Glass forming ability of Zr-based Zr–Cu–Ni–Al–(Ag) alloys

A series of Zr-based glasses has been obtained by rapid melt quenching and their thermal characteristics (Tg, Tx, Tl) were determined by DSC/DTA. Some of the most recognized glass forming ability (GFA) criteria were applied to predict the glass forming ability of the alloys and to study its dependence on the alloy composition. As a result, the best glass forming compositions among the studied alloys could be selected. It was found that as a general trend an increase of the Zr content results in Tx and ΔTx (= Tx − Tg) decrease. The determined liquidus temperatures of chosen compositions with substantially different ΔTx were found to vary relatively little. Therefore, it was concluded that the variation in the GFA of the studied alloys is mainly due to difficulties in the crystallization process, rather than to different stability of the melts.


Introduction
Zr-based alloys are among the intensively studied multicomponent metallic glasses [1][2][3]. The reasons for this interest are their good mechanical properties included high tensile strength, high bending strength, high fracture toughness, high hardness and fatigue strength, high impact fracture energy, good castability and cutting machinability [3][4][5] as well as high corrosion stability [6][7][8]. They are also known as easy glass formers and appropriate alloys to be prepared in a bulk amorphous form [9]. Thus, having the size in the centimeter-scale, bulk metallic glasses (BMG) allow being developed as good structural and functional materials [3,[10][11][12][13].
The first Pd-based BMG, which belongs to such a group, was prepared at the beginning of 1980s by Turnbull. Later large number of BMG on the base of La, Mg, and Zr were developed at the end of the decade by the group of Inoue [14][15][16] and Johnson [2]. Zr-Al-TM system was also studied, showing a large supercooled liquid region. The multicomponent alloys offer significantly larger glass forming ability due to their relatively lower critical cooling rate and bigger critical glass formation size in comparison to the binary systems investigated. The additional elements improve both, stabilize the liquid phase and prevent the crystallization process. The improvement of GFA in Zr-Al-Ni alloys could be related to an optimum atomic configuration achieved with the dissolution of Al and increased short-range packing density of the amorphous phase [17][18][19].
The glass forming ability is usually expressed by the maximum obtainable diameter of a fully amorphous ingot, called critical diameter [20]. However, the methods for the critical diameter determination are time and materials exhausting. Therefore, many papers were devoted to attempts to correlate the GFA with the thermal characteristics of the alloys and already accepted criteria, such as temperature interval of supercooled liquid region ΔT x = T x − T g [1], reduced glass transition temperature T rg = T g /T l [21], α = T x /T l [22], γ m = (2T x − T g )/T l [23], γ = T x /(T l + T g ]) [24], K = (T x − T g )/ (T l − T x ) [25], δ = T x /(T l − T g ) [26], and ω = (T g /T x − 2T g )/ (T g + T l ) [27]. Some of these criteria take into account the difficulties in the crystallization process of the glasses and others the stability (thermodynamic and rheological properties) of their melts [28][29][30][31][32][33].
Some of the amorphous alloys included in the present work have already been the subject of our previous studies, which mainly traced their crystallization behavior, 1 3 crystalline products (quasicrystalline and equilibrium crystalline phases) at different annealing conditions [34]. For the purpose of the present work a large series of similar in composition Zr-based amorphous alloys has been selected, all obtained under well-controlled and uniform quenching conditions, and on the base of determined thermal characteristics T g , T x , T l using known criteria their GFA has been predicted. Dependence of GFA on the composition of the glasses was sought, as well as the specific heat change ΔC p at glass transition temperature was determined for the whole series of zirconium glasses studied.

Experimental part
The alloy ingots were prepared by arc-melting the mixture of Zr (99.5%), Cu (99.9%), Ni (99.9%), Al (99.99%) and Ag (99.95%) in a pure argon atmosphere approx. 1.2 bar on a water-cooled copper substrate. Homogeneous alloys composition was achieved by 3 times remelting the master alloys. The alloys were then rapidly solidified by melt spinning in a He atmosphere of 300 mbar to obtain ribbons with a thickness of 40 µm and width of 3 mm. The temperature of the melt was between 1100 and 1200 °C. Since oxygen contamination significantly affects the thermal stability [34], special care was taken to ensure comparable conditions for all ribbons produced.
The microstructure of the as-cast samples was characterized by X-ray diffraction (XRD) with Cu-Kα radiation (diffractometer Siemens D 500, at a step of 0.05Θ and counting time 2 s/step). Thermal analysis was carried out by differential scanning calorimetry (Perkin-Elmer DSC7) and differential thermal analysis (Perkin-Elmer DTA). Scanning rate of 10 K.min −1 , sample mass of 15-20 mg and purified nitrogen purging gas were used for all thermal experiments.

Results and discussion
X-ray diffraction analysis reveals the amorphous character of all studied Zr-based alloys. The diffraction patterns of the amorphous alloys, shown in Fig. 1, are ordered according to their composition. Figure 1c shows XRD patterns of Zr alloys containing Ag and Au in their composition. For all studied alloys two clearly observed diffraction halloes with little difference in the position of their maximums could be detected. The differences in the positions of the diffraction maxima are minor due to the small differences in the overall glasses' composition. Therefore, it can be concluded that the most frequently observed interatomic distances in the studied glasses are comparable (2.45 ± 0.04 Å). Nevertheless, some minor differences in the interatomic distance could be detected, e.g., Zr 70 Cu 12 Ni 10.5 Al 7.5 glass is characterized with the smallest mean interatomic distance (2.41 Å), whereas Zr 63.5 Cu 16 Ni 13 Al 7.5 with the largest one (2.48 Å). The observed minor variations can be explained with the difference in the alloy compositions and with the atomic radii of the metals constituting the alloys.
The DSC scans of the Zr-based amorphous alloys are presented in Fig. 2, as the glasses were again grouped in the same order as in Fig. 1. It is noteworthy that despite the relatively similar compositions of the glasses, there are significant differences in their thermal behavior. However, on all thermal curves, the glass transition is clearly detected, as well as the temperature range of the supercooled liquid (T x ÷ T g ) is well determined. The last observation allowed relatively accurate definition of the region of supercooled liquid, as well as of the specific heat difference due to the glass transition, ΔC p . The enthalpies of crystallization were also reliably determined. All the above thermal characteristics of the glasses are presented in Table 1.
Comparing the thermal characteristics of the studied Zr-based glasses and applying the simplest GFA criterion 25  In the first place, the general trend shows crystallization temperature, T x , increase with the decrease in the Zr content of the glasses (with all the conditionality that the concentration of other metals also slightly changes), Fig. 3. Depending on the alloy composition the glass transition temperature, T g , varies within a range of only about 30 °C (360-390 °C), while T x differs significantly more (from 390 to 465 °C). Therefore, the different temperature range of the supercooled liquid (ΔT x = T x − T g ) of the glasses is mainly due to their different crystallization temperatures, and to lesser extend to variations in their glass transition temperatures.
Another notable compositional dependence is related to the Zr glasses containing silver/gold. These amorphous alloys generally show higher T g and lower T x , i.e., smaller ΔT x , compared to the glasses without noble metal, Table 1. Figure 4 reveals the dependence of T g and T x on the Ag/Au content of the studied glasses. While T g does not vary with the Ag concentration, the crystallization temperature of the   Table 1 Glass transition (T g ) and crystallization (T x ) temperatures, liquidus temperatures (T l ) and specific heat change (ΔC p ) at the glass transition, determined by DSC/DTA The last leads to the conclusion that the GFA of the studied Zr-based glasses deteriorates when the silver concentration is increased. The only glass with Au in its composition also reveals lower ΔT x , compared to the Zr glasses that do not contain precious metals. Thus, it can be summarized that the presence of precious metals in the Zr-Cu-Ni-Al glasses results in a GFA decrease. The comparison between the two glasses with the same composition (Zr 68.5 Cu 11 Ag 2 Ni 11 Al 7.5 ) but vitrified by different melt temperatures (1000 and 1100 °C) is also interesting. The melt, cooled from a higher temperature, leads to a glass with higher T g and T x compared to that vitrified from a lower temperature, the difference in T x being greater. This reasonably results in a larger ΔT x for the glass obtained from the melt annealed at 1100 °C, which result can be explained by a better homogenization of the melt at the higher temperature and hence to a impeded nucleation of the crystalline phase(s).
Other dependence that can also be established form the thermal behavior of the studied Zr glasses (without Ag/Au) is related to alloys with the same or very similar Zr and Al content, but with varying Ni and Cu content. These glasses showed similar T g and T x values, and therefore comparable ΔT x value, Table 1.
Although the glass transition is a kinetic phenomenon strongly dependent on the heating rate, and ΔC p is a value at a particular temperature (T g ) it is worth noting the observation that glasses with large ΔT x are generally characterized with larger ΔC p (> 0.35-0.50 J g −1 K −1 ) and the opposite, glasses with smaller ΔT x show also reduced ΔC p (< 0.3 J g −1 K −1 ). To confirm and clarify this fact, a separate systematic measurement of the absolute value of the specific heat at different temperatures, both in the alloys' glassy state and as supercooled liquid, is necessary.
DTA analysis to higher temperatures of Zr-based amorphous alloys, showing very different ΔT x , was also carried out, Fig. 5. Besides, two glasses with the same composition (Zr 68.5 Cu 11 Ag 2 Ni 11 Al 7.5 ) but solidified from different temperature of the melt were also analyzed. This study revealed the liquidus temperature of the alloys, T l (Table 1), and thus allowed the GFA criteria ΔT x /T l or (2T x − T g )/T l , T g /T l , T x /(T g + T l ) related to the overall liquid phase stability (both at undercooled state and at the equilibrium), to be also applied [23,37]. There is an agreement between the result of the applied GFA criteria: glasses with a larger ΔT x follow a larger ΔT x /T l , (2T x − T g )/T l and T x /(T g + T l ), as well (Table 1). Some exception is observed only when we apply the T g /T l criterion for glasses with similar T g and T l , but with a significant difference in T x . Thus, Zr 69.5 Cu 8 Ag 4 Ni 11 Al 7.5 alloy shows a better GFA compared to Zr 68.5 Cu 11 Ag 2 Ni 11 Al 7.5 (T m = 1000 °C), although ΔT x is significantly larger for the latter alloy. Based on the thermal characteristics of the studied Zrbased glasses, it can be concluded that the evaluated higher GFA of certain alloy compositions should be attributed mainly to their higher crystallization temperatures T x , reflecting the crystallization resistance of the glass, than to the increased liquid phase stability.

Conclusions
The thermal characteristics (T g , T x , T l ) of a series of Zrbased amorphous alloys, obtained by rapid quenching (melt spinning), were determined. The specific heat change at glass transition ΔCp was also determined. The alloys were obtained under the same well-controlled conditions of quenching (temperature of the melt, quenching disk size and velocity, He atmosphere), which fact allowed to search for a relationship between the compositional variations and thermal behavior of the glasses. From the obtained thermal parameters and using well accepted glass forming ability (GFA) criteria the compositions with the higher GFA were determined. An answer was also given to the question of what determines the high GFA for these glasses: the difficulties in crystallization or the stability of the melt.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.