Coupling between heavy fermion superconductor CeCoIn$_5$ and antiferromagnetic metal CeIn$_3$ through the atomic interface

Abstract

To study the mutual interaction between unconventional superconductivity and magnetic order through an interface, we fabricate Kondo superlattices consisting of alternating layers of heavy-fermion superconductor CeCoIn$_5$ and antiferromagnetic (AFM) heavy-fermion metal CeIn$_3$. The strength of the AFM fluctuations is tuned by applying hydrostatic pressure to CeCoIn$_5(m)$/CeIn$_3(n)$ superlattices with $m$ and $n$ unit-cell-thick layers of CeCoIn$_5$ and CeIn$_3$, respectively. Superconductivity in CeCoIn$_5$ and AFM order in CeIn$_3$ coexist in spatially separated layers. At ambient pressure, N\'{e}el temperature $T_N$ of the CeIn$_3$ block layers (BLs) of CeCoIn$_5$(7)/CeIn$_3(n)$ shows little dependence on $n$, in contrast to CeIn$_3(n)$/LaIn$_3$(4) superlattices where $T_N$ is strongly suppressed with decreasing $n$. This suggests that each CeIn$_3$ BL is magnetically coupled by the RKKY interaction through the adjacent CeCoIn$_5$ BL and a 3D magnetic state is formed. With applying pressure to CeCoIn$_5$(7)/CeIn$_3$(13), $T_N$ of the CeIn$_3$ BLs is suppressed up to 2.4 GPa, showing a similar pressure dependence as CeIn$_3$ single crystals. An analysis of upper critical field reveals that the superconductivity in the CeCoIn$_5$ BLs is barely influenced by the AFM fluctuations in the CeIn$_3$ BLs, even when the CeIn$_3$ BLs are in the vicinity of the AFM quantum critical point. This is in stark contrast to CeCoIn$_5$/CeRhIn$_5$ superlattice where the superconductivity in the CeCoIn$_5$ BLs is profoundly affected by AFM fluctuations in the CeRhIn$_5$ BLs. The present results show that although AFM fluctuations are injected into the CeCoIn$_5$ BLs from the CeIn$_3$ BLs through the interface, they barely affect the force which binds superconducting electron pairs. These results demonstrate that 2D AFM fluctuations are essentially important for the pairing interactions in CeCoIn$_5$.