Synthesis of new nickel hydrides at high pressure

We predict two new nickel hydrides, $\mathrm{N}{\mathrm{i}}_2{\mathrm{H}}_3$ ($C2/m$) and $\mathrm{Ni}{\mathrm{H}}_2$ ($I4/mmm$), to be thermodynamically stable at 60 GPa by DFT calculations. The calculated structure of $\mathrm{Ni}{\mathrm{H}}_2$ is similar to the one observed earlier for $\mathrm{Fe}{\mathrm{H}}_2$. However, to the best of our knowledge, the monoclinic structure predicted for $\mathrm{N}{\mathrm{i}}_2{\mathrm{H}}_3$ was never reported in the other hydrides. We successfully synthesized the monoclinic $\mathrm{N}{\mathrm{i}}_2{\mathrm{H}}_3$ phase using a laser-heating technique in a diamond anvil cell at pressures around 60 GPa. The $\mathrm{N}{\mathrm{i}}_2{\mathrm{H}}_3$ phase can be retained down to 17 GPa, and DFT calculations predict it to be nonmagnetic and nonsuperconducting. However, we did not find in our experiments the theoretically predicted $\mathrm{Ni}{\mathrm{H}}_2$ phase. Higher pressures and/or different synthesis conditions might be needed to synthesize this and other polyhydrides of nickel. Our results show that the Ni-H system behaves differently in comparison to the Fe-H system: Ni extends its own chemical identity to pressures as high as 60–80 GPa. This finding may have significant implications for the high-pressure behavior of Fe-Ni alloys at the high-pressure conditions relevant for the Earth and planetary sciences.

Figure 1

(a) The calculated convex hull of the Ni-H system at 60 GPa. NiH, $\mathrm{N}{\mathrm{i}}_2{\mathrm{H}}_3$, and $\mathrm{Ni}{\mathrm{H}}_2$ are thermodynamically stable at 60 GPa. We did not consider the zero-point energy in the calculated formation enthalpy of the Ni-H phases. The representation of structures for the three theoretically predicted nickel hydrides at high pressure: fcc NiH (b), $C2/m-\mathrm{N}{\mathrm{i}}_2{\mathrm{H}}_3$ (c), and $I4/mmm-\mathrm{Ni}{\mathrm{H}}_2$ (d).

Figure 2

The upper panel shows the cake-type image of the two-dimensional x-ray diffraction pattern for $\mathrm{N}{\mathrm{i}}_2{\mathrm{H}}_3$ at 57 GPa with temperature around 200 K. The lower panel shows the Le Bail fit result for the integrated pattern by using the theoretically predicted structure model ($C2/m$): nearly all the peaks can be fit well using a theoretical structural model ($\lambda =0.3344\AA$).

Figure 3

Comparison of the experimental lattice parameters measured around 200 K and the theoretical ones resulting from the DFT calculations for the monoclinic $\mathrm{N}{\mathrm{i}}_2{\mathrm{H}}_3$ phase at various pressures.

Figure 4

The volume per Ni atom as a function of pressure for nickel hydrides measured at 200 K. NiH was measured during the compression process. For $\mathrm{N}{\mathrm{i}}_2{\mathrm{H}}_3$, the DFT calculations (red dashed line) are consistent with the experimental results. The solid lines represent the BM fit results as given in Table 1. The blue dashed line indicates the BM fitting result of NiH data at ambient temperature obtained by Ishimatsu et al. below 4 GPa [29]. The blue and black diamond represent the $V_0$ value for NiH and Ni measured previously.

Figure 5

The calculated band structure and density of states for predicted $C2/m-\mathrm{N}{\mathrm{i}}_2{\mathrm{H}}_3$ structure at 80 GPa. The calculation indicates $\mathrm{N}{\mathrm{i}}_2{\mathrm{H}}_3$ is a metal and its Fermi surface is dominated by the Ni $d$ orbitals.

Figure 6

The calculated phonon dispersion for the predicted $C2/m-\mathrm{N}{\mathrm{i}}_2{\mathrm{H}}_3$ at 80 GPa. The absence of any imaginary frequency modes in the whole Brillouin zone suggests the dynamical stability of $C2/m-\mathrm{N}{\mathrm{i}}_2{\mathrm{H}}_3$ at high pressures.