Micrometer thick single crystal iron-garnet films on a diamagnetic buffer layer for cryogenic applications
A. N. Kuzmichev, P. M. Vetoshko, E. I. Pavluk, A. A. Holin, G. A. Knyazev, A. S. Kaminskiy, S. S. Demirchan, R. Tyumenev, D. S. Kalashnikov, V. S. Stolyarov, V. I. Belotelov
TL;DR
The paper tackles the challenge of achieving ultra-low magnetic damping in YIG films at cryogenic temperatures by isolating interfacial paramagnetic contributions using a diamagnetic buffer layer grown by liquid phase epitaxy on a gadolinium gallium garnet substrate. A Y3(GaScIn)5O12 buffer layer is engineered (with In3+ substitution) to minimize lattice mismatch and suppress Gd diffusion, enabling micrometer-thick, single-crystal YIG films with highly uniform thickness. The authors report record-low ferromagnetic resonance linewidths of about 4.9 MHz at 4 K and 5.9 MHz at 16 mK, along with lattice-mismatch values that keep strain negligible and high-quality interfaces verified by SEM. This diamagnetic-buffer strategy paves the way for integrating high-coherence YIG components into cryogenic quantum devices and scalable spintronic platforms.
Abstract
This work advances the frontier of low-damping magnetic materials, directly addressing the demand for ultra-low-loss components in quantum computing and cryogenic electronics. Here we demonstrate a new approach to get single crystal micrometer-thick yttrium iron garnet (YIG) films with low damping through isolating and mitigating interfacial paramagnetic contributions of a paramagnetic substrate by a buffer-layer. The YIG films with the diamagnetic yttrium scandium gallium garnet buffer layer grown by liquid phase epitaxy on a gadolinium gallium substrate demonstrate homogeneity unprecedented for the thin planar YIG structures, yielding ferromagnetic resonance linewidths of 4.9 MHz at 4 K and 5.9 MHz at 16 mK, the lowest values reported to date. These results underscore the critical role of interfacial engineering in overcoming intrinsic material limitations, opening avenues for further optimization in spin-based technologies.
