Elastoresistance as probe of strain-controlled entropy from Kondo scattering

TL;DR

This work uses symmetry-resolved elastoresistance to probe how in-plane strain controls Kondo scattering in the heavy-fermion lattice YbRh2Si2. By combining longitudinal and transverse resistivity under uniaxial strain along [100] and [110], the authors decompose the response into $A_{1g}$, $B_{1g}$, and $B_{2g}$ channels, finding a dominant $A_{1g}$ channel with negligible symmetry-breaking components, signaling no nematic fluctuations. The $A_{1g}$ elastoresistance shows strong, temperature- and field-dependent behavior that tracks the strain derivative of magnetic entropy via Fisher-Langer scaling with in-plane thermal expansion, implying strain tunes the Kondo temperature and can drive quantum-criticality upon cooling. Correcting for out-of-plane compression and finite strain-transfer, the study provides a robust framework linking elastoresistance to magnetic entropy in a Kondo lattice, offering a non-disordered route to explore quantum criticality at milli-Kelvin temperatures.

Abstract

Heavy-fermion metals are prototype correlated electron systems for the study of Kondo entanglement and quantum criticality. We use the symmetry decomposed elastoresistance to uncover the fingerprints of strain-dependent Kondo scattering as function of temperature and magnetic field in the prototypical tetragonal Kondo lattice YbRh$_2$Si$_2$. By combining longitudinal and transverse resistance measurements under uniaxial strain applied along the tetragonal $[100]$ and $[110]$ directions, we obtain the elastoresistive responses in the $A_{1g}$, $B_{1g}$, and $B_{2g}$ symmetry channels. While the responses in the symmetry-breaking channels are negligible, the isotropic $A_{1g}$ elastoresistance displays characteristic sign changes and approaches huge values at low temperatures. Scaling analysis and comparison with linear thermal expansion measurements reveals that the elastoresistance probes the contribution of Kondo scattering to the strain dependence of magnetic entropy and signals strain-controlled quantum criticality upon cooling to 2 K.

Figures (10)

• Figure 1: Schematic temperature (a) and field dependence (b) of the electrical resistance $\rho$ of the Kondo lattice Cox1988Schlottmann1983Batlogg1987, indicated by black lines. The blue and green lines indicate the effect of an increase and decrease of $T_{\rm K}$, respectively. The elastoresistance $(d\rho/d\varepsilon)/\rho$ (in red, with right y-axis in (a) and left y-axis in (b)) is estimated under the assumption of a decrease/increase of $T_{\rm K}$ under compression/tensile strain $\varepsilon$. Note the sign change in (a) and positive elastoresistance under field in (b).
• Figure 2: Longitudinal strain-induced change in electrical resistance $\Delta R/R_0$ for YRS at various temperatures for strain and current parallel to the $[100]$ (a) and $[110]$ (b) directions. $R_0$ denotes the resistance without applied strain.
• Figure 3: a): Temperature dependence of the relative resistance change under strain $d(\Delta R/R_0)/d\varepsilon$ along the [100] and [110] directions, at zero field and under applied magnetic fields. (b): Magnetoresistance $\Delta R(B)/R(0)$ for current along the [100] and [110] directions at 2 K.
• Figure 4: Symmetry decomposition of elastoresistance. (a) and (b) depict the strain derivatives of the longitudinal and transverse resistance for uniaxial strain applied along the [100] and [110] directions, respectively. (c) and (d) depict the sum and difference of longitudinal and transverse strain derivatives along [100] and [110], respectively. (e) Temperature dependence of the elastoresistance coefficients $m_{A1g}$, $m_{B1g}$, and $m_{B2g}$, calculated from the data in (c) and (d) using eqs. (S2)-(S5) supple. (f) Schematic illustration of $A_{1g}$, $B_{1g}$, and $B_{2g}$ strain symmetry channels in YRS.
• Figure 5: In-plane thermal expansion coefficient $\alpha(T)$ vs $T$ (on log scale) for YRS measured along the [100] and [110] directions, indicated by red and black squares, respectively (left y-axis). The red and black circles display the function $\int[(dT/T)(d(m_{A1g}R/R_{300\rm K})/dT)]$ vs $T$ (right y-axis) with strain along the [100] (red) and [110] (black) direction. The inset displays the same thermal expansion data as $\alpha/T$ vs $T$ (left y-axis) together with $T^{-1}\int[(dT/T)(d(m_{A1g}R/R_{300\rm K})/dT) +const.]$ vs $T$ with the integration constant as determined in supple.