Characterization of Ni63Cu9Fe8P20 glass forming alloy

Journal of Alloys and Compounds 387 (2005) 179–182

Characterization of Ni63Cu9Fe8P20 glass forming alloy Krzysztof Ziewiec∗ Institute of Technology, Pedagogical University, ul. Podchorazych 2, Krakow 30-084, Poland Received 20 April 2004; received in revised form 9 June 2004; accepted 9 June 2004

Abstract Quaternary Ni63 Cu9 Fe8 P20 alloy was prepared using 99.95 wt% Ni, 99.95 wt% Cu, 99.95 wt% Fe and Ni–P master alloy. The precursors were induction melted in quartz tubes under vacuum (10−2 bar). The microstructure of the alloy was investigated by the use of a light microscope, scanning electron microscope (SEM) with energy dispersive spectroscope (EDS). Analysis of morphology was carried out and volume fraction of microstructural constituents in as-cast state was assessed. Melt spinning was used to produce ribbons from the alloy. The ribbons were then investigated by means of differential scanning calorimetry (DSC) and differential thermal analysis (DTA). High temperature X-ray studies of melt-spun ribbon as a function of temperature were carried in order to characterize thermal stability of the alloy. The results were discussed and compared with other experimental data. © 2004 Elsevier B.V. All rights reserved. Keywords: Metallic glass; Glass forming ability; Thermal stability; Melt spinning

1. Introduction Metallic glassy alloys have been developed recently in several multi-component systems [1–3]. The amorphous metallic alloys have interesting properties, e.g. high elasticity limit and low coercivity. These characteristics, which can be hardly found in crystalline materials, are attractive for practical uses of structural and functional materials. These features are associated with an amorphous disordered atomic structure [4]. In the past several years, a large number of alloys with a good glass forming ability have been discovered, i.e.: Zr41 Ti14 Cu12.5 Ni10 Be22.5 [5], Pd40 Ni10 Cu30 P20 [6], La55 Al25 Ni5 Cu10 Co5 [7], Nd60 Fe20 Al10 Co10 [8], Zr57 Nb5 Cu15.4 Ni12.6 Al10 [9], Zr48 Nb8 Fe8 Cu12 Be24 [10]. However, due to limited resources and high prices of such constituents as Pd, La, Nd and Zr, applications of metallic glasses with high glass forming ability are still very restricted. Therefore, extensive use of good metallic glass formers lies probably behind cheaper precursors and more common elements. On the other hand, analysis of available binary and ∗

Tel.: +48 607 237 425. E-mail address: [email protected] (K. Ziewiec).

0925-8388/$ – see front matter © 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.jallcom.2004.06.040

ternary phase diagrams containing Ni, Pd, Cu and P indicates that there are deep eutectics [11], which show a good glass forming ability, especially in compositions where one of the constituents is P. This feature in connection with sufficiently large difference of atomic diameters indicates that there is a chance to obtain the compositions with a large glass forming ability. This expectation is confirmed by the studies of Pd40 Ni40 P20 alloy [12,13] presenting a critical cooling rate as low as 0.16 K/s as well as the Pd40 Ni10 Cu30 P20 alloy [14] with a critical cooling rate of 0.1 K/s. Glass forming ability could be even further improved by adjusting the Pd composition with respect to Cu [15,16]. The Pd43 Ni10 Cu27 P20 alloy [16] has the lowest critical cooling rate of all BMG’s discovered so far. Schroers and Johnson [16] found that even a cooling rate as low as 0.005 K/s there are 10–15% particles that still do not crystallize. With regards to glass forming ability (GFA) the most important parameter is Trg . Trg is defined most frequently as Tg /Tm or Tg /Tl ratios (Tg , glass transition temperature; Tm , onset melting point; Tl , liquidus temperature) and it is generally considered that the higher Trg , the lower the critical cooling rate for a given composition [17]. However, recently Lu and Liu [18] have introduced a new glass forming ability criterion γ = Tx /(Tg + Tl ), which should also be

180

K. Ziewiec / Journal of Alloys and Compounds 387 (2005) 179–182

considered because it seems to have even better correlation with GFA than Trg [18]. The thermal stability Tx is defined as Tx − Tg , (interval between glass transition temperature Tg and crystallization temperature Tx ) [3,4,6,19]. Although Tx is not directly related to the GFA [20,21], alloys having Tx larger than 50 K are reported as good glass formers and they are able to be cast into fully amorphous rods which are 2 mm in diameter [20]. On the other hand, it is worth to note that a low critical cooling rate does not necessarily imply a high thermal stability in bulk glass-forming compositions [21]. It is considered that large Tx indicates high resistance of the liquid alloy to crystallization. That is why the existence of large Tx values is of scientific and technological interest. This gives a chance to create a new group of structural and functional materials that can be processed by means of plastic deformation with the use of viscous flow at temperatures above Tg and below Tx [12,13]. It is known that in the Ni–Cu–P and Fe–Cu–P systems regions with deep eutectics exist [22] as it is the case of the Pd–Ni–P and Pd–Ni–Cu–P systems, and probably the Ni–Cu–Fe–P system is a good potential candidate for finding compositions of a good glass forming ability. Therefore, this work is focused on the regions close to the eutectic in Ni–Cu–Fe–P system. The aim of this work was to study thermal stability of the quaternary Ni63 Cu9 Fe8 P20 near-eutectic alloy and finding the main indicators of GFA.

10−2 bar vacuum the alloy in capsules was quenched in water. Then the ingot was cut for metallographic observations and EDS analysis. The quantity of structural constituents was assessed using pixel analysis of a microscopic picture. The following stage was melt spinning of the alloys using a wheel with linear speed of 22 m/s (approximate cooling rate of 105 K/s). The melt spun ribbons were then investigated by means of differential scanning calorimetry (NETZSCH STA 409 PC/PG) differential thermal analysis (DTA-STD 2960 TA Instruments). DSC measurements were applied to investigate solid state transformations while DTA was focused on determining the melting point. Both DSC and DTA runs were made at a heating rate of 20 K/min. Then X-ray high temperature studies of melt spun ribbons as a function of temperature were carried out using a Siemens D5005 (Bruker-AXS) diffractometer equipped with a Cu lamp with 30 mA and 40 kV power supply. The radiation was monochromatized with graphite monochromator. As-cast ribbons of 20 ␮m were stuck to a glass plate. In some measurements high purity silicon placed on the plate was used as an external standard to ensure the correct sample position during the operation. High temperature measurements were made at 296, 373, 473, 623, 673, 773 and 873 K in XRK-900 reaction chamber (Anton Paar). Diffraction patterns were registered in the range of 30–60◦ (2θ) using θ–θ scan with 0.04◦ /1 s speed. During heating of the samples with an average heating rate of 10 ◦ C/min selected diffraction peaks were monitored.

2. Experimental 3. Results Quaternary Ni63 Cu9 Fe8 P20 alloy was prepared using 99.95 wt% Ni, 99.95 wt% Cu, 99.95 wt% Fe and Ni–P master alloy. The precursors were induction melted in quartz tubes under vacuum (10−2 bar). After melting at 1050 ◦ C in

The results of EDS analysis are shown in Fig. 1. The microstructure of the alloy in as-cast state contains primary faceted crystals and an eutectic constituent (Fig. 1).

Fig. 1. SEM microstructure of Ni63 Cu9 Fe8 P20 alloy quenched in water. Regions A and B analysed with EDS; (A) composition of faceted primary crystals Ni: 65.9 ± 3.1 at.%, Cu: 2.6 ± 1.6 at.%, Fe: 7.9 ± 0.7 at.%, P: 23.6 ± 0.8 at.%; (B) composition of eutectic Ni: 68.5 ± 3.4 at.%, Cu: 2.9 ± 1.2 at.%, Fe: 7.3 ± 1.2 at.%, P: 21.3% ± 0.8 at.%.

K. Ziewiec / Journal of Alloys and Compounds 387 (2005) 179–182

181

Table 1 Temperatures of glass transition, crystallization, melting point, liquidus and different indicators of GFA derived from the DSC and DTA data Ni63 Cu9 Fe8 P20 alloy (Fig. 3) Tg [K] Tx [K] Tm [K] Tl [K] Tx − T g Tg /Tl Tg /Tm Tx /(Tg + Tl )

598 648 1149 1181 50 0.506 0.521 0.364

Fig. 2. Optical microstructure of Ni63 Cu9 Fe8 P20 alloy quenched in water.

The optical microstructure of the Ni63 Cu9 Fe8 P20 alloy is shown on Fig. 2. The results of metallographic study show that faceted primary crystals cover 24 ± 4% of the volume fraction while the eutectic is the major microstructural constituent −76 ± 4%. The results of DSC and DTA obtained from the ribbon produced by melt spinning technique are presented in Fig. 3a and b, respectively. The data derived from DSC and DTA as well as the indicators of GFA are presented in Table 1. The results of X-ray investigations are shown in Fig. 4. At room temperature Ni63 Cu9 Fe8 P20 alloy shows a broad

Fig. 4. X-ray patterns at different temperatures for Ni63 Cu9 Fe8 P20 alloy.

peak between 39 and 51◦ . It seems that during heating the Ni63 Cu9 Fe8 P20 alloy to 623 K the X-ray patterns do not change significantly. At 673 K the broad peak splits into peaks that can be attributed to the Ni(Cu, Fe) solid solution and (Ni, Cu, Fe)3 P [23]. One can observe that at higher temperatures further changes of X-ray patterns are rather small. However, at 873 K the peaks from crystalline phases mentioned above are more distinct than at lower temperatures.

4. Discussion

Fig. 3. Thermal characteristics of the Ni63 Cu9 Fe8 P20 alloy. (a) DSC curve: glass transition temperature and crystallization sequence; (b) DTA curve: melting temperature and liquidus temperature.

EDS analysis of the samples quenched in water shows that the faceted primary crystals contain ca. 77.7 at.% of (Ni + Fe + Cu) and ca. 22.3 at.% of P. On the other hand, the eutectic constituent contains less phosphorus, which may be due to the presence of the solid solution Ni(Fe,Cu) that contains slightly less phosphorus (ca. 22.7 at.%) than a phosphide phase. This suggests that the crystals are (Ni, Cu, Fe)3 P phosphides. The cooling conditions provided by the melt spinning process were sufficient for amorphization of Ni63 Cu9 Fe8 P20 alloy. This is proved by X-ray studies of the as-cast melt-spun ribbon. This is in accordance with the fact that the alloy has a high volume fraction of eutectic in the samples cooled with much lower cooling rates (Figs. 1 and 2). The same observation was made in the cases of amorphous alloys or nanostructures in quaternary Pr–Cu–Ni–Al [24], ternary [25] and binary Nb–Si [26] systems. The authors of these papers carefully control the composition to be near the eutectic. With reference to the results obtained for Ni–Ti–Zr and Ni–Ti–Zr–Si

182

K. Ziewiec / Journal of Alloys and Compounds 387 (2005) 179–182

alloys [19] which have the value of supercooled liquid range Tx about 50 K, Ni63 Cu9 Fe8 P20 alloy presents the parameter on a similar level (50 K). Such high values as reported by Yi et al. [19] were due to fulfilling all of the empirical rules indicated by Shen and Schwarz [27], which are important in the search for amorphous alloys in the metal-metalloid system. In the case of the investigated alloy, we operated near the deep ternary eutectic. The alloy melts between Tm = 1148 K and Tl = 1181 K. While increasing the volume fraction of the eutectic, the glass forming ability could be also increased. Our present results clearly show that a composition with a substantial volume fraction of liquid eutectic can be undercooled to low temperatures at which the atomic motion is sufficiently slow to result in easy freezing of the liquid structure. It has been found that Ni63 Cu9 Fe8 P20 alloy has Tg /Tm , Tg /Tl , Tx /(Tg + Tl ) GFA indicators on a relatively high level, i.e.: Tg /Tm = 0.521, Tg /Tl = 0.506 and Tx /(Tg + Tl ) = 0.364, respectively. The coefficients mentioned above are slightly lower than the values obtained for the top glass formers [18,28]. However, the Ni63 Cu9 Fe8 P20 alloy is made of definitively cheaper precursors than most of the compositions mentioned by Lu et al. [18,28]. High-temperature X-ray studies correlate well with the DSC curves. The melt spun ribbons remain amorphous even at 623 K, i.e. at the temperature within the supercooled liquid region. Above that range the ribbons crystallize which is observed in the X-ray pattern at 673 K. The main DSC peak may be connected with the crystallization of amorphous alloy into a mixture of two phases, i.e. Ni(Cu, Fe) solid solution and (Ni, Cu, Fe)3 P phosphide. Once created the phases do not change significantly with increasing temperature. The only change concerns slight sharpening of the peaks.

5. Conclusions 1. The primary microstructure of the Ni63 Cu9 Fe8 P20 alloy in as-cast state after quenching in water consists of (Ni, Cu, Fe)3 P phosphide and eutectic. The eutectic is the leading microstructural constituent. 2. The melt spinning technique provided sufficient cooling conditions for amorphization of the Ni63 Cu9 Fe8 P20 alloy. The observation correlates well with the large volume fraction of the eutectic and with the fact that the alloy melts at a relatively low temperature range (between Tm = 1148 K and Tl = 1181 K). These two factors probably favour glass formation. 3. The Ni63 Cu9 Fe8 P20 alloy shows a distinct glass transition temperature at the level of 598 K, and crystallization with onset at 648 K, and the alloy presents the main GFA indicators Tg /Tm , Tg /Tl , Tx /(Tg + Tl ) on a relatively high level, i.e.: Tg /Tm = 0.521, Tg /Tl = 0.506 and Tx /(Tg + Tl ) = 0.364, respectively.

4. The relatively high extent of the supercooled liquid region Tx (50 K) for Ni63 Cu9 Fe8 P20 alloy suggests that the alloy may be a good potential material for consolidation in a bulk shape.

Acknowledgements The financial support by the KBN research project no 4 T08D 00524 is gratefully acknowledged.

References [1] A. Peker, W.L. Johnson, Appl. Phys. Lett. 63 (1993) 2342. [2] Y. He, R.B. Schwarz, J.I. Archuleta, Appl. Phys. Lett. 69 (1996) 1861. [3] A. Inoue, T. Nakamura, N. Nishiyama, T. Masumoto, Mater. Trans. JIM 33 (1990) 937. [4] A. Inoue, Acta Mater. 48 (2000) 279. [5] W.L. Johnson, MRS Bull. (October 1995) 42–56. [6] A. Inoue, A. Takeuchi, B. Shen, Mater. Trans. 42 (6) (2001). [7] Z.P. Lu, T.T. Goh, Y. Li, S.C. Ng, Acta Mater. 47 (7) (1999) 2215–2224. [8] G.J. Fan, W. Loeser, S. Roth, J. Eckert, L. Schultz, J. Mater. Res. 15 (7) (2000) 1556–1563. [9] C.C. Hays, J. Schroers, W.L. Johnson, Appl. Phys. Lett. 79 (11) (2001) 1605–1607. [10] Y. Zhang, D.Q. Zhao, B.C. Wei, P. Wen, M.X. Pan, W.H. Wang, J. Mater. Res. 16 (6) (2001) 1675–1679. [11] R. Willnecker, K. Wittmann, G.P. G¨orler, J. Non-Cryst. Solids 156–158 (1993) 450. [12] A. Inoue, N. Nishiyama, Mater. Sci. Eng. A 226–228 (1997) 401–405. [13] Z.P. Lu, Y. Li, S.C. Ng, J. Non-Cryst. Solids 270 (2000) 103–114. [14] N. Nishiyama, A. Inoue, Mater. Trans. JIM 38 (1997) 464. [15] I.R. Lu, G. Wilde, G.P. Gr¨oler, R. Wilnecker, J. Non-Cryst. Solids 250 (1999) 577. [16] J. Schroers, W.L. Johnson, Appl. Phys. Lett. 80 (2002) 2069. [17] W.L. Johnson, MRS Bull. 24 (1999) 42. [18] Z.P. Lu, C.T. Liu, Acta Mater. 50 (2002) 3501–3512. [19] S. Yi, J.K. Lee, W.T. Kim, D.H. Kim, J. Non-Cryst. Solids 291 (2001) 132–136. [20] X. Wang, I. Yoshii, A. Inoue, Y.H. Kim, I.B. Kim, Mater. Trans. JIM 40 (1999) 1130. [21] T. Waniuk, J. Schroers, W.L. Johnson, Phys. Rev. B 67 (2003) 184–203. [22] P. Villars, A. Prince, H. Okamoto, Handbook of Ternary Alloy Phase Diagrams, vols. 8 and 10, ASM International, The Materials Information Society, 1995. [23] JCPDS-International Centre for Diffraction Data v. 2.00, 1998. [24] Y. Zhang, Y. Li, Glass Forming Ability in Pr–(Cu,Ni)–Al Alloys, Symposium of The Singapore MIT Alliance, 2001. [25] K. Ziewiec, P. Olszewski, R. Gajerski, A. Małecki, J. Alloys Compd. 373 (2004) 115–121. [26] R. Abbaschian, M.D. Lipschutz, Mater. Sci. Eng. A 226–228 (1997) 13–21. [27] T.D. Shen, R.B. Schwarz, Appl. Phys. Lett. 75 (1999) 49. [28] Z.P. Lu, H. Tan, Y. Li, S.C. Ng, Scripta Mater. 42 (2000) 667– 673.