Figure 1

Temperature-composition ($T\mathrm{\text{−}}c$) phase diagram for as-quenched ${\mathrm{Ti}}_{50-x}{\mathrm{Ni}}_{50+x}$, bisected into regions of appearance (blue region) and disappearance of $B2\to B{19}^{\prime }$ martensitic transformation; the border composition is slightly above ${\mathrm{Ti}}_{49}{\mathrm{Ni}}_{51}$. The $B2\to B{19}^{\prime }$ transformation, appearing as an enthalpy peak in differential scanning calorimetry (DSC) measurement and resistivity $\rho (T)$ drop in $\rho$ vs $T$ measurement for ${\mathrm{Ti}}_{50}{\mathrm{Ni}}_{50}$, vanishes abruptly for $x>1$. The $\rho (T)$ curve for ${\mathrm{Ti}}_{48}{\mathrm{Ni}}_{52}$ exhibits negative temperature dependence $d\rho /dT<0$, being similar to the case of $R$-phase formation, but there is no signature for a long-range $B2\to R$ transformation. Here WQ stands for “water quenched.”Reuse & Permissions

Figure 2

(a) and (b) show temperature dependence of ac storage modulus $S(\omega ,T)$ and internal friction $Q^{-1}(\omega ,T)$ for the “nontransforming” ${\mathrm{Ti}}_{48.5}{\mathrm{Ni}}_{51.5}$, respectively, at various frequencies (0.1 Hz–10 Hz). Although the DSC and $\rho (T)$ measurements suggest the “absence” of martensitic transformation for such specimen, $S(\omega ,T)$ produces a dip at $T_g$ [as shown in (a)], which has logarithmic frequency dependence and the corresponding $Q^{-1}(\omega ,T)$ exhibits a maximum at lower temperature [as seen in (b)]. By contrast, the storage modulus $S(\omega ,T)$ for transforming ($B2\to B{19}^{\prime }$) specimen (${\mathrm{Ti}}_{49}{\mathrm{Ni}}_{51}$, quenched from 1123 K) shows no frequency dependence of the dip temperature (c). The dip height of $S(\omega ,T)$ decreases by an order of magnitude with increasing Ni content just by 1% (i.e., samples with more chemical disorder), exhibiting influence of disorder towards suppressing the long-range strain ordering (martensitic transformation) (d). Inset of (a) shows the “non-Arrhenius” $T_g$ vs $\ln (\omega )$ curve for ${\mathrm{Ti}}_{48.5}{\mathrm{Ni}}_{51.5}$, fitted well by the Vogel-Fulcher equation, a signature of most glass transitions. All the samples were water quenched (WQ) from high temperature to ensure a homogenous solid solution.Reuse & Permissions

Figure 3

Electron diffraction pattern of $[1\overline{1}1{]}_{B2}$ zone axis for ${\mathrm{Ti}}_{48}{\mathrm{Ni}}_{52}$ is shown as a function of temperature in (a). The diffuse superlattice reflections at incommensurate $1/3$ position (marked by yellow arrow), which is almost invisible at room temperature, become sharper and of higher intensity with lowering temperature; however, do not change below 158 K ($T_g$ for ${\mathrm{Ti}}_{48}{\mathrm{Ni}}_{52}$) and fail to reach commensurate $1/3$ value (which is characteristic of a long-range ordered $R$ phase). The corresponding HREM dark-field images (b), obtained by using the reflections marked by green circle in (a), reveal random distribution of tiny nanodomains of $R$-like phase in $B2$-like matrix. The nanodomains grow to some extent in size but are finally frozen in that configuration.Reuse & Permissions

Figure 4

Modified temperature-composition ($T\mathrm{\text{−}}c$) phase diagram for as-quenched Ni-rich ${\mathrm{Ni}}_{50-x}{\mathrm{Ni}}_{50+x}$, where the “glassy-martensite” phase is mapped in the yellow region, separated by a dashed line from the long-range correlated $B{19}^{\prime }$ martensitic phase (blue region). Two TEM images, one with large domains of $B{19}^{\prime }$ phase (for specimen undergoing long-range strain-ordered $B2\to B{19}^{\prime }$ transformation) and one with tiny nanodomains (for “glassy-martensite” specimen), have been shown on a comparative ground. $x_c$ is the threshold concentration of excess Ni, above which the as-quenched ${\mathrm{Ti}}_{50-x}{\mathrm{Ni}}_{50+x}$ transforms into the glassy phase.Reuse & Permissions