Magnetic phase diagram of NiCl2·6H2O: The low-temperature canted-paramagnetic boundary and the bicritical point

The magnetic phase diagram of the antiferromagnetic Ni${\mathrm{Cl}}_2$·6${\mathrm{H}}_2$O ($T_N=5.34\mathrm{}$ K) was determined from $\frac{\mathrm{dM}}{\mathrm{dH}}$ measurements down to temperatures $T$ as low as 0.35 K, and in magnetic fields $H$ up to 150 kG. The $T=0\mathrm{}$ spin-flop field and canted-paramagnetic critical field were determined as $H_{\mathrm{SF}}\mathrm{}\left(0\right)=39.45\mathrm{}\pm 0.30$ kG and $H_c\mathrm{}\left(0\right)=143.9\mathrm{}\pm 1.5$ kG, respectively. From these values, the exchange field $H_E=77.4\mathrm{}\pm 0.9$ kG and the uniaxial anisotropy field $H_A=10.8\mathrm{}\pm 0.3$ kG were obtained. The corresponding single-ion parameters are $\frac{z\left|J\right|\mathrm{}}{k_B}=11.5\mathrm{}\pm 0.3$ K and $\frac{\mathrm{}\left|D\right|\mathrm{}}{k_B}=1.61\mathrm{}\pm 0.07$ K. These values are compared with the ones obtained previously from other experiments. A careful analysis of the temperature dependence of the canted-paramagnetic critical field $H_c$ is presented. For $T\lesssim 0.25T_N$, the $T^{\frac{3}{2}}$ dependence predicted by Green's-function and spin-wave calculations is observed. For $0.3T_N\lesssim T\lesssim 0.6T_N$, the observed $T$ dependence is approximately $T^{\frac{5}{2}}$, similar to the results of Rives et al. in Mn${\mathrm{Cl}}_2$·4${\mathrm{H}}_2$O and Co${\mathrm{Cl}}_2$·6${\mathrm{H}}_2$O for comparable reduced temperatures. Our results suggest that temperatures above $0.3T_N$ are too high for comparing the $T$ dependence of $H_c$ with spin-wave theory. Near the bicritical point ($H_b=44.22\mathrm{}$ kG and $T_b=3.940\mathrm{}$ K) the boundaries were determined with higher precision, and compared with the recent predictions based on scaling and renormalization-group theories. Colinearity between $\overrightarrow{\mathrm{H}}$ and the easy axis was achieved to about 0.1 degrees. Least-squares fits indicated good overall agreement with the theoretical predictions for the case where the number of critical spin components is $n=2\mathrm{}$.