Quantifying Mn Diffusion through Transferred versus Directly Grown Graphene Barriers

We quantify the mechanisms for manganese (Mn) diffusion through graphene in Mn/graphene/Ge (001) and Mn/graphene/GaAs heterostructures samples prepared by layer transfer versus growth directly on semiconductor substrate. These are important applications spintronics; however, challenges synthesizing technologically substrates such as GaAs necessitate anneal steps, which introduce defects into graphene. \textit{In-situ} photoemission spectroscopy measurements reveal that Mn grown a Ge substrate is 1000 times lower than without ($D_{gr,direct} \sim 4\times10^{-18}$cm$^2$/s, $D_{no-gr} 5 \times 10^{-15}$ cm$^2$/s at 500$^\circ$C). Transferred suppresses factor of 10 compared to no ($D_{gr,transfer} 4\times10^{-16}cm^2/s$). For both transferred directly-grown graphene, low activation energy ($E_a 0.1-0.5$ eV) suggests occurs primarily defects. This further confirmed diffusivity prefactor, $D_0$, scales with defect density sheet. Similar barrier performance found substrates; it not currently possible grow GaAs. Our results highlight importance developing functional substrates, avoid damage induced annealing.