Physical properties of the magnetically frustrated very-heavy-fermion compound ${\mathrm{YbCu}}_4\mathrm{Ni}$

The physical properties of the very-heavy-fermion compound ${\mathrm{YbCu}}_4\mathrm{Ni}$ were characterized through structural, magnetic, thermal, and transport studies along nearly four decades of temperature ranging between 50 mK and 300 K. At high temperature, the crystal electric field level splitting was determined with ${\mathrm{\Delta }}_1\left({\mathrm{\Gamma }}_6\right)=85\mathrm{K}$ and ${\mathrm{\Delta }}_2\left({\mathrm{\Gamma }}_8\right)\approx 200\mathrm{K}$, the latter being a quartet in this cubic symmetry. An effective magnetic moment ${\mu }_{\mathrm{eff}}\approx 3{\mu }_B$ is evaluated for the ${\mathrm{\Gamma }}_7$ ground state, while at high temperature the value for a ${\mathrm{Yb}}^{3+}$ ion is observed. At low temperature this compound shows the typical behavior of a magnetically frustrated system undergoing a change of regime at a characteristic temperature $T^{*}\approx 200\mathrm{mK}$ into of Fermi-liquid-type “plateau” of the specific heat: $C_m{/T|}_{T\to 0}$ = const. The change in the temperature dependence of the specific heat coincides with a maximum and a discontinuity in respective inductive and dissipative components of the ac susceptibility. More details about the nature of this ground state are revealed by the specific heat behavior under applied magnetic field.

Figure 1

(a) The experimental x-ray diffraction powder pattern of ${\mathrm{YbCu}}_4\mathrm{Ni}$ compared with the calculated diffraction diagram. The experimental data are shown by symbols, whereas the line through the data represents the results of the Rietveld refinement. The lower curve is the difference curve. The ticks indicate Bragg peak positions. (b) Crystal structure seen as edge-sharing tetrahedra with Yb ions located at the vertices (unit cell indicated by black lines) after [15].

Figure 2

(a) High-temperature dependence of the inverse magnetic susceptibility and (b) low-temperature inductive component of ac susceptibility scaled with high-temperature dc susceptibility in a double logarithmic representation. Inset: inductive (${\chi }^{\prime }$, left axis) and dissipative (${\chi }^{″}$, right axis) components of ac susceptibility measured at $f=4$ kHz in a semilogarithmic representation.

Figure 3

Magnetization measurements versus field up to 9 T performed between 2 K and room temperature.

Figure 4

Logarithmic temperature dependence of the electrical resistivity within four decades of temperature at $B=0$ and 6 T, measured in different cryostats. The dashed vertical line indicates the data matched at $T=4\mathrm{K}$ using a ratio of 1.8 in the form factor (see the text). The straight purple line remarks the logarithmic $T$ dependence in the range $2<T<10\mathrm{K}$. The $\rho (T\le 4,B=2T)\mathrm{K}$ data are included to show the $T_{\max }^{\rho }$ variation with field.

Figure 5

Specific heat temperature dependencies up to room temperature of ${\mathrm{YbCu}}_4\mathrm{Ni}$ and ${\mathrm{LuCu}}_4\mathrm{Ni}$. Inset: high-temperature magnetic contribution and a fit accounting for the CEF levels and GS contributions.

Figure 6

Low-temperature dependencies of the specific heat of ${\mathrm{YbCu}}_4\mathrm{Ni}$ under different magnetic fields, 0–6 T. Inset: field dependence of the temperature of respective $C_m\left(T\right)$ maxima $T^{C_{\max }}$. The dashed line shows values computed applying the $T^{C_{\max }}\propto B$ formula for the ${\mathrm{\Gamma }}_7$ doublet (see the text), and the solid curve shows the actual $C_m(T,B)$ observed maxima.

Figure 7

The temperature dependence of the magnetic entropy of ${\mathrm{YbCu}}_4\mathrm{Ni}$ in a logarithmic temperature scale. The inset shows the entropy gain with different applied fields, in linear temperature scale.

Figure 8

Comparison between measured $C_m\left(T\right)/T$ and fitted $C_{\mathrm{fit}}{/T|}_{T>T^{*}}$ (solid curve) dependencies referred to the left axis in double-logarithmic representation. On the right axis (linear scale), the temperature dependencies of respective entropies are referred, showing how $S_{\mathrm{fit}}\to 0$ at $T>0$. The value of $S_{\mathrm{fit}}(T\to \infty )=\mathrm{R}ln2$ is taken as the reference for the doublet GS. CEF indicates the onset of the excited crystal field level contribution above about 3 K.

Figure 9

(a) Low-temperature thermal dependence of the magnetic contribution to the specific heat $C_m\left(T\right)$. The solid curve shows theoretic prediction for a Kondo $S=1/2$ impurity [25]. The dashed curve shows the Schottky anomaly divided by a factor 2, and the dash-dotted line shows an exemplary spin-glass system [22]. (b) Comparison between the ${\chi }^{″}\left(T\right)$ jump (left axis) and that of the $C_m\left(T\right)$ derivative $\partial C_m/\partial T$ (right axis).