Fig. 4: Normalized torques, their temperature dependence and determination of SHE and spin swapping contribution.
From: Generation of out-of-plane polarized spin current by spin swapping
a–c \({\tau }_{y,{AD}}^{{\prime} }\) (\({|\tau }_{{{{{{\rm{y}}}}}},{{{{{\rm{AD}}}}}}}\)/\({\tau }_{{{{{{\rm{y}}}}}},{{{{{\rm{FL}}}}}}}|\)), \({\tau }_{z,{AD}}^{{\prime} }\) (\({|\tau }_{{{{{{\rm{z}}}}}},{{{{{\rm{AD}}}}}}}\)/\({\tau }_{{{{{{\rm{y}}}}}},{{{{{\rm{FL}}}}}}}|\)) and \({\tau }_{z,{FL}}^{{\prime} }\) (\(|{\tau }_{{{{{{\rm{z}}}}}},{{{{{\rm{FL}}}}}}}\)/\({\tau }_{{{{{{\rm{y}}}}}},{{{{{\rm{FL}}}}}}}|\)) for Cu(5 nm)/Py(5 nm), Ru(5 nm)/Py(5 nm), Mn3Sn (dAFM = 3-12 nm)/Py(5 nm), Re(5 nm)/Py(5 nm) and Pt(5 nm)/Py(5 nm) films, which are grown on Al2O3 substrate. d Temperature dependence of \({\tau }_{y,{AD}}^{{\prime} }\), \({\tau }_{z,{AD}}^{{\prime} }\) and \({\tau }_{z,{FL}}^{{\prime} }\) for the Mn3Sn(12 nm)/Py(5 nm) film. e, f The absolute value of \({\tau }_{{{{{{\rm{y}}}}}},{{{{{\rm{AD}}}}}}}\)/\({\tau }_{{{{{{\rm{z}}}}}},{{{{{\rm{AD}}}}}}}\) and \({\tau }_{{{{{{\rm{y}}}}}},{{{{{\rm{AD}}}}}}}\)/\({\tau }_{{{{{{\rm{z}}}}}},{{{{{\rm{FL}}}}}}}\) for Mn3Sn/Py and non-magnetic (Cu, Ru, Re and Pt)/Py bilayers. SHE and spin-swapping contribute to all the bilayers and SHE dominates in Pt and Re whereas spin-swapping influences in Cu and Ru. As a function of disorder, spin swapping is large for Cu/Py and Pt/Py bilayers grown on Al2O3 substrates whereas SHE is relatively significant when the same structures are grown on Si/SiO2 substrates. Note that the value of \({\tau }_{{{{{{\rm{y}}}}}},{{{{{\rm{AD}}}}}}}\)/\({\tau }_{{{{{{\rm{z}}}}}},{{{{{\rm{AD}}}}}}}\) and \({\tau }_{{{{{{\rm{y}}}}}},{{{{{\rm{AD}}}}}}}\)/\({\tau }_{{{{{{\rm{z}}}}}},{{{{{\rm{FL}}}}}}}\) for Al2O3/Cu/Py structure is multiplied by 10 for better visibility compared to other data points. All the data presented in this figure are measured on 0° device.
