Applicability of Pre-Plastic Deformation Method for Improving Mechanical Properties of Bulk Metallic Glasses

Pre-plastic deformation (PPD) treatments on bulk metallic glasses (BMGs) have previously been shown to be helpful in producing multiple shear bands. In this work, the applicability of the PPD approach on BMGs with different Poisson’s ratios was validated based on experimental and simulation observations. It was found that for BMGs with high Poisson’s ratios (HBMGs, e.g., Zr56Co28Al16 and Zr46Cu46Al8), the PPD treatment can easily trigger a pair of large plastic deformation zones consisting of multiple shear bands. These PPD-treated HBMGs clearly display improved strength and compressive plasticity. On the other hand, the mechanical properties of BMGs with low Poisson’s ratios (LBMG, e.g., Fe48Cr15Mo14Y2C15B6) become worse due to a few shear bands and micro-cracks in extremely small plastic deformation zones. Additionally, for the PPD-treated HBMGs with similar high Poisson’s ratios, the Zr56Co28Al16 BMG exhibits much larger plasticity than the Zr46Cu46Al8 BMG. This phenomenon is mainly due to more defective icosahedral clusters in the Zr56Co28Al16 BMG, which can serve as nucleation sites for shear transformation zones (STZs) during subsequent deformation. The present study may provide a basis for understanding the plastic deformation mechanism of BMGs.


Introduction
In past decades, bulk metallic glasses (BMGs) have attracted widespread interest due to their outstanding mechanical properties [1][2][3][4][5][6], e.g., large elastic limit, high strength, high hardness, excellent fracture toughness, good corrosion resistance, good wear resistance, etc. However, due to the characteristics of long-range disorders and short-range orders in homogeneous macrostructures of BMGs, a highly localized plastic deformation via shear bands usually governs mechanical properties, resulting in catastrophic room-temperature brittleness [1][2][3][4][5][6], severely limiting future practical applications as engineering materials. In order to overcome catastrophic fracture and improve the plastic deformability of BMGs, various approaches have been suggested [7][8][9][10][11][12][13][14][15], which are classified into three basic strategies: The first is to create structural heterogeneity by tailoring chemical compositions, cooling rates, cryogenic thermal cycling, elastic loading, and so on [7,8]. The second is to introduce the second crystalline phase into a glassy matrix to prepare BMG composites by heating, manipulating the solidification process, or altering chemical compositions [9][10][11][12][13][14][15]. Among them, the BMG composites containing metastable B2 crystals with transformationinduced plasticity exhibit remarkable tensile plasticity and high strength with an obvious work-hardening effect [11][12][13][14][15]. The third is to change the deformation conditions or artificially produce macroscopic defects (e.g., notches, holes, shear bands) [16][17][18]. These Uniaxial compression tests were conducted using an electronic tensile testing machine under a strain rate of 2.5 × 10 −4 s −1 . The dimension of the specimens for compression was Φ2 mm × 4 mm. The shear bands induced by PPD treatments were observed, and the surface morphologies after fracture were also investigated using SEM. Furthermore, Uniaxial compression tests were conducted using an electronic tensile testing machine under a strain rate of 2.5 × 10 −4 s −1 . The dimension of the specimens for compression was Φ2 mm × 4 mm. The shear bands induced by PPD treatments were observed, and the surface morphologies after fracture were also investigated using SEM. Furthermore, Young's modulus was measured using a nanoindentation device (Anton Paar CSM-NHT 2 , Ashland, VA, USA). During nanoindentation, a maximum force of 50 mN was achieved Materials 2022, 15,7574 4 of 14 at a loading rate of 100 mN/min with a duration time of 10 s. Furthermore, the atomic structural features of the Zr 56 Co 28 Al 16 BMG were observed by the ab initio molecular dynamics simulations using the Vienna ab initio simulation package (VASP) based on the density functional theory (DFT). The detailed simulation processing can be found in Reference [33].  (Figure 1a,f), which display a very large hemispherical shape. Obviously, the deformation morphology is governed by the radial shear bands, which is very similar to the plastic flow zone achieved during Vickers indentation for BMGs [34]. The enlarged SEM images of these plastic deformation regions (i.e., R1-R4) are shown in Figure 1b,c,e,f. It is clearly seen that the plastic deformation zone consists of two types of shear bands, i.e., relatively long semi-circular (red dotted lines) shear bands and dense wavy (red arrows) shear bands that appear between semi-circular ones. During deformation, the initial elastic and plastic deformations first concentrate on two tiny plastic zones at both the top and bottom of HBMG rods. When the plastic deformation occurs at both contact regions, the shear bands (yellow dash-dotted lines) initiate at both contact areas with the help of the normal stress upon loading. At the same time, the neighboring regions should bear significant horizontal pressure when the semicircular contact regions gradually become flat. A resultant force does not strictly follow and starts to depart from the loading direction, while the deviation degree increases when the plastic deformation zones become larger. As a result, shear bands will not be straight but semi-circular. Meanwhile, these shear bands also interact with each other, leading to the appearance of dense wavy shear bands. On the contrary, only very small plastic deformation zones (green dotted circle) appear at the top and bottom of the Fe 48 Cr 15 Mo 14 Y 2 C 15 B 6 LBMG ( Figure 2g). Take the plastic deformation zone R5 for an example, and the corresponding enlarged view was shown in Figure 2h. The initial semi-circular shear bands (dotted red lines) appear and accumulate at the top and bottom of LBMG rods. The subsequent shear banding propagation does not occur after the LBMG is subjected to a loading displacement of~0.1 mm. However, some micro-cracks appear along the induced shear bands (solid yellow arrows in Figure 2i). These observations indicate that it becomes more difficult for LBMGs to be deformed than HBMGs. The green dotted lines represent the plastic deformation zone induced by PPD treatments. The yellow dot-dash lines represent the semi-circular shear bands developed from the initial shear bands induced by PPD treatments. In contrast, the red dotted lines represent the induced semi-circular shear bands during the later stage of PPD treatments. The solid red arrows represent the dense wavy shear bands between semi-circular ones, and the dotted yellow arrows represent micro-cracks. Figure 3 shows the mechanical properties of the Zr46Cu46Al8, Zr56Co28Al16, and Fe48Cr15Mo14Y2C15B6 BMGs in the as-cast state and after PPD treatments. Before PPD treatments, the mechanical properties are different even though the as-cast HBMGs with a similar high Poisson's ratio. The yield strength, maximum compressive strength, and plastic strain of the as-cast Zr56Co28Al16 HBMG are measured to be 2115 ± 10 MPa, 2168 ± 20 MPa, and ~4.4%, while those of the as-cast Zr46Cu46Al8 HBMG are 1992 ± 10 MPa, 1995 ± 20 MPa, and ~0.2%, respectively (Figure 3a,b and Table 1). Moreover, the as-cast Fe48Cr15Mo14Y2C15B6 LBMG shows a representative brittle behavior, and no plasticity is observed (Figure 3c), whose fracture strength is 3239 ± 20 MPa. However, as shown in Figure 3a,b, the mechanical properties of both Zr46Cu46Al8 and Zr56Co28Al16 HBMGs after PPD treatments are enhanced. After PPD treatments, the yield strength and maximum compressive strength of the Zr56Co28Al16 HBMG are enhanced to be 2122 ± 15 MPa and 2420 ± 30 MPa, while those of the Zr46Cu46Al8 HBMG are improved to be 1992 ± 15 MPa and 2120 ± 30 MPa, respectively (Figure 3a,b and Table 1). It is worth noting that the enhancement of yield strength of both Zr46Cu46Al8 and Zr56Co28Al16 HBMGs is very limited, The green dotted lines represent the plastic deformation zone induced by PPD treatments. The yellow dot-dash lines represent the semi-circular shear bands developed from the initial shear bands induced by PPD treatments. In contrast, the red dotted lines represent the induced semi-circular shear bands during the later stage of PPD treatments. The solid red arrows represent the dense wavy shear bands between semi-circular ones, and the dotted yellow arrows represent micro-cracks.  Table 1). Moreover, the as-cast Fe 48 Cr 15 Mo 14 Y 2 C 15 B 6 LBMG shows a representative brittle behavior, and no plasticity is observed (Figure 3c), whose fracture strength is 3239 ± 20 MPa. However, as shown in Figure 3a,b, the mechanical properties of both Zr 46 16 HBMGs is very limited, approximately 0.33% and 0.25%, respectively. However, the maximum compressive strength is distinctly improved, while the increased proportion is about 11.6% and 6.3% for the Zr 46 Cu 46 Al 8 and Zr 56 Co 28 Al 16 HBMGs, respectively. It has been elucidated that the enhanced strength should be attributed to the residual stress upon pre-deformation [35,36].

Mechanical Properties of BMGs in As-Cast States and after PPD Treatments
strength is distinctly improved, while the increased proportion is about 11.6% and 6.3% for the Zr46Cu46Al8 and Zr56Co28Al16 HBMGs, respectively. It has been elucidated that th enhanced strength should be attributed to the residual stress upon pre-deformation [35,36].
Moreover, the enhancement of plasticity is also remarkable for both HBMGs. Th plastic strain of the Zr56Co28Al16 HBMG increases from ~4.4% to ~21.8%, while that of th Zr46Cu46Al8 HBMG rises from 0.2% to ~3.5%. Even though both BMGs show a similar Pois son's ratio, the improved mechanical performance of the Zr56Co28Al16 HBMG is far mor significant than that of the Zr46Cu46Al8 HBMG after PPD treatments. On the other hand the mechanical properties of the Fe48Cr15Mo14Y2C15B6 LBMGs become worse after PPD treatments, whose fracture strength decreases to 2909 ± 30 MPa without detectable plas ticity.

Fracture Morphologies
As shown in Figure 4a, only several shear bands are observed along the fracture plane for the as-cast Zr 46 Cu 46 Al 8 HBMG, while multiple shear bands appear after PPD treatments (Figure 4b). The newly formed shear bands initiate from rod surfaces where the PPD treatment was conducted (region A in Figure 4b) and then propagate parallel to the main shear band. For the as-cast Zr 56 Co 28 Al 16 HBMG, a lot of shear bands appear after failure without bearing PPD treatments (Figure 4c). After PPD treatment, more multiple shear bands form in the Zr 56 Co 28 Al 16 HBMG than in the Zr 46 Cu 46 Al 8 HBMG. More importantly, some shear bands cross with each other along~43º, while other dense shear bands perpendicular to the loading direction exist between the crossed shear bands (Figure 4d), implying that the newly formed shear bands induced by PPD treatments play a vital role in the subsequent plastic deformation.

Fracture Morphologies
As shown in Figure 4a, only several shear bands are observed along the fracture plane for the as-cast Zr46Cu46Al8 HBMG, while multiple shear bands appear after PPD treatments (Figure 4b). The newly formed shear bands initiate from rod surfaces where the PPD treatment was conducted (region A in Figure 4b) and then propagate parallel to the main shear band. For the as-cast Zr56Co28Al16 HBMG, a lot of shear bands appear after failure without bearing PPD treatments (Figure 4c). After PPD treatment, more multiple shear bands form in the Zr56Co28Al16 HBMG than in the Zr46Cu46Al8 HBMG. More importantly, some shear bands cross with each other along ~43º , while other dense shear bands perpendicular to the loading direction exist between the crossed shear bands (Figure 4d), implying that the newly formed shear bands induced by PPD treatments play a vital role in the subsequent plastic deformation. However, for the Fe48Cr15Mo14Y2C15B6 LBMGs, the samples are broken down into pieces and exhibit typical brittle features ( Figure 5) in the as-cast state and after PPD treatments. These observations imply that the plastic deformation of the Fe48Cr15Mo14Y2C15B6 LBMGs cannot be affected by the pre-existing shear bands induced by PPD treatment and is still governed by crack propagation. The existence of micro-cracks induced by PPD treatments can provide crack-initiation sites for subsequent crack-linking and propagation.

Fracture Morphologies
As shown in Figure 4a, only several shear bands are observed along the fracture plane for the as-cast Zr46Cu46Al8 HBMG, while multiple shear bands appear after PPD treatments (Figure 4b). The newly formed shear bands initiate from rod surfaces where the PPD treatment was conducted (region A in Figure 4b) and then propagate parallel to the main shear band. For the as-cast Zr56Co28Al16 HBMG, a lot of shear bands appear after failure without bearing PPD treatments (Figure 4c). After PPD treatment, more multiple shear bands form in the Zr56Co28Al16 HBMG than in the Zr46Cu46Al8 HBMG. More importantly, some shear bands cross with each other along ~43º , while other dense shear bands perpendicular to the loading direction exist between the crossed shear bands (Figure 4d), implying that the newly formed shear bands induced by PPD treatments play a vital role in the subsequent plastic deformation. However, for the Fe48Cr15Mo14Y2C15B6 LBMGs, the samples are broken down into pieces and exhibit typical brittle features ( Figure 5) in the as-cast state and after PPD treatments. These observations imply that the plastic deformation of the Fe48Cr15Mo14Y2C15B6 LBMGs cannot be affected by the pre-existing shear bands induced by PPD treatment and is still governed by crack propagation. The existence of micro-cracks induced by PPD treatments can provide crack-initiation sites for subsequent crack-linking and propagation.

Discussion
Compared to the Fe 48 Cr 15 Mo 14 Y 2 C 15 B 6 LBMG, multiple shear bands govern the whole plastic deformation of both Zr 46 Cu 46 Al 8 and Zr 56 Co 28 Al 16 HBMGs. As we know, the formation of shear bands is caused by the percolation of a serial of the activated shear transformation zones (STZs) [1][2][3][37][38][39]. The STZ is one local cluster of atoms that experiences inelastic shear distortion by conquering the activation energy barrier between two different energy configurations [1][2][3][37][38][39]. During deformation, the nucleation of a shear band is believed to be controlled by the activation of local STZs at looser atomic structures. The activation of local STZs proceeds through two steps [1][2][3][37][38][39][40], i.e., a flow-induced dilatation that broadens homogeneously during the early stage and the localization of plastic strain into a narrow shear band at the expense of shear flow in the surrounding regions. With further deformation, the shear banding propagation is closely linked with the percolated connection and the cooperative shear of a large amount of local STZs. So far, many investigations have been conducted to evaluate STZ sizes or volumes in BMGs [40][41][42], among which a cooperative shear model is widely accepted to calculate them [43]. The prevailing view is that a large STZ volume facilitates the plasticity of BMGs [44][45][46]. Moreover, a high Poisson's ratio or a low ratio of shear modulus (G) and bulk modulus (B) may be responsible for good plasticity of BMGs [47]. The critical Poisson's ratio for the ductile-brittle transition was determined to be 0.31~0.32 [47]. Thus, the STZ volume, Poisson's ratio, and plasticity are usually correlated in different BMGs.
Pan et al. found that the STZ volume increases with increasing Poisson's ratio, leading to the improved plasticity of BMGs [48], while Qiao et al. believed that this rule is no longer applicable [49]. Herein, we collected the reported STZ volumes and Poisson's ratios of BMGs, which are shown in Figure 6a. It is obvious that the STZ volumes roughly increase with increasing Poisson's ratios for the reported BMGs (detailed data listed in the Supplementary Materials). After carefully examining previous observations [42,[50][51][52], the STZ volumes strongly depend on the loading conditions (e.g., strain rates, applied pressure, tip radius, etc.), experimental technologies, calculation methods, and so on. For instance, in Zr-based BMGs, simulations and experimental observations based on the statistical analysis of the first pop-in data give STZ volumes of approximately 0.2~0.4 nm 3 [53], while the STZ volumes calculated based on the rate-dependent hardness data are estimated to be 1~16 nm 3 [46,[54][55][56][57] (Figure 6a). As shown in Figure 6b, when the calculation method is fixed, the STZ volume exhibits a strong loading rate-dependence or tip radius-dependence during nanoindentation. Moreover, the Poisson's ratios of some BMGs are mismeasured, leading to different Poisson's ratios for a same glass-forming composition. As a result, the correlation between the STZ volume and Poisson's ratio does not show a good change tendency. When the calculation methods or experimental processing are fixed, the plasticity indeed increases with increasing STZ volumes (Figure 6c). Furthermore, PPD treatments (e.g., cold rolling) on BMGs indeed show the increase in the released enthalpy prior to the glass transition temperature (T g ) (Figure 6d), which reflects the increase in the free volume. Since the liquid-like regions containing high local free volume provide potential sites for the STZ operation, the increased free volume can cause the easy nucleation and activation of STZs [42,58]. In our case, compared with the Fe 48 Cr 15 Mo 14 Y 2 C 15 B 6 LBMG, both Zr 46 Cu 46 Al 8 and Zr 56 Co 28 Al 16 LBMGs exhibit a higher Poisson's ratio and then a potential larger STZ volume, leading to the easier initiation of shear bands during PPD treatments. Therefore, more STZs will be easily activated in the PPD-treated Zr 46 Cu 46 Al 8 and Zr 56 Co 28 Al 16 glassy matrix during compression, especially in the previous plastically deformed regions. As a result, more multiple shear bands appear, resulting in enhanced plasticity.

Figure 6. (a) Correlation between the STZ volumes and Poisson's ratios in different systems BMGs
(Red arrow indicates the increasing tendency), (b) the change in the STZ volumes with increasing loading rate or tip radius during nanoindentation [42,53], (c) correlation between the STZ volumes and plastic strains [54,59], and (d) the increased enthalpy before Tg after cold rolling [60,61].

Conclusions
In this work, three types of BMGs with different Poisson's ratios, i.e., Zr 46 16 BMGs with high Poisson's ratios, a pair of larger plastic deformation zones can be induced at loading positions after pre-plastic deformation. Many long semi-circular shear bands appear within both plastic deformation zones, while some dense wavy shear bands emerge between these long shear bands. The dominant reason is that the resultant force changes continuously when the pre-plastic deformation proceeds. For the Fe 48 Cr 15 Mo 14 Y 2 C 15 B 6 BMG with a high Poisson's ratio, only very small plastic deformation zones are caused after the pre-plastic deformation. Within these small zones, only a few semi-circular shear bands appear, accompanied by several micro-cracks.
Then the room-temperature compression tests were conducted on these pre-deformed BMG rods to measure their mechanical properties. It was observed that those BMGs with high Poisson's ratios show higher strength and larger plasticity after pre-plastic deformation. However, the Fe 48 Cr 15 Mo 14 Y 2 C 15 B 6 BMG with a high Poisson's ratio after the pre-plastic deformation exhibits worse mechanical properties than the as-cast state, which should be attributed to the relatively small STZ volume because of a low Poisson's ratio. Additionally, despite the Poisson's ratios of the Zr 56 Co 28 Al 16 and Zr 46 Cu 46 Al 8 BMGs being similar, the improvement of mechanical properties of the Zr 56 Co 28 Al 16 BMG is more evident than those of the Zr 46 Cu 46 Al 8 BMG. Molecular dynamics simulation observations indicate that the Co-and Al-centered architectures in the as-cast Zr 56 Co 28 Al 16 BMG contain more defective icosahedral clusters, which can provide more nucleation sites for STZs during deformation and lead to better mechanical properties.

Author Contributions:
The manuscript was written thanks to the contributions of all the authors. C.Z., investigation, methodology, formal analysis, and writing-original draft; H.Z., investigation, methodology; X.Y., data curation, methodology, visualization, writing-original draft; K.S., conceptualization, writing-review and editing, supervision, project administration, funding acquisition; D.L., formal analysis, writing-review and editing. All authors have read and agreed to the published version of the manuscript. Acknowledgments: The authors are grateful to Kunlun Wang and Yanqing Xin for technical assistance from the Physical-Chemical Materials Analytical and Testing Center of Shandong University at Weihai.

Conflicts of Interest:
The authors declare no conflict of interest.