11:00 AM - 11:15 AM
▼ [14a-501-8] Time-resolved measurement of spin-transfer-torque-induced magnetization switching in CoFeB-MgO magnetic tunnel junctions with perpendicular easy axis
Keywords:spintronics, MTJ, high speed switching
It was shown that single-shot time-resolved measurements is useful to understand the magnetization switching mode induced by spin-transfer-torque (STT) in nanoscale magnetic tunnel junctions with perpendicular easy axis (p-MTJ). In this study, we investigate the STT-induced magnetization switching in CoFeB-MgO based p-MTJs from the time-resolved measurements.
A stack, from substrate side, Ta(5)/Pt(5)/[Co(0.4)/Pt(0.4)]5/Co(0.4)/Ru(0.52)/[Co(0.4)/Pt(0.4)]2/ Co(0.4)/Ta(0.3)/CoFeB(1)/MgO/CoFeB(1.6)/Ta(5)/Ru(5) is deposited on a sapphire substrate by dc/rf magnetron sputtering. Numbers in parentheses are nominal thickness in nm. The stack is processed into circular MTJs with a diameter D ranging from 40 to 125 nm on a coplanar waveguide. We apply pulse voltage V to the MTJ from a pulse-generator, and measure the transmitted voltage by an oscilloscope to detect magnetization dynamics during the switching. The current I flowing through the MTJ is also measured by the oscilloscope.
We evaluate two characteristic times; incubation time tA and dynamical switching time tB after tA. From 1000-time events at various I, we evaluate median values, tA and tB, of tA and tB. For P-to-AP and AP-to-P switching, both tA and tB decrease with increase of I. The dependence of tA and tB on I is compared with that obtained from macrospin simulation. The simulation reproduces the overall trend in the switching behavior, however, tB (tA) in P-to-AP (AP-to-P) switching shows a larger reduction with increasing I for the experiment than the simulation.
A stack, from substrate side, Ta(5)/Pt(5)/[Co(0.4)/Pt(0.4)]5/Co(0.4)/Ru(0.52)/[Co(0.4)/Pt(0.4)]2/ Co(0.4)/Ta(0.3)/CoFeB(1)/MgO/CoFeB(1.6)/Ta(5)/Ru(5) is deposited on a sapphire substrate by dc/rf magnetron sputtering. Numbers in parentheses are nominal thickness in nm. The stack is processed into circular MTJs with a diameter D ranging from 40 to 125 nm on a coplanar waveguide. We apply pulse voltage V to the MTJ from a pulse-generator, and measure the transmitted voltage by an oscilloscope to detect magnetization dynamics during the switching. The current I flowing through the MTJ is also measured by the oscilloscope.
We evaluate two characteristic times; incubation time tA and dynamical switching time tB after tA. From 1000-time events at various I, we evaluate median values, tA and tB, of tA and tB. For P-to-AP and AP-to-P switching, both tA and tB decrease with increase of I. The dependence of tA and tB on I is compared with that obtained from macrospin simulation. The simulation reproduces the overall trend in the switching behavior, however, tB (tA) in P-to-AP (AP-to-P) switching shows a larger reduction with increasing I for the experiment than the simulation.