Zigzag magnetic order in the Kitaev spin-liquid candidate material ${\mathrm{RuBr}}_3$ with a honeycomb lattice

We successfully synthesize ${\mathrm{RuBr}}_3$, which is a candidate material of a Kitaev spin liquid, via a high-pressure synthesis. The material ${\mathrm{RuBr}}_3$ possesses a ${\mathrm{BiI}}_3$-type structure (space group $R\overline{3}$) where ${\mathrm{Ru}}^{3+}$ forms an ideal honeycomb lattice even at room temperature and it does not show a structural transition at low temperatures. ${\mathrm{RuBr}}_3$ has a negative Weiss temperature and it undergoes a zigzag antiferromagnetic transition at $T_{\mathrm{N}}=34$ K, as does $\alpha \text{−}{\mathrm{RuCl}}_3$ $(T_{\mathrm{N}}=7–14\mathrm{K}\mathrm{)}$ which is a promising candidate for a Kitaev spin liquid. Although both compounds of ${\mathrm{RuBr}}_3$ and $\alpha \text{−}{\mathrm{RuCl}}_3$ have a zigzag magnetic order at low temperature, the magnetic order in ${\mathrm{RuBr}}_3$ is more robust than in $\alpha \text{−}{\mathrm{RuCl}}_3$. Our results indicate that the Kitaev and non-Kitaev interactions can be modified in ruthenium trihalides by changing the ligand sites, and will provide another platform for exploring Kitaev spin liquids.

Figure 1

(a) Schematic of crystal and magnetic structures refined from the neutron diffraction pattern. This figure is drawn using vesta [53]. $\alpha$ is the angle of the magnetic moment from the honeycomb plane, and $J_3$ is the third-nearest-neighbor Heisenberg interaction. (b) Neutron diffraction pattern collected at 300 K by the backscattering (BS) bank. The red dots, cyan curve, and blue curve represent the observed intensities, calculated intensities, and difference between them. The difference is shifted for clarity. The vertical black bars indicate the peak position expected from the space group $R\overline{3}$. (c) Neutron diffraction patterns focused on the magnetic reflections. Intensities at 3 and 100 K are represented by the red and black dots, respectively. The upper and lower panels are data collected by the quarter-angle (QA) and low-angle (LA) detector banks, respectively.

Figure 2

(a) Temperature ($T$) dependence of the resistivity ($\rho$). The inset shows the Arrhenius plot, and the fitting result is shown by the red line. (b) Temperature dependence of the magnetic susceptibility ($\chi$) under a magnetic field of ${\mu }_{0}H=1$ T. The inset shows the inverse of $\chi$. The data at $200–300$ K are fitted with the Curie-Weiss law, and the fitting results are shown by the red line.

Figure 3

(a)–(c) Temperature ($T$) dependences of (a) the magnetic susceptibility ($\chi$) and temperature differential of $\chi (d\chi /dT)$ under a magnetic field of ${\mu }_{0}H=1$ T, (b) the specific heat ($C$) under a zero magnetic field, and (c) the ${}^{79}\mathrm{Br}$ and ${}^{81}\mathrm{Br}$ nuclear spin-lattice relaxation rate $1/T_1$. In the inset of (b), $C/T$ is plotted as a function of $T^2$. The solid curves in (c) are guides for the eye.

Figure 4

(a), (b) NQR spectra of (a) ${}^{79}\mathrm{Br}$ and (b) ${}^{81}\mathrm{Br}$ above the antiferromagnetic transition temperature. (c) Zero-field Br and Ru nuclear magnetic resonance (NMR) spectra in the antiferromagnetically ordered state. A spectrum below 75 MHz comes from a ${}^{99}\mathrm{Ru}$ NMR spectrum.