.0. Little information was
available concerning the microstructural developments of the Cu3−xMnxAl alloys containing higher Mn content. Therefore,
the purpose of this work is an attempt to study the phase trans-formations in the Cu–35 at.% Mn–25 at.% Al alloy.
2. Experimental procedure
The alloy, Cu–35 at.% Mn–25 at.% Al, was prepared in a vacuum induc-tion furnace by using 99.9% Cu, 99.9% Mn and 99.9% Al. The melt was chill
0925-8388/$ – see front matter © 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.jallcom.2005.09.021
Fig. 1. Electron micrographs of the as-quenched alloy: (a) BF; (b) and (c) two SADPs. The zone axes of the L21phase are (b) [1 0 0] and (c) [1 1 0], respectively
(h k l = L21, h k l1,2= L-J phase, 1: variant 1; 2: variant 2); (d) and (e) (¯1 1 1) and (0 0 2) L21DF, respectively; (f) (0 ¯2 01) L-J DF.
cast into a 30 mm× 50 mm × 200 mm copper mold. After being homogenized at 900◦C for 72 h, the ingot was sectioned into 2.0-mm thick slices. These slices were subsequently solution-treated at 850◦C for 1 h and then quenched into room-temperature water rapidly. The aging process was performed at 460◦C for various times in a vacuum heat-treated furnace and then quenched rapidly.
TEM specimens were prepared by means of a double-jet electropolisher with an electrolyte of 70% methanol and 30% nitric acid. The polishing tem-perature was kept in the range from−30 to −15◦C, and the current density was kept in the range from 3.0× 104to 4.0× 104A/m2. Electron microscopy
was performed on a JEOL JEM-2000FX scanning transmission electron
micro-scope operating at 200 kV. This micromicro-scope was equipped with a Link ISIS 300 energy-dispersive X-ray spectrometer (EDS) for chemical analysis. Quantitative analyses of elemental concentrations for Cu, Mn and Al were made with the aid of a Cliff-Lorimer Ratio Thin Section method.
3. Results
Fig. 1(a) shows a bright-field (BF) electron micrograph of the as-quenched alloy.Fig. 1(b) and (c) are two selected-area
diffrac-S.Y. Yang, T.F. Liu / Journal of Alloys and Compounds 417 (2006) 63–68 65
tion patterns (SADPs) of the as-quenched alloy. When compared with our previous studies in the Cu2.2Mn0.8Al and Cu2MnAl
alloys[5,6], it is found in these SADPs that the brighter and well-arranged reflection spots are of the ordered L21phase and
the extra spots with streaks are of the L-J phase with two variants. Although the brighter and well-arranged reflection spots could be analyzed as a single L21phase, the L21reciprocal lattices
contain all the B2-type reflections[7,8]. Therefore, in order to decide whether the ordered B2-type phase coexists with the L21
phase, both electron diffraction method and dark-field technique were performed. In our previous study[6], it was found that the intensity of the (¯1 1 1) and (0 0 2) reflection spots of a single L21phase should be almost equivalent. However, it is clearly
seen inFig. 1(c) that the (0 0 2) and (¯2 2 2) reflection spots are much stronger than the (¯1 1 1) reflection spot. Therefore, it is strongly suggested that the (0 0 2) and (¯2 2 2) reflection spots
should derive from not only L21 phase but also the B2 phase,
since the (¯1 1 1) reflection spot comes from the L21phase only;
while the (0 0 2) and (¯2 2 2) reflection spots can come from both the L21 and B2 phases (the (0 0 2) and (¯2 2 2) L21 reflection
spots are equal to the (0 0 1) and (¯1 1 1) B2 reflection spots, respectively).Fig. 1(d) and (e) are (¯1 1 1) and (0 0 2) L21
dark-field (DF) electron micrographs of the as-quenched alloy. It is obviously seen that the bright region in the (0 0 2) DF image is much more than that in the (¯1 1 1) DF image. This demonstrates that both B2 and L21phases are present, rather than single L21
phase; otherwise these two DF images should be morphologi-cally identical.Fig. 1(f) is a (0 ¯2 01) L-J DF electron micrograph,
revealing the presence of fine L-J precipitates. Accordingly, it is concluded that the microstructure of the alloy in the as-quenched condition was a mixture of (L21+ B2 + L-J) phases.
When the as-quenched alloy was aged at 460◦C for less than 10 min, the sizes of both the B2 and L-J phases existing within
the L21 matrix increased and the microstructure of the alloy
was still the mixture of (L21+ B2 + L-J) phases. An example is
shown inFig. 2. However, after prolonged aging at the same
temperature, heterogeneous precipitation started to occur at the APBs, as illustrated inFig. 3(a).Fig. 3(b) is an SADP taken from an area including the precipitate marked as “R” inFig. 3(a) and its surrounding matrix. Based on the analyses of the diffrac-tion pattern, it is confirmed that the heterogeneously precipitated phase isbrass and the orientation relationship between the ␥-brass and the L21matrix was determined to be cubic to cubic.
This is similar to that reported by previous workers in the aged
Cu2MnAl alloy [4]. With continued aging at 460◦C, the
␥-brass precipitates grew and another type of precipitates started to occur at the regions contiguous to the␥-brass precipitates, as shown inFig. 4(a).Fig. 4(b), an SADP taken from the precip-itate marked as “B” inFig. 4(a), indicates that the new type of precipitate was-Mn with lattice parameter a = 0.641 nm[6].
Fig. 4(c) is an SADP taken from an area covering two pre-cipitates marked as “R” and “B” in Fig. 4(a), indicating that
the orientation relationship between the␥-brass and -Mn was
(0 0 1)␥-brass//(0 1 2)-Mnand (0 1 1)␥-brass//(0 3 1)-Mn. With the subsequent aging at 460◦C, the precipitation of (␥-brass + -Mn) would tend toward the inside of the L21matrix, as illustrated
inFig. 5. It is thus anticipated that the microstructure of the alloy
Fig. 2. Electron micrographs of the alloy aged at 460◦C for 10 min: (a) (0 0 2) L21DF; (b) (0 ¯2 01) L-J DF.
in the equilibrium stage at 460◦C was a mixture of (␥-brass + -Mn).
4. Discussion
That the B2 phase could be detected in the as-quenched or aged at 460◦C alloy is a remarkable feature in the present study. This result is different from that examined by previous workers in the Cu3−xMnxAl alloys with 0.5 x 1.0 [1–4],
in which they reported that the as-quenched microstructure of the Cu3−xMnxAl alloys with 0.5 x 0.8 was the (D03+ L21)
phases, and that of the Cu2MnAl alloy was the L21phase; and
the B2 phase could exist only at temperatures above 600◦C.
Compared to the previous studies[1–4], it is clear that besides containing higher Mn content, the chemical composition of the present alloy is similar to that of the Cu3−xMnxAl alloys with
0.5 x 1.0. Therefore, it is reasonable to expect that the addi-tion of the higher Mn content in the Cu3−xMnxAl alloys would
pronouncedly enhance the formation of the B2 phase. However, the reason why the higher addition of Mn could lead to this result is unclear.
A second important feature of the present study is that when
the alloy was aged at 460◦C for moderate times, the -Mn
precipitates started to occur at the regions contiguous to the ␥-brass precipitates. This precipitation behavior has never been
Fig. 3. Electron micrographs of the alloy aged at 460◦C for 30 min: (a) (002) L21DF; (b) an SADP. The zone axis of the L21phase is [1 0 0] (h k l = L21,
h k l =␥-brass).
observed by previous workers in the aged Cu2MnAl alloy[3,4],
in which they found that when the Cu2MnAl alloy was aged at
temperatures ranging from 350 to 650◦C, the␥-brass and -Mn precipitates were formed separately at the grain boundaries or on other structural defects. In order to clarify this difference,
an STEM-EDS study was undertaken.Fig. 6(a)–(c) represents
three typical EDS spectra taken from the as-quenched alloy and the␥-brass as well as the -Mn precipitates in the alloy aged at 460◦C for 6 h, respectively. The average concentrations of alloy-ing elements obtained by analyzalloy-ing a number of EDS spectra of each phase are listed inTable 1. It is clearly seen inTable 1that the concentration of Mn in the␥-brass is only about 2.25 at.%, which is much less than that in the as-quenched alloy. It is thus expected that along with the growth of the␥-brass precipitates,
Table 1
Chemical compositions of the phases revealed by energy-dispersive X-ray spec-trometer (EDS)
Heat treatment Phase Chemical compositions (at.%)
Cu Mn Al
As-quenched 39.81 35.11 25.08
460◦C, 6 h ␥-Brass 67.63 2.25 30.12 460◦C, 6 h -Mn 11.99 67.97 20.04
Fig. 4. Electron micrographs of the alloy aged at 460◦C for 6 h: (a) BF; (b) an SADP. The zone axis of the-Mn is [0 0 1]. (c) An SADP. The zone axes of the-Mn and the ␥-brass is [1 0 0] and [1 0 0], respectively (h k l = -Mn, h k l =␥-brass).
the surrounding regions would be enriched in Mn. In Cu–Mn phase diagram[9], it is clearly seen that the-Mn phase could exist only when the Mn content was greater than 75 at.% and the temperature was in the range from 707 to 1100◦C; whereas the
-Mn phase region was pronouncedly expanded to below 427◦C
with 61≤ Mn ≤ 90 at.% and 10 ≤ Al ≤ 39 at.% in Al–Mn binary
alloys[10]. Therefore, it is reasonable to propose that at lower temperature, the concentrations of both Al and Mn would be the predominant factor for the formation of the-Mn precipitates.
S.Y. Yang, T.F. Liu / Journal of Alloys and Compounds 417 (2006) 63–68 67
Fig. 5. BF electron micrograph of the alloy aged at 460◦C for 12 h.
InTable 1, it is obvious that the concentrations of both Al and Mn in the-Mn precipitate are located within the composition range
of the-Mn phase region in the Al–Mn binary alloys.
There-fore, the coexistence of (␥-brass + -Mn) is expected to occur. In contrast to the observations in the present alloy, although the Mn-lack␥-brass precipitates were also observed to occur in
Fig. 6. Three typical EDS spectra obtained from (a) as-quenched alloy, (b) a␥-brass precipitate as well as (c) a -Mn precipitate in the alloy aged at 460◦C for 6 h.
the aged Cu2MnAl alloy, no evidence of the-Mn precipitates
could be detected at the regions contiguous to the␥-brass pre-cipitates[4,6]. The reason is probably that along with the growth of␥-brass precipitates, the Mn concentration at the regions sur-rounding the␥-brass may not be sufficient to cause the formation of the-Mn precipitates.
Finally, it is worthwhile to note that during the early stage
of isothermal aging at 460◦C, the ␥-brass precipitates have
occurred preferentially at APBs. This feature is similar to that observed by other workers in an aged Cu–14.6 wt.% Al–1.95 wt.% Ni alloy[11].
5. Conclusion
The as-quenched microstructure of the Cu–35 at.%
Mn–25 at.% Al alloy was a mixture of (L21+ B2 + L-J) phases.
When the as-quenched alloy was aged at 460◦C for moderate
times,-Mn precipitates were formed at the regions contiguous to the␥-brass precipitates. The orientation relationship between
the ␥-brass and -Mn was (0 0 1)␥-brass//(0 1 2)-Mn and
(0 1 1)␥-brass//(0 3 1)-Mn. The coexistence of (␥-brass + -Mn) has never been observed by previous workers in Cu–Mn–Al alloy systems before.
Acknowledgments
The authors are pleased to acknowledge the financial support of this research by the National Science Council, Republic of China under Grant NSC93-2216-E009-016. He is also grateful to M.H. Lin for typing.
References
[1] M. Bouchard, G. Thomas, Acta Metall. 23 (1975) 1485–1500. [2] R. Kainuma, N. Satoh, X.J. Liu, I. Ohnuma, K. Ishida, J. Alloys Compd.
266 (1998) 191–200.
[3] R. Kozubski, J. Soltys, J. Mater. Sci. 18 (1983) 1689–1697.
[4] R. Kozubski, J. Soltys, J. Dutkiewicz, J. Morgiel, J. Mater. Sci. 22 (1987) 3843–3846.
[5] S.C. Jeng, T.F. Liu, Metall. Mater. Trans. A 26 (1995) 1353–1365. [6] K.C. Chu, T.F. Liu, Metall. Mater. Trans. A 30 (1999) 1705–1716. [7] T.F. Liu, G.C. Uen, C.Y. Chao, Y.L. Lin, C.C. Wu, Metall. Trans. A 22
(1991) 1407–1415.
[8] C.C. Wu, J.S. Chou, T.F. Liu, Metall. Trans. A 22 (1991) 2265–2276. [9] F.A. Shunk, in: Thaddeus B. Massalski (Ed.), Binary Alloy Phase
Dia-grams, American Society for Metals, Metals Park, Ohio, 1990, pp. 1435–1436.
[10] X.J. Liu, I. Ohnuma, R. Kainuma, K. Ishida, J. Phase Equilib. 20 (1999) 45–56.