- Email: [email protected]
Glass-forming ability and thermal stability of a new bulk metallic glass in the quaternary Zr–Cu–Ni–Al system
Journal of Non-Crystalline Solids 351 (2005) 2519–2523 www.elsevier.com/locate/jnoncrysol
Glass-forming ability and thermal stability of a new bulk metallic glass in the quaternary Zr–Cu–Ni–Al system J. Shen
a,b,*
, J. Zou c, L. Ye a, Z.P. Lu d, D.W. Xing b, M. Yan b, J.F. Sun
b
a
c
School of Aerospace, Mechanical and Mechatronic Engineering, The University of Sydney, NSW 2006, Australia b School of Materials Science and Engineering, Harbin Institute of Technology, Harbin 150001, China School of Engineering and Center for Microscopy and Microanalysis, The University of Queensland, QLD 4066, Australia d Metals and Ceramics Division, Oak Ridge National Laboratory, Oak Ridge, TN 37831-6115, USA Received 27 February 2005; received in revised form 26 April 2005
Abstract A new quaternary Zr-based bulk metallic glass, Zr51Cu20.7Ni12Al16.3, was developed by using a newly proposed approach. It has shown that the reduced glass transition temperature, defined as the ratio of the glass transition temperature to the melting point temperature, of the new quaternary Zr-based bulk metallic glass is as high as 0.66. The Gibbs free energy difference between the glassy state and the corresponding equilibrium crystals is estimated to be 0.99 kJ/mol. The maximum diameter of the glassy rod-like sample prepared through conventional copper-mold casting is close to 5 mm. The single amorphous phase remains unchanged after holding at 400 °C for 60 min. These suggest this new alloy having a reasonable glass-forming ability and a high thermal stability against crystallization, promising a base composition for creating other Zr-based bulk glass formers. Ó 2005 Elsevier B.V. All rights reserved. PACS: 61.43.Dq; 61.82.Bg; 64.70.Pf; 68.60.Dv
1. Introduction The last 10 years have witnessed the discovery of multi-component bulk metallic glasses (BMGs) mainly in Zr- [1–3], Pd- [4,5], Mg- [6–10], Ti- [11–13], Ni- [14– 16], Fe- [17–23], Cu- [24–26], La- [27], and Y- [28] based systems. The new BMGs can be readily manufactured by using conventional casting techniques at a cooling rate as low as 1–100 K/s. These BMGs have demonstrated various superior characteristics and properties over conventional amorphous metals, promising a variety of applications in engineering as a class of novel * Corresponding author. Address: School of Materials Science and Engineering, Harbin Institute of Technology, Harbin 150001, China. Tel.: +86 451 86418317; fax: +86 451 86413903. E-mail address: [email protected] (J. Shen).
0022-3093/$ - see front matter Ó 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.jnoncrysol.2005.07.009
structural and functional materials. As such, considerable efforts have been devoted to searching novel BMGs with high glass-forming ability (GFA) and high thermal stability. Due to the complexity in the dominant factors governing the formation of multi-component BMGs, however, it is rather difficult to expect a theory which can be used to guide the searching of good glass formers in metallic systems. As a pioneering work, Inoue [29] has proposed the three empirical rules for searching new BMGs: (a) requirement of three or more elements; (b) a difference in atomic size ratios above 12% among the main constituent elements; and (c) negative heats of mixing among the main constituent elements. Although those three empirical rules have been widely used to interpret the GFA in many BMG systems, the method of trail-and-error through extensive experimental searching and screening is still dominating the development of
J. Shen et al. / Journal of Non-Crystalline Solids 351 (2005) 2519–2523
new BMG systems. Recently, Xing [30] proposed a simple and effective approach for efficiently locating the basic compositions of BMGs. The new approach for predicting new BMG compositions is based on the concept that the formation of glassy metals requires suppression of nucleation and growth of the competing crystalline phases. If the chemical affinity between the atomic pairs of the constitutional elements can be balanced to a certain degree, the molten metallic liquid would have a strong ability to prevent the precipitation of any compounds from the melt during the cooling process. It is assumed that such balance in the chemical affinity can be achieved by proportionally mixing the corresponding binary eutectic compositions with low eutectic points, leading to a base composition of a new BMG. By using this approach, we obtained a new Zrbased BMG alloy in the quaternary Zr–Cu–Ni–Al system, i.e., Zr51Cu20.7Ni12Al16.3 (hereafter referred as alloy Zr51, all compositions are in atomic percentage), which is deduced based on the mixing of binary eutectics Zr38Cu62, Zr64Ni36 and Zr51Al49 with the ratio of 1/3(Zr38Cu62) + 1/3(Zr64Ni36) + 1/3(Zr51Al49). In this article, we study the glass-forming ability and the thermal stability of alloy Zr51. In order to evaluate the alloy Zr51, the well-known Zr65Cu17.5Ni10Al7.5 BMG alloy [1,29] (referred as alloy Zr65) was manufactured for comparison.
2. Experimental methods Alloy button ingots having target compositions were prepared by non-consumable arc melting under a Ti-gettered argon atmosphere. Each starting element has a purity of better than 99.9%. For each ingot the melting was repeated for four times to ensure composition homogeneity. Glassy alloys of cylindrical rods forms were prepared by remelting the button ingots and drop casting the melts into a copper-mould cavity. Differential scanning calorimetry (DSC) was conducted under flowing argon using TA 2920 and Netzsch STA 409C/CD DSC instruments to measure the thermal parameters of studied alloys. Annealing was also performed in the DSC cells to examine the thermal stability of the glassy alloys. The amorphous nature of the as-cast samples was monitored by X-ray diffraction (XRD) analysis using Cu radiation. Microstructures for both as-cast and annealed alloys were investigated using transmission electron microscopy (TEM). To prevent potential annealing during TEM specimen preparation, the TEM specimens in this study were prepared by grinding the slices, taken from the alloy ingots, on a 9 lm diamond lapping films in ethyl alcohol to produce fine powders. The powders were then spread out on holy carbon films supported by copper grids. TEM investiga-
tion was carried out in a Tecnai F20 electron microscope operating at 200 kV.
3. Results Fig. 1 shows the XRD patterns of the as-cast Zr51 alloy with different sample size. Fig. 1(a) shows that the as-cast sample with a diameter of 3 mm is mostly amorphous, evidenced by the broad diffusion halo dominating its XRD pattern. However, for the as-cast alloy Zr51 sample of 5 mm in diameter, Fig. 1(b) exhibits several peaks superimposed on the diffusion halo, indicating that this sample contains a certain amount of crystalline phases. This means that, under our experimental condition, we have achieved to produce glassy sample of the Zr51 alloy with a diameter of close to 5 mm. This achievement suggests this newly developed quaternary Zr-based BMG alloy has a reasonable glass-forming ability. Fig. 2 shows the DSC thermograms of the alloys Zr51 and Zr65 in the as-cast and annealed states. Table 1 summarizes the characteristic thermal parameters, deduced from the DSC data, namely, the onset glass transition temperature Tg, the onset glass crystallization temperature Tx, the onset melting point Tm, the supercooled liquid region DTx(=Tx Tg), the reduced glass transition temperature Trg(=Tg/Tm), the crystallization enthalpies DHc1 and DHc2, defined as the integrated area of the crystallization peak, respectively for the samples before and after annealing, and the crystallized fraction due to the annealing [=(DHc1 DHc2)/DHc1]. As can be seen from Fig. 2(a) that, both alloys show a distinct glass transition and a significant supercooled liquid region, illustrating characteristics of BMG alloys with high GFA and high thermal stability against crystallization. It is noted that the alloy Zr51 shows a much higher
(c)
Intensity, a.u.
2520
(b)
(a)
10
20
30
40
50
60
70
80
90
2Theta, Degree Fig. 1. XRD patterns of the as-cast (a) alloy Zr51, B3 mm, (b) alloy Zr51, B5 mm, and (c) alloy Zr51Y0.5, B10 mm.
J. Shen et al. / Journal of Non-Crystalline Solids 351 (2005) 2519–2523
Heat Flow Exo. (a.u.)
Tx
Tg
(a) as-cast alloy Zr51 annealed alloy Zr51 as-cast alloy Zr65 annealed alloy Zr 65
500
550
600
650
700
750
800
850
Heat Flow Exo. (mW)
Temperature (K) 3
Cooling
(b)
Cooling
(c)
2 1 0 -1
Tm
Tm
Heating -2 1000
1050
1100
Heating 1150
1200 1000 1050
Temperature (K)
1100
1150
1200
Temperature (K)
Fig. 2. DSC thermograms showing the glass transition and crystallization for the alloys Zr51 and Zr65 in the as-cast and annealed states (a) and the melting and solidification behaviors for the as-cast alloys Zr51 (b) and Zr65 (c). The heating and cooling rates were 5 K/min.
Tg, Tx and Trg, compared to those of alloy Zr65. The Trg of alloy Zr51 is even high up to those of the quinary Vitreloy series of BMGs [31], the best Zr-based metallic glass formers reported so far. It is worthwhile to emphasize that the alloy Zr51 has the highest glass transition temperature Tg, amongst all reported Zr-based BMGs (see Table 2 in Ref. [31]). From these results, one can conclude that Trg cannot be used as a criterion for measuring the GFA of the alloy Zr51. Furthermore, both the melting and solidification DSC curves of alloy Zr51 show only a single endothermic peak upon melting and a single exothermic peak upon cooling from the melt [see Fig. 2(b)]. This implies a concurrent melting and formation of the equilibrium compounds in the heating and cooling cycle, respectively. For this reason, the composition of this alloy is close to the eutectic. In contrast, two or three distinctly separated endothermic and exothermic peaks appear in the heating and cooling
2521
DSC curves of the alloy Zr65 [see Fig. 2(c)], indicating a melting and solidification behavior of an off-eutectic alloy. It is well known that glass formation favors for the alloys with or near eutectic compositions [20]. After annealing at 400 °C for 60 min, the alloys Zr51 and Zr65 also show a similar DSC curve to their corresponding as-cast alloy. The alloy Zr65, however, has already experienced noticeable crystallization, whilst the alloy Zr51 remains a primary amorphous state with no indication of crystallization (see Table 1). This reflects the alloy Zr51 having higher thermal stability than the alloy Zr65, although the latter has a much wider supercooled liquid region. It is known that undercooled melts can be frozen to be glassy as long as the formation of the competing crystalline phases can be suppressed. Thermodynamically, the tendency for the glass formation is favorable if the Gibbs free energy difference between the undercooled melts and the corresponding crystalline solids (i.e., DGl!s) is small. In this case, the critical nucleation work for the formation of a crystal becomes large and therefore the nucleation rates are greatly reduced. As a consequence, DGl!s is a good measurement for GFA – the less the DGl!s, the higher is the GFA [32]. Here we can estimate the DGl!s for the alloy Zr51 by using the following expression [28]: DH f ðT m T Þ cDH f Tm DGl!s ¼ ðT m T Þ T ln ; Tm Tm T ð1Þ where DHf is the enthalpy of fusion, T, the temperature of the undercooled melt, and c, the proportionality coefficient. By taking c = 0.8 for metallic glass-forming liquids [28], DHf = 5.44 kJ/mol [deduced from Fig. 2(b)] and Tm = 1076.2 K (see Table 1), we can obtain the DGl!s = 0.99 kJ/mol for alloy Zr51 at T = 0.8Tm. It is evident that the DGl!s for alloy Zr51 is significantly lower than that (DGl!s = 2.0 kJ/mol, estimated from Fig. 11 in Ref. [32]) for alloy Zr65. The relatively small DGl!s for the alloy Zr51 provides an indication that this alloy essentially possesses a stronger glass-forming ability, compared to that of the alloy Zr65. Fig. 3 presents the high-resolution transmission electron microscopy (HRTEM) images and their corresponding selected area electron diffraction (SAED) patterns of the as-cast and annealed alloys. All TEM specimens were taken from the rod samples with a diameter of 3 mm. Fig. 3(a) shows a homogenous maze
Table 1 Thermal data of as-cast alloys Zr51 and Zr65, obtained from their DSC traces Alloy
Tg (K)
Tx (K)
DTx (K)
Tm (K)
Trg
DHc1 (kJ/mol)
DHc2 (kJ/mol)
Cfa (%)
Zr51 Zr65
714.1 620.3
776.0 704.1
51.9 83.8
1076.2 1101.6
0.66 0.56
2.37 3.69
2.37 2.66
0.0 27.9
a
Crystallized fraction due to annealing.
2522
J. Shen et al. / Journal of Non-Crystalline Solids 351 (2005) 2519–2523
Fig. 3(c). However, Fig. 3(d) shows that the annealed alloy Zr65 has gone through a significant crystallization. This comparison confirms the conclusion made from the DSC data that the alloy Zr51 has a higher thermal stability than the alloy Zr65.
4. Discussion
Fig. 3. High-resolution TEM images and corresponding selected area electron diffraction patterns (insets) for (a) as-cast alloy Zr51, (b) ascast alloy Zr65, (c) annealed alloy Zr51, and (d) annealed alloy Zr65.
HREM contrast and a single halo ring in its corresponding electron diffraction patterns, indicating the nature of completely amorphous phase in the as-cast alloy Zr51. In the case of the as-cast alloy Zr65, although the amorphous contrast still dominates, as evidenced by the HRTEM image shown in Fig. 3(b), lattice fringes from crystalline clusters with a size of the nanometer scale [as encircled in Fig. 3(b)] and sharp electron diffraction spots co-existing with the halo ring in the SAED pattern can also observed, indicating that the as-cast alloy Zr65 has experienced a certain degree of crystallization under the identical manufacturing condition with that used for the as-cast alloy Zr51 in the present study. Interestingly, after annealed at 400 °C for 60 min, the alloy Zr51 still remains its amorphous structure as verified in
The above experimental results support the alloy Zr51 being a true alloy with a larger GFA and higher thermal stability than the alloy Zr65. Their significant difference in GFA cannot be simply explained in the frame of three empirical rules, because they have same components (namely Zr, Al, Ni, Cu). The atomic sizes, the mixing heats between the elements in these two studied alloys should be the same qualitatively. We believed that the reason of the better GFA of the alloy Zr51 is primarily due to the fact that the chemical interaction between the base element (Zr) and the glass-forming elements (Al, Ni, and Cu) has been balanced in the melt prior to solidification, and consequently providing an extent more favorable for glass formation. This study suggests that such a balance in the chemical interaction can be achieved by the appropriate choice of the mixing ratios of the corresponding binary eutectic compositions. The easiness of glass formation for the alloy Zr51 can be explained by that the selection of the mixing ratios and the binary eutectic compositions is more reasonable for suppression of the formation of the competing crystals, when compared with those for the alloy Zr65. This success of developing a new BMG system suggests our new approach for searching new BMGs is an easy and efficient approach to pin-pointing the required composition for a chosen BMG system. Based on the resultant composition, other BMGs with large GFA can be obtained either by appropriately adjusting the contents of the constitutional elements, or by adding alloying elements. As an example, we can significantly improve the GFA of the alloy Zr51 by adding 0.5% yttrium (the resultant alloy is referred as Zr51Y0.5). The modified alloy can be fabricated into a BMG ingot with a diameter of 10 mm, as shown in Fig. 1(c). The investigation on the GFA of alloy Zr51Y0.5 and the role of yttrium addition is beyond the scope of this study and will be discussed elsewhere.
5. Conclusions A new BMG alloy, Zr51Cu20.7Ni12Al16.3, is designed by using a newly developed approach for determining the basic compositions of BMG formers. The GFA and thermal stability of this new alloy have been assessed from the thermodynamic and microstructural
J. Shen et al. / Journal of Non-Crystalline Solids 351 (2005) 2519–2523
points of view. The most striking result from this study is the easy locating of the multi-component alloy compositions with a good GFA using a powerful and perspective method other than following tedious experimental work. It is expected that the resultant basic compositions can be further modified by appropriate adjustment of the constituent elements or alloying additions in the base alloys to achieve compositions which can be easily produced into amorphous materials in bulk forms through conventional metallurgical routes. Acknowledgements One of authors (Z.P. Lu) would like to acknowledge the financial support by the Division of Materials Sciences and Engineering, Office of Basic Energy Sciences, U.S. Department of Energy under contract DE-AC0500OR-22725 with UT-Battelle, LLC. References [1] A. Inoue, T. Zhang, N. Nishiyama, K. Ohba, T. Masumoto, Mater. Trans. JIM 34 (1993) 1234. [2] A. Peker, W.L. Johnson, Appl. Phys. Lett. 63 (1993) 2342. [3] L.Q. Xing, P. Ochin, M. Harmelin, F. Faudot, J. Bigot, J. NonCryst. Solids 205–207 (1996) 597. [4] N. Nishiyama, A. Inoue, Mater. Trans. JIM 43 (2002) 1913. [5] A. Inoue, N. Nishiyama, H.M. Kimura, Mater. Trans. JIM 38 (1997) 179. [6] A. Inoue, T. Nakamura, N. Nishiyama, T. Masumoto, Mater. Trans. JIM 33 (1992) 937. [7] Y. Li, H.Y. Liu, H. Jones, J. Mater. Sci. 31 (1996) 1957.
2523
[8] H. Ma, J. Xu, E. Ma, Appl. Phys. Lett. 83 (2003) 2793. [9] H. Men, D.H. Kim, J. Mater. Res. 18 (2003) 1502. [10] X.K. Xi, R.J. Wang, D.Q. Zhao, M.X. Pan, W.H. Wang, J. NonCryst. Solids 344 (2004) 105. [11] X.H. Lin, W.L. Johnson, J. Appl. Phys. 78 (1995) 6514. [12] T. Zhang, A. Inoue, Mater. Trans. JIM 39 (1998) 1001. [13] G. He, J. Eckert, M. Hagiwara, J. Appl. Phys. 95 (2004) 1816. [14] X. Wang, I. Yoshii, A. Inoue, Y.H. Kim, I.B. Kim, Mater. Trans. JIM 40 (1999) 1130. [15] S. Yi, T.G. Park, D.H. Kim, J. Mater. Res. 15 (2000) 2425. [16] H. Choi-Yim, D.H. Xu, W.L. Johnson, Appl. Phys. Lett. 82 (2003) 1030. [17] A. Inoue, Y. Shinohara, J.S. Gook, Mater. Trans. JIM 36 (1995) 1427. [18] A. Inoue, B.L. Shen, A.R. Yavari, A.L. Greer, J. Mater. Res. 18 (2003) 1487. [19] T.D. Shen, R.B. Schwarz, Appl. Phys. Lett. 75 (1999) 49. [20] Z.P. Lu, C.T. Liu, J.R. Thompson, W.D. Porter, Phys. Rev. Lett. 92 (2004) 245503. [21] V. Ponnambalam, S.J. Poon, G.J. Shiflet, J. Mater. Res. 19 (2004) 1320. [22] J. Shen, Q.J. Chen, J.F. Sun, H.B. Fan, G. Wang, Appl. Phys. Lett. 86 (2005) 151907. [23] A. Inoue, B.L. Shen, C.T. Chang, Acta Mater. 52 (2004) 4093. [24] T. Zhang, T. Yamamoto, A. Inoue, Mater. Trans. 43 (2002) 3222. [25] D.H. Xu, G. Duan, W.L. Johnson, Phys. Rev. Lett. 92 (2004) 245504. [26] E.S. Park, W.T. Kim, D.H. Kim, Mater. Trans. 45 (2004) 2693. [27] Y. Zhang, H. Tan, Y. Li, Mater. Sci. Eng. A 375–377 (2004) 436. [28] F.Q. Guo, S.J. Poon, G.J. Shiflet, Appl. Phys. Lett. 83 (2003) 2575. [29] A. Inoue, Acta Mater. 48 (2000) 279. [30] D.W. Xing, Doctoral Dissertation, Harbin Institute of Technology, China, 2003. [31] W.H. Wang, C. Dong, C.H. Shek, Mater. Sci. Eng. R 44 (2004) 45. [32] G.J. Fan, J.F. Loffler, R.K. Wunderlich, H.J. Fecht, Acta Mater. 52 (2004) 667.
Glass-forming ability and thermal stability of a new bulk metallic glass in the quaternary Zr–Cu–Ni–Al system J. Shen
a,b,*
, J. Zou c, L. Ye a, Z.P. Lu d, D.W. Xing b, M. Yan b, J.F. Sun
b
a
c
School of Aerospace, Mechanical and Mechatronic Engineering, The University of Sydney, NSW 2006, Australia b School of Materials Science and Engineering, Harbin Institute of Technology, Harbin 150001, China School of Engineering and Center for Microscopy and Microanalysis, The University of Queensland, QLD 4066, Australia d Metals and Ceramics Division, Oak Ridge National Laboratory, Oak Ridge, TN 37831-6115, USA Received 27 February 2005; received in revised form 26 April 2005
Abstract A new quaternary Zr-based bulk metallic glass, Zr51Cu20.7Ni12Al16.3, was developed by using a newly proposed approach. It has shown that the reduced glass transition temperature, defined as the ratio of the glass transition temperature to the melting point temperature, of the new quaternary Zr-based bulk metallic glass is as high as 0.66. The Gibbs free energy difference between the glassy state and the corresponding equilibrium crystals is estimated to be 0.99 kJ/mol. The maximum diameter of the glassy rod-like sample prepared through conventional copper-mold casting is close to 5 mm. The single amorphous phase remains unchanged after holding at 400 °C for 60 min. These suggest this new alloy having a reasonable glass-forming ability and a high thermal stability against crystallization, promising a base composition for creating other Zr-based bulk glass formers. Ó 2005 Elsevier B.V. All rights reserved. PACS: 61.43.Dq; 61.82.Bg; 64.70.Pf; 68.60.Dv
1. Introduction The last 10 years have witnessed the discovery of multi-component bulk metallic glasses (BMGs) mainly in Zr- [1–3], Pd- [4,5], Mg- [6–10], Ti- [11–13], Ni- [14– 16], Fe- [17–23], Cu- [24–26], La- [27], and Y- [28] based systems. The new BMGs can be readily manufactured by using conventional casting techniques at a cooling rate as low as 1–100 K/s. These BMGs have demonstrated various superior characteristics and properties over conventional amorphous metals, promising a variety of applications in engineering as a class of novel * Corresponding author. Address: School of Materials Science and Engineering, Harbin Institute of Technology, Harbin 150001, China. Tel.: +86 451 86418317; fax: +86 451 86413903. E-mail address: [email protected] (J. Shen).
0022-3093/$ - see front matter Ó 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.jnoncrysol.2005.07.009
structural and functional materials. As such, considerable efforts have been devoted to searching novel BMGs with high glass-forming ability (GFA) and high thermal stability. Due to the complexity in the dominant factors governing the formation of multi-component BMGs, however, it is rather difficult to expect a theory which can be used to guide the searching of good glass formers in metallic systems. As a pioneering work, Inoue [29] has proposed the three empirical rules for searching new BMGs: (a) requirement of three or more elements; (b) a difference in atomic size ratios above 12% among the main constituent elements; and (c) negative heats of mixing among the main constituent elements. Although those three empirical rules have been widely used to interpret the GFA in many BMG systems, the method of trail-and-error through extensive experimental searching and screening is still dominating the development of
J. Shen et al. / Journal of Non-Crystalline Solids 351 (2005) 2519–2523
new BMG systems. Recently, Xing [30] proposed a simple and effective approach for efficiently locating the basic compositions of BMGs. The new approach for predicting new BMG compositions is based on the concept that the formation of glassy metals requires suppression of nucleation and growth of the competing crystalline phases. If the chemical affinity between the atomic pairs of the constitutional elements can be balanced to a certain degree, the molten metallic liquid would have a strong ability to prevent the precipitation of any compounds from the melt during the cooling process. It is assumed that such balance in the chemical affinity can be achieved by proportionally mixing the corresponding binary eutectic compositions with low eutectic points, leading to a base composition of a new BMG. By using this approach, we obtained a new Zrbased BMG alloy in the quaternary Zr–Cu–Ni–Al system, i.e., Zr51Cu20.7Ni12Al16.3 (hereafter referred as alloy Zr51, all compositions are in atomic percentage), which is deduced based on the mixing of binary eutectics Zr38Cu62, Zr64Ni36 and Zr51Al49 with the ratio of 1/3(Zr38Cu62) + 1/3(Zr64Ni36) + 1/3(Zr51Al49). In this article, we study the glass-forming ability and the thermal stability of alloy Zr51. In order to evaluate the alloy Zr51, the well-known Zr65Cu17.5Ni10Al7.5 BMG alloy [1,29] (referred as alloy Zr65) was manufactured for comparison.
2. Experimental methods Alloy button ingots having target compositions were prepared by non-consumable arc melting under a Ti-gettered argon atmosphere. Each starting element has a purity of better than 99.9%. For each ingot the melting was repeated for four times to ensure composition homogeneity. Glassy alloys of cylindrical rods forms were prepared by remelting the button ingots and drop casting the melts into a copper-mould cavity. Differential scanning calorimetry (DSC) was conducted under flowing argon using TA 2920 and Netzsch STA 409C/CD DSC instruments to measure the thermal parameters of studied alloys. Annealing was also performed in the DSC cells to examine the thermal stability of the glassy alloys. The amorphous nature of the as-cast samples was monitored by X-ray diffraction (XRD) analysis using Cu radiation. Microstructures for both as-cast and annealed alloys were investigated using transmission electron microscopy (TEM). To prevent potential annealing during TEM specimen preparation, the TEM specimens in this study were prepared by grinding the slices, taken from the alloy ingots, on a 9 lm diamond lapping films in ethyl alcohol to produce fine powders. The powders were then spread out on holy carbon films supported by copper grids. TEM investiga-
tion was carried out in a Tecnai F20 electron microscope operating at 200 kV.
3. Results Fig. 1 shows the XRD patterns of the as-cast Zr51 alloy with different sample size. Fig. 1(a) shows that the as-cast sample with a diameter of 3 mm is mostly amorphous, evidenced by the broad diffusion halo dominating its XRD pattern. However, for the as-cast alloy Zr51 sample of 5 mm in diameter, Fig. 1(b) exhibits several peaks superimposed on the diffusion halo, indicating that this sample contains a certain amount of crystalline phases. This means that, under our experimental condition, we have achieved to produce glassy sample of the Zr51 alloy with a diameter of close to 5 mm. This achievement suggests this newly developed quaternary Zr-based BMG alloy has a reasonable glass-forming ability. Fig. 2 shows the DSC thermograms of the alloys Zr51 and Zr65 in the as-cast and annealed states. Table 1 summarizes the characteristic thermal parameters, deduced from the DSC data, namely, the onset glass transition temperature Tg, the onset glass crystallization temperature Tx, the onset melting point Tm, the supercooled liquid region DTx(=Tx Tg), the reduced glass transition temperature Trg(=Tg/Tm), the crystallization enthalpies DHc1 and DHc2, defined as the integrated area of the crystallization peak, respectively for the samples before and after annealing, and the crystallized fraction due to the annealing [=(DHc1 DHc2)/DHc1]. As can be seen from Fig. 2(a) that, both alloys show a distinct glass transition and a significant supercooled liquid region, illustrating characteristics of BMG alloys with high GFA and high thermal stability against crystallization. It is noted that the alloy Zr51 shows a much higher
(c)
Intensity, a.u.
2520
(b)
(a)
10
20
30
40
50
60
70
80
90
2Theta, Degree Fig. 1. XRD patterns of the as-cast (a) alloy Zr51, B3 mm, (b) alloy Zr51, B5 mm, and (c) alloy Zr51Y0.5, B10 mm.
J. Shen et al. / Journal of Non-Crystalline Solids 351 (2005) 2519–2523
Heat Flow Exo. (a.u.)
Tx
Tg
(a) as-cast alloy Zr51 annealed alloy Zr51 as-cast alloy Zr65 annealed alloy Zr 65
500
550
600
650
700
750
800
850
Heat Flow Exo. (mW)
Temperature (K) 3
Cooling
(b)
Cooling
(c)
2 1 0 -1
Tm
Tm
Heating -2 1000
1050
1100
Heating 1150
1200 1000 1050
Temperature (K)
1100
1150
1200
Temperature (K)
Fig. 2. DSC thermograms showing the glass transition and crystallization for the alloys Zr51 and Zr65 in the as-cast and annealed states (a) and the melting and solidification behaviors for the as-cast alloys Zr51 (b) and Zr65 (c). The heating and cooling rates were 5 K/min.
Tg, Tx and Trg, compared to those of alloy Zr65. The Trg of alloy Zr51 is even high up to those of the quinary Vitreloy series of BMGs [31], the best Zr-based metallic glass formers reported so far. It is worthwhile to emphasize that the alloy Zr51 has the highest glass transition temperature Tg, amongst all reported Zr-based BMGs (see Table 2 in Ref. [31]). From these results, one can conclude that Trg cannot be used as a criterion for measuring the GFA of the alloy Zr51. Furthermore, both the melting and solidification DSC curves of alloy Zr51 show only a single endothermic peak upon melting and a single exothermic peak upon cooling from the melt [see Fig. 2(b)]. This implies a concurrent melting and formation of the equilibrium compounds in the heating and cooling cycle, respectively. For this reason, the composition of this alloy is close to the eutectic. In contrast, two or three distinctly separated endothermic and exothermic peaks appear in the heating and cooling
2521
DSC curves of the alloy Zr65 [see Fig. 2(c)], indicating a melting and solidification behavior of an off-eutectic alloy. It is well known that glass formation favors for the alloys with or near eutectic compositions [20]. After annealing at 400 °C for 60 min, the alloys Zr51 and Zr65 also show a similar DSC curve to their corresponding as-cast alloy. The alloy Zr65, however, has already experienced noticeable crystallization, whilst the alloy Zr51 remains a primary amorphous state with no indication of crystallization (see Table 1). This reflects the alloy Zr51 having higher thermal stability than the alloy Zr65, although the latter has a much wider supercooled liquid region. It is known that undercooled melts can be frozen to be glassy as long as the formation of the competing crystalline phases can be suppressed. Thermodynamically, the tendency for the glass formation is favorable if the Gibbs free energy difference between the undercooled melts and the corresponding crystalline solids (i.e., DGl!s) is small. In this case, the critical nucleation work for the formation of a crystal becomes large and therefore the nucleation rates are greatly reduced. As a consequence, DGl!s is a good measurement for GFA – the less the DGl!s, the higher is the GFA [32]. Here we can estimate the DGl!s for the alloy Zr51 by using the following expression [28]: DH f ðT m T Þ cDH f Tm DGl!s ¼ ðT m T Þ T ln ; Tm Tm T ð1Þ where DHf is the enthalpy of fusion, T, the temperature of the undercooled melt, and c, the proportionality coefficient. By taking c = 0.8 for metallic glass-forming liquids [28], DHf = 5.44 kJ/mol [deduced from Fig. 2(b)] and Tm = 1076.2 K (see Table 1), we can obtain the DGl!s = 0.99 kJ/mol for alloy Zr51 at T = 0.8Tm. It is evident that the DGl!s for alloy Zr51 is significantly lower than that (DGl!s = 2.0 kJ/mol, estimated from Fig. 11 in Ref. [32]) for alloy Zr65. The relatively small DGl!s for the alloy Zr51 provides an indication that this alloy essentially possesses a stronger glass-forming ability, compared to that of the alloy Zr65. Fig. 3 presents the high-resolution transmission electron microscopy (HRTEM) images and their corresponding selected area electron diffraction (SAED) patterns of the as-cast and annealed alloys. All TEM specimens were taken from the rod samples with a diameter of 3 mm. Fig. 3(a) shows a homogenous maze
Table 1 Thermal data of as-cast alloys Zr51 and Zr65, obtained from their DSC traces Alloy
Tg (K)
Tx (K)
DTx (K)
Tm (K)
Trg
DHc1 (kJ/mol)
DHc2 (kJ/mol)
Cfa (%)
Zr51 Zr65
714.1 620.3
776.0 704.1
51.9 83.8
1076.2 1101.6
0.66 0.56
2.37 3.69
2.37 2.66
0.0 27.9
a
Crystallized fraction due to annealing.
2522
J. Shen et al. / Journal of Non-Crystalline Solids 351 (2005) 2519–2523
Fig. 3(c). However, Fig. 3(d) shows that the annealed alloy Zr65 has gone through a significant crystallization. This comparison confirms the conclusion made from the DSC data that the alloy Zr51 has a higher thermal stability than the alloy Zr65.
4. Discussion
Fig. 3. High-resolution TEM images and corresponding selected area electron diffraction patterns (insets) for (a) as-cast alloy Zr51, (b) ascast alloy Zr65, (c) annealed alloy Zr51, and (d) annealed alloy Zr65.
HREM contrast and a single halo ring in its corresponding electron diffraction patterns, indicating the nature of completely amorphous phase in the as-cast alloy Zr51. In the case of the as-cast alloy Zr65, although the amorphous contrast still dominates, as evidenced by the HRTEM image shown in Fig. 3(b), lattice fringes from crystalline clusters with a size of the nanometer scale [as encircled in Fig. 3(b)] and sharp electron diffraction spots co-existing with the halo ring in the SAED pattern can also observed, indicating that the as-cast alloy Zr65 has experienced a certain degree of crystallization under the identical manufacturing condition with that used for the as-cast alloy Zr51 in the present study. Interestingly, after annealed at 400 °C for 60 min, the alloy Zr51 still remains its amorphous structure as verified in
The above experimental results support the alloy Zr51 being a true alloy with a larger GFA and higher thermal stability than the alloy Zr65. Their significant difference in GFA cannot be simply explained in the frame of three empirical rules, because they have same components (namely Zr, Al, Ni, Cu). The atomic sizes, the mixing heats between the elements in these two studied alloys should be the same qualitatively. We believed that the reason of the better GFA of the alloy Zr51 is primarily due to the fact that the chemical interaction between the base element (Zr) and the glass-forming elements (Al, Ni, and Cu) has been balanced in the melt prior to solidification, and consequently providing an extent more favorable for glass formation. This study suggests that such a balance in the chemical interaction can be achieved by the appropriate choice of the mixing ratios of the corresponding binary eutectic compositions. The easiness of glass formation for the alloy Zr51 can be explained by that the selection of the mixing ratios and the binary eutectic compositions is more reasonable for suppression of the formation of the competing crystals, when compared with those for the alloy Zr65. This success of developing a new BMG system suggests our new approach for searching new BMGs is an easy and efficient approach to pin-pointing the required composition for a chosen BMG system. Based on the resultant composition, other BMGs with large GFA can be obtained either by appropriately adjusting the contents of the constitutional elements, or by adding alloying elements. As an example, we can significantly improve the GFA of the alloy Zr51 by adding 0.5% yttrium (the resultant alloy is referred as Zr51Y0.5). The modified alloy can be fabricated into a BMG ingot with a diameter of 10 mm, as shown in Fig. 1(c). The investigation on the GFA of alloy Zr51Y0.5 and the role of yttrium addition is beyond the scope of this study and will be discussed elsewhere.
5. Conclusions A new BMG alloy, Zr51Cu20.7Ni12Al16.3, is designed by using a newly developed approach for determining the basic compositions of BMG formers. The GFA and thermal stability of this new alloy have been assessed from the thermodynamic and microstructural
J. Shen et al. / Journal of Non-Crystalline Solids 351 (2005) 2519–2523
points of view. The most striking result from this study is the easy locating of the multi-component alloy compositions with a good GFA using a powerful and perspective method other than following tedious experimental work. It is expected that the resultant basic compositions can be further modified by appropriate adjustment of the constituent elements or alloying additions in the base alloys to achieve compositions which can be easily produced into amorphous materials in bulk forms through conventional metallurgical routes. Acknowledgements One of authors (Z.P. Lu) would like to acknowledge the financial support by the Division of Materials Sciences and Engineering, Office of Basic Energy Sciences, U.S. Department of Energy under contract DE-AC0500OR-22725 with UT-Battelle, LLC. References [1] A. Inoue, T. Zhang, N. Nishiyama, K. Ohba, T. Masumoto, Mater. Trans. JIM 34 (1993) 1234. [2] A. Peker, W.L. Johnson, Appl. Phys. Lett. 63 (1993) 2342. [3] L.Q. Xing, P. Ochin, M. Harmelin, F. Faudot, J. Bigot, J. NonCryst. Solids 205–207 (1996) 597. [4] N. Nishiyama, A. Inoue, Mater. Trans. JIM 43 (2002) 1913. [5] A. Inoue, N. Nishiyama, H.M. Kimura, Mater. Trans. JIM 38 (1997) 179. [6] A. Inoue, T. Nakamura, N. Nishiyama, T. Masumoto, Mater. Trans. JIM 33 (1992) 937. [7] Y. Li, H.Y. Liu, H. Jones, J. Mater. Sci. 31 (1996) 1957.
2523
[8] H. Ma, J. Xu, E. Ma, Appl. Phys. Lett. 83 (2003) 2793. [9] H. Men, D.H. Kim, J. Mater. Res. 18 (2003) 1502. [10] X.K. Xi, R.J. Wang, D.Q. Zhao, M.X. Pan, W.H. Wang, J. NonCryst. Solids 344 (2004) 105. [11] X.H. Lin, W.L. Johnson, J. Appl. Phys. 78 (1995) 6514. [12] T. Zhang, A. Inoue, Mater. Trans. JIM 39 (1998) 1001. [13] G. He, J. Eckert, M. Hagiwara, J. Appl. Phys. 95 (2004) 1816. [14] X. Wang, I. Yoshii, A. Inoue, Y.H. Kim, I.B. Kim, Mater. Trans. JIM 40 (1999) 1130. [15] S. Yi, T.G. Park, D.H. Kim, J. Mater. Res. 15 (2000) 2425. [16] H. Choi-Yim, D.H. Xu, W.L. Johnson, Appl. Phys. Lett. 82 (2003) 1030. [17] A. Inoue, Y. Shinohara, J.S. Gook, Mater. Trans. JIM 36 (1995) 1427. [18] A. Inoue, B.L. Shen, A.R. Yavari, A.L. Greer, J. Mater. Res. 18 (2003) 1487. [19] T.D. Shen, R.B. Schwarz, Appl. Phys. Lett. 75 (1999) 49. [20] Z.P. Lu, C.T. Liu, J.R. Thompson, W.D. Porter, Phys. Rev. Lett. 92 (2004) 245503. [21] V. Ponnambalam, S.J. Poon, G.J. Shiflet, J. Mater. Res. 19 (2004) 1320. [22] J. Shen, Q.J. Chen, J.F. Sun, H.B. Fan, G. Wang, Appl. Phys. Lett. 86 (2005) 151907. [23] A. Inoue, B.L. Shen, C.T. Chang, Acta Mater. 52 (2004) 4093. [24] T. Zhang, T. Yamamoto, A. Inoue, Mater. Trans. 43 (2002) 3222. [25] D.H. Xu, G. Duan, W.L. Johnson, Phys. Rev. Lett. 92 (2004) 245504. [26] E.S. Park, W.T. Kim, D.H. Kim, Mater. Trans. 45 (2004) 2693. [27] Y. Zhang, H. Tan, Y. Li, Mater. Sci. Eng. A 375–377 (2004) 436. [28] F.Q. Guo, S.J. Poon, G.J. Shiflet, Appl. Phys. Lett. 83 (2003) 2575. [29] A. Inoue, Acta Mater. 48 (2000) 279. [30] D.W. Xing, Doctoral Dissertation, Harbin Institute of Technology, China, 2003. [31] W.H. Wang, C. Dong, C.H. Shek, Mater. Sci. Eng. R 44 (2004) 45. [32] G.J. Fan, J.F. Loffler, R.K. Wunderlich, H.J. Fecht, Acta Mater. 52 (2004) 667.