Influence of the structural quality of Pt on the spin Hall magnetoresistance in Pt/YIG hybrids

For the generation and detection of pure spin currents via the (inverse) spin Hall effect, a combination of a non-ferromagnetic metal and an insulating ferrimagnet is used. Thin films of platinum (Pt) grown on yttrium iron garnet (Y3Fe5O12, YIG) often serve as prototype structures [1-3]. Recently, the influence of the interface quality between the two materials on the spin Hall effect came into focus and the importance of a surface treatment of YIG was discussed [3-7]. We present a systematic study of Pt thin film growth on YIG single crystals. The YIG crystals were exposed to different surface treatments prior to the Pt deposition. Pt thin films were grown under UHV conditions at different substrate temperatures by thermal evaporation utilizing the JCNS molecular beam epitaxy system which is also open to users in the framework of neutron experiments performed at the MLZ (www.mlz-garching.de). Surface science methods like in-situ Auger-electron-spectroscopy, reflection high/low energy electron diffraction, ex-situ x-ray reflectivity and atomic force microscopy were utilized for sample quality control. The spin Hall magnetoresistance is characterized via magnetization orientation dependent resistance measurements in a superconducting magnet cryostat. Best Pt growth results are achieved when the YIG substrate is annealed prior to thin film growth. Growth at room temperature yields higher film roughness while at higher temperatures the risk of Pt oxidation increases. At about 900°C substrate temperature interdiffusion takes place. Optimum Pt thin film growth is found at about 500°C where a texture of the Pt film is observed. In our presentation we discuss the influence of the YIG crystal surface treatment and different growth parameters on the quality of the Pt films in detail and address their impact on the spin Hall magnetoresistance effect.References[1] Sun Y., Chang H., Kabatek M., Song Y.-Y., Wang Z. et al., Phys. Rev. Lett. 111, 106601 (2013)[2] Nakayama H., Althammer M., Chen Y.-T., Uchida K., Kajiwara Y., et al., Phys. Rev. Lett. 110, 206601 (2013).[3] Weiler M., Althammer M., Czeschka F. D., Huebl H., Wagner M. S. et al., Phys. Rev. Lett. 108, 106602 (2012).[4] Czeschka F. D., Dreher L., Brandt M. S., Weiler M., Althammer M. et al., Phys. Rev. Lett. 107, 046601 (2011).[5] Jungfleisch M. B., Lauer V., Neb R., Chumak A. V., and Hillebrands B., Appl. Phys. Lett. 103, 022411 (2013).[6] Qiu Z., Ando K., Uchida K., Kajiwara Y., Takahashi R. et al., Appl. Phys. Lett. 103, 09404 (2013).[7] Aqeel A., Vera-Marun I. J., van Wees B. M., and Palstra T. T. M., J. Appl. Phys. 116, 15703 (2014)