Comparison of mechanical behavior between bulk and ribbon Cu-based metallic glasses

Materials Science and Engineering A 430 (2006) 350 354 Short communication Comparison of mechanical behavior between bulk and ribbon Cu-based metallic glasses W.H. Jiang a,, F.X. Liu a, Y.D. Wang a,b, H.F. Zhang c, H. Choo a,d, P.K. Liaw a a Department of Materials Science and Engineering, The University of Tennessee, Knoxville, TN 37996, USA b School of Materials and Metallurgy, Northeastern University, Shenyang 110006, China c Shenyang National Laboratory for Materials Science, Institute of Metals Research, Chinese Academy of Sciences, Shenyang 110016, China d Materials Science and Technology Division, Oak Ridge National Laboratory, Oak Ridge, TN 37831, USA Received 10 April 2006; received in revised form 11 May 2006; accepted 14 May 2006 Abstract As-cast bulk and as-spun ribbon Cu 60 Zr 30 Ti 10 metallic glasses were characterized using differential-scanning calorimetry and instrumented nanoindentation. Two alloys show a significant difference in the amount of free volume, which is attributed to the difference in a cooling rate, while exhibiting a similar serrated plastic flow. Atomic-force-microscopy observations demonstrate the pile-ups containing shear bands around the indents in both alloys. The as-cast bulk alloy has higher hardness and elastic modulus than the as-spun ribbon alloy. The difference in the strengths of two alloys may be related to the different amount of free volume. The strength seems to be more sensitive to a cooling rate during solidification than the plastic-flow behavior in the Cu 60 Zr 30 Ti 10. 2006 Elsevier B.V. All rights reserved. Keywords: Metallic glasses; Nanoindentation; Plastic deformation; Hardness; Free volume 1. Introduction Over the last decade, significant progress has been made in searching alloy compositions with sluggish crystallization kinetics, resulting in the development of numerous bulk-metallic glasses (BMGs) that include Zr-, Ni-, Ti-, Mg-, Cu-, Pt-, Fe-, Laand Pd-based alloy systems. As their critical cooling rates are as low as 100 K/s, the BMGs can be produced using conventional casting techniques and their size is as large as millimeters or even centimeters [1]. This progress makes the applications of metallic glasses, particularly as structural materials, close to realization [1]. The BMGs have outstanding mechanical properties, such as high strength of up to 5 GPa [2], large elastic-deformation limit of around 2% [1], as well as good fatigue properties [3 6]. The cooling rate is one of the most important factors affecting structures and properties of the metallic glasses. Conventionally, the metallic glasses are synthesized by a rapid solidification process, typically melt spinning, which gives a cooling rate of about 10 6 K/s. The resulting products are in the form of ribbons with Corresponding author. Tel.: +1 865 974 5413; fax: +1 865 974 4115. E-mail address: wjiang5@utk.edu (W.H. Jiang). a thickness of tens of micrometers [7]. The rapidly solidified ribbons must be different in structures and properties from the BMG ingots produced at a much lower cooling rate, generally less than 10 2 K/s. Illekováetal.[8] observed a substantial difference in the structural relaxation kinetics between the bulk and ribbon Zr 55 Ni 25 Al 20 metallic glasses and the different degrees of the short-range order. Bobrov et al. [9] found that the density of the bulk Zr 52.5 Ti 5 Cu 17.9 Ni 14.6 Al 10 glass is higher than that of the ribbon counterpart, indicating that the amount of free volume in a metallic glass is related to the cooling rate. Understanding the difference in the mechanical behavior between the rapidly solidified ribbons and BMG ingots is of both practical significance and fundamental interest. However, small thicknesses of as-spun ribbons (i.e., tens of micrometers) inhibit an investigation through the conventional mechanical testing, such as tension, compression or microhardness tests. An instrumented nanoindentation is a useful probe for studying the mechanical response of various materials to the microto nano-scale loads. This technique has been applied widely to the study of mechanical behavior of metallic glasses as recently reviewed by Schuh and Nieh [10]. In this paper, an instrumented nanoindenter is utilized to study the mechanical properties of the bulk and ribbon Cu 60 Zr 30 Ti 10 0921-5093/$ see front matter 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.msea.2006.05.042

(at.%) metallic glasses. It is expected that this work can provide insight to the effect of cooling rates on the mechanical properties of metallic glasses. 2. Experimental W.H. Jiang et al. / Materials Science and Engineering A 430 (2006) 350 354 351 The bulk Cu 60 Zr 30 Ti 10 metallic glass was prepared by arcmelting a mixture of the constituent elements in a purified argon atmosphere. In order to obtain the homogeneity, the alloy ingots were remelted several times before cast into a water-cooled copper mold using an in situ suction-casting facility. The resulting cylindrical BMG rods have a dimension of 3 mm in diameter and 30 mm in length. For the ribbon specimens, the master alloy with the nominal composition of Cu 60 Zr 30 Ti 10 was first synthesized by arc-melting a mixture of the constituent elements in a purified argon atmosphere. Then, the amorphous ribbons (2 mm 0.06 mm) were prepared from the master-alloy ingots using a single-roller melt-spinning apparatus with a copper wheel in a purified argon atmosphere. The amorphous structure of both bulk and ribbon alloys was examined by X-ray diffraction with Cu K radiation using Philips X pert Diffractometer. Differential-scanning calorimetry (DSC, Perkin-Elmer DSC7) was used to characterize their thermal properties. DSC was run two times for each specimen from 50 to 600 C at a heating rate of 20 C/min in an argon atmosphere. Taking the second scan as a base line, any thermal effect from the structural evolution during heating was investigated. The surfaces of the bulk and ribbon alloys were electropolished using a twin-jet thinning electropolisher for nanoindentation tests. For the ribbon alloy, the wheel side was subjected to polishing. Polishing was performed in a solution of 25% nitric acid and 75% methanol (vol.%) at 243 K and a voltage of 10 V. Nanoinstruments Nanoindenter II with a Berkovich diamond indenter was used to characterize their mechanical properties. At least 10 indents were made on each sample with the separation between adjacent indents of 20 m. The loading phase of the indentation was carried out under a displacement control at a loading rate of 10 nm/s to a maximum displacement of 350 nm. At 90% unloading, a dwell period of 100 s was imposed to correct for any thermal drift in the system, which was less than 0.5 nm/s. The load displacement curves, hardness and elastic modulus were obtained from indentation experiments. The atomic-force-microscopy (AFM) observation on the indents was conducted, using a Digital Instruments Nanoscope IIIa in the contact mode. Fig. 1. X-ray diffraction patterns of the bulk and ribbon Cu 60 Zr 30 Ti 10. DSC curves, the typical thermal properties were determined, as shown in Table 1. The bulk and ribbon alloys exhibit a similar thermal behavior. There are two exothermic peaks for the crystallization and an obvious endothermic peak for a glass transition for both alloys. The exothermic heats of the crystallization, H x ( H x = H 1 + H 2, where H 1 and H 2 are the exothermic heats of the first and the second crystallization, respectively.) are 41.7 and 45.5 J/g for the bulk and ribbon alloys, respectively. The crystallization peaks correspond to the precipitation of Cu 8 Zr 3 and Cu 10 Zr 7, respectively [11]. The 3. Results Fig. 1 shows the X-ray diffraction patterns of both bulk and ribbon Cu 60 Zr 30 Ti 10 metallic glasses. The broad diffraction maxima demonstrate the amorphous structure of the bulk and ribbon alloys. It has been observed that Cu 60 Zr 30 Ti 10 alloy has a good glass-forming ability, and the BMG rod with the maximum diameter of 4 mm has been reported [11]. Fig. 2 presents the DSC thermograms for the bulk and ribbon alloys that were normalized using the base lines. From the Fig. 2. (a) DSC traces of the bulk and ribbon Cu 60 Zr 30 Ti 10 at a heating rate of 20 min/s and (b) an enlarged portion of the dashed-line box in (a) showing an exothermic heat of the structural relaxation. The arrows in (b) indicate T g.

352 W.H. Jiang et al. / Materials Science and Engineering A 430 (2006) 350 354 Table 1 Thermal properties of the bulk and ribbon Cu 60 Zr 30 Ti 10 measured from DSC Alloys T g ( C) T x1 ( C) T x2 ( C) T x ( C) H o (J/g) H 1 (J/g) H 2 (J/g) H 1 + H 2 (J/g) Bulk 442.6 479.6 529.2 37.0 8.9 22.1 19.6 41.7 Ribbon 422.3 462.7 516.4 40.4 24.0 22.6 22.9 45.5 T g denotes the glass-transition temperature, T x1 and T x2 the first and the second crystallization temperatures, T x the supercooled liquid region, H o the structuralrelaxation exothermic heat, and H 1 and H 2 the exothermic heats of the first and the second crystallization. glass-transition temperatures, T g, and the supercooled liquid regions, T x ( T x = T x1 T g, where T x1 is the onset temperature of the first crystallization), are 442.6 and 37.0 C, and 422.3 and 40.4 C for the bulk and ribbon alloys, respectively. This trend shows that the supercooled liquid regions of the bulk and ribbon alloys are quite similar. But, the supercooled liquid region of the bulk alloy is shifted upward by about 20 C compared to that of the ribbon alloy. Fig. 2(b) exhibits the enlarged DSC scanning curves at low temperatures. The exothermic reaction manifesting structural relaxation for both bulk and ribbon alloys are clearly observed just below the T g. The structural relaxation exothermic heat ( H o ) of the bulk alloy (8.9 J/g) is much smaller than that of the ribbon alloy (24.0 J/g). This trend can be related to the large difference in the amount of the free volume between the bulk and ribbon alloys; that is, the amount of the free volume of the bulk alloy is smaller than that of the ribbon alloy. Table 2 shows the values of hardness and elastic modulus of the bulk and ribbon alloys obtained from the nanoindentation tests. A correction was applied, using the actual contact area of indents observed by AFM to account for the pile-ups. It is important to mention that the hardness was obtained from the nanoindentation with a maximum displacement of 350 nm and a loading rate of 10 nm/s. Due to an indentation-size effect, hardness values for amorphous alloys, derived from the nanoindentation, are higher than those obtained from macro- or microhardness testers [12,13]. Thus, the numbers we present here are only used for the comparison between the current two alloys. It can be seen that the bulk alloy has a higher hardness and elastic modulus than the ribbon alloy. Fig. 3(a) displays the typical load displacement curves for the bulk and ribbon alloys. The obvious serrations are observed for both alloys. From the nanoindentation data in Fig. 3(a), the strain rates, ε = (1/h)(dh/dt), where h is the indenter displacement during loading, were calculated. The values change with the depth, as shown in Fig. 3(b). The strain rates for both alloys show the same decreasing trend with increasing the depth. However, the curves are not smooth, and there are obvious serrations. These curves highlight the serration behavior observed in the load displacement curves (Fig. 3(a)) and demonstrate that there is little difference in the serrations between the bulk and ribbon alloys. Table 2 Hardness (H) and elastic modulus (E) of the bulk and ribbon Cu 60 Zr 30 Ti 10 Alloys H (GPa) E (GPa) Bulk 7.61 ± 0.33 93.88 ± 1.70 Ribbon 6.94 ± 0.13 61.39 ± 0.92 Fig. 3. (a) The load displacement curves for the bulk and ribbon Cu 60 Zr 30 Ti 10 during nanoindentation and (b) the corresponding strain rate vs. displacement curves during loading. In order to avoid overlapping, the curves were shifted along the displacement in (a) and the strain rate in (b), respectively. AFM observations of the indents on the bulk and ribbon alloys show that the indents were rather regular in shape and caused no cracking. Typical images are shown in Fig. 4(a and c). There are obvious pile-ups containing shear bands around the indents in both bulk and ribbon samples (Fig. 4(b and d)). The pile-ups have the height of about 22 and 34% of the indent depth for the bulk and ribbon alloys, respectively. 4. Discussion Amorphous materials contain a significant amount of the free volume that is frozen during cooling. The free volume is one of the most important structural features in amorphous materials, which significantly affects the mechanical, physical and chemical properties. DSC is one of the effective methods to characterize the structural relaxation that is closely related to the free volume in amorphous materials. In a DSC thermogram, the exothermic reaction just below the T g is a result of the annihi-

W.H. Jiang et al. / Materials Science and Engineering A 430 (2006) 350 354 353 Fig. 4. AFM images (a and c) and corresponding height profiles (b and d) of the indents formed at a maximum displacement of 350 nm in the bulk alloy (a and b) and ribbon alloy (c and d). The lines in the AFM images indicate the traces where height profiles were taken. lation of free volume and structural relaxation. The exothermic heat is proportional to the amount of the escaping free volume [14]. The present work demonstrates that the exothermic heat in the bulk Cu 60 Zr 30 Ti 10 is significantly less than that in the ribbon Cu 60 Zr 30 Ti 10 (Fig. 2(b) and Table 1), indicating that the bulk alloy contains less free volume than the ribbon alloy. A high cooling rate used in the process of the ribbon suppresses the escape of free volume during cooling. By measuring the density, Bobrov et al. [9] demonstrated that the bulk Zr 52.5 Ti 5 Cu 17.9 Ni 14.6 Al 10 has a higher density than the as-spun ribbon by about 1.3%. Using the positron annihilation, Nagel et al. [15] also observed that in Zr 46.7 Ti 8.5 Cu 7.5 Ni 10 Be 27.5, the rapid solidification froze in more free volumes. Even though the quantitative distinction in free volume between the bulk and ribbon alloys cannot be made here, a qualitative (or semi-quantitative) difference is clearly observed, which resulted from the large difference (about four orders of magnitude) in a cooling rate between two alloys [1,7]. Nanoindentation indicates that the bulk alloy has an appreciably higher hardness and elastic modulus than the ribbon alloy (Table 2). Their difference in mechanical strength may be closely related to the amount of the free volume in the metallic glasses. Evidently, an increase in free volume may increase the average atomic distance that affects the material stiffness, i.e., the elastic modulus. On the other hand, the free volume also influences the resistance to the plastic deformation. The increase in free volume results in the lower resistance to the plastic deformation. Jiang and Atzmon [16] also observed the decreases in the hardness values in the order of the as-relaxed, as-spun, and as-rolled states of the amorphous Al 86.8 Ni 3.7 Y 9.5, while the free volume in these states is expected to increase in the opposite order. In fact, it is widely observed that the relaxation annealing causes an increase in hardness/strength in metallic glasses [16 20]. The relatively high values of hardness and elastic modulus of the bulk alloy may be attributed to the smaller amount of free volume in its matrix, which was caused by the lower cooling rate. In the nanoindentation, the load displacement curves (Fig. 3(a)) of both bulk and ribbon alloys display obvious serrations. The serrated plastic flow has been observed extensively in metallic glasses [10,21 23]. At low temperatures (e.g., room temperature) and high strain rates, metallic glasses exhibit an inhomogeneous plastic deformation, and the deformation is confined to highly localized narrow regions, i.e., shear bands. The shear bands observed around the indents in both alloys indicate that the plastic deformation of both alloys is inhomogeneous. Wright et al. suggested that the serrated plastic flow is caused by the formation of individual shear bands [23]. Using an infrared camera, Jiang et al. recently observed in situ dynamic shearbanding processes during the uniaxial compression of a Zr-based BMG and demonstrated that the serrations in the plastic flow correspond to individual shear-banding events [24]. No appreciable distinction in serrations between the bulk and ribbon alloys can be made in the strain rate versus displacement curves (Fig. 3(b)). Therefore, the effect of free volume on the inhomogeneous plastic flow cannot be identified in the present study. 5. Conclusions The as-cast bulk and as-spun ribbon Cu 60 Zr 30 Ti 10 metallic glasses were characterized for the thermal and mechanical behaviors. The main results can be summarized as follows. (1) The as-cast bulk Cu 60 Zr 30 Ti 10 alloy has less free volume than the ribbon alloy, which is attributed to the relatively slower cooling rate. (2) The bulk alloy has a higher hardness and elastic modulus than the ribbon alloy. The difference in strength may be related to the different amount of free volume in the amorphous matrix. (3) The bulk and ribbon alloys exhibit a similar serrated plastic flow during the nanoindentation test. The pile-ups around the indents are observed in both alloys. (4) The hardness and modulus seem to be more sensitive to a cooling rate during solidification than the plastic-flow behavior in the Cu 60 Zr 30 Ti 10. Acknowledgements This work was supported by the National Science Foundation [NSF] International Materials Institutes [IMI] Program [DMR- 0231320], the Combined Research and Curriculum Development [CRCD] Program [EEC-0203415] and the Integrative Graduate Education and Research Training [IGERT] Program [DGE-9987548] with Dr. C. Huber, Ms. M. Poats, Dr. C. J. Van Hartesveldt, Dr. D. Dutta, Dr. L. Clesceri, Dr. W. Jennings and Dr. L. Goldberg as the Program Directors. References [1] W.L. Johnson, MRS Bull. 24 (1999) 42 56. [2] A. Inoue, B.L. Shen, H. Koshiba, H. Kato, A.R. Yavari, Nat. Mater. 2 (2003) 661 663.

354 W.H. Jiang et al. / Materials Science and Engineering A 430 (2006) 350 354 [3] W.H. Peter, P.K. Liaw, R.A. Buchanan, C.T. Liu, C.R. Brooks, J.A. Horton, C.A. Carmichael, J.L. Wright, Intermetallics 10 (2002) 1125 1129. [4] W.H. Peter, R.A. Buchanan, C.T. Liu, P.K. Liaw, J. Non-Cryst. Solids 317 (2003) 187 192. [5] G.Y. Wang, P.K. Liaw, A. Peker, B. Yang, M.L. Benson, W. Yuan, W.H. Peter, L. Huang, A. Freels, R.A. Buchanan, C.T. Liu, C.R. Brooks, Intermetallics 13 (2005) 429 435. [6] G.Y. Wang, P.K. Liaw, W.H. Peter, B. Yang, M. Freels, Y. Yokoyama, M.L. Benson, B.A. Green, T.A. Saleh, R.L. McDaniels, R.V. Steward, R.A. Buchanan, C.T. Liu, C.R. Brooks, Intermetallics 12 (2004) 1219 1227. [7] H.H. Liebermann, in: F.E. Luborsky (Ed.), Amorphous Metallic Alloys, Butterworths, London, 1983, pp. 26 41. [8] E. Illeková, M. Jergel, P. Duhaj, A. Inoue, Mater. Sci. Eng. A. 226-228 (1997) 388 392. [9] O.P. Bobrov, V.A. Khonik, K. Kitagawa, S.N. Laptev, J. Non-Cryst. Solids 342 (2004) 152 159. [10] C.A. Schuh, T.G. Nieh, J. Mater. Res. 19 (2004) 46 57. [11] A. Inoue, W. Zhang, T. Zhang, K. Kurosaka, Mater. Trans. 42 (2001) 1149 1151. [12] D.C.C. Lam, A.C.M. Chong, Mater. Sci. Eng. A. 318 (2001) 313 319. [13] M.J. Mayo, R.W. Siegel, A. Narayanasami, W.D. Nix, J. Mater. Res. 5 (1990) 1073 1082. [14] A. Slipenyuk, J. Eckert, Scr. Mater. 50 (2004) 39 44. [15] C. Nagel, K. Rätzke, E. Schmidke, J. Wolf, U. Geyer, F. Faupel, Phys. Rev. B. 57 (1998) 10224 10227. [16] W.H. Jiang, M. Atzmon, Acta Mater. 53 (2005) 3469 3477. [17] A. Concustell, G. Alcalá, S. Mato, T.G. Woodcock, A. Gebert, J. Eckert, M.D. Baró, Intermetallics 13 (2005) 1214 1219. [18] Z.H. Zhang, J.X. Xie, Mater. Sci. Eng. A. 407 (2005) 161 166. [19] U. Ramamurty, M.L. Lee, J. Basu, Y. Li, Scr. Mater. 47 (2002) 107 111. [20] P. Murah, U. Ramamurty, Acta Mater. 53 (2005) 1467 1478. [21] W.H. Jiang, M. Atzmon, J. Mater. Res. 18 (2003) 755 757. [22] W.J. Wright, R. Saha, W.D. Nix, Mater. Trans. JIM 42 (2001) 642 649. [23] W.J. Wright, R.B. Schwarz, W.D. Nix, Mater. Sci. Eng. A 319-321 (2001) 229 232. [24] W.H. Jiang, G.J. Fan, F.X. Liu, G.Y. Wang, H. Choo, P.K. Liaw, J. Mater. Res., in press.