The domain-wall depinning boundary, showing the variation in critical current density with magnetic field, is measured for notched permalloy wires using pulsed-current measurements. The structure of domain walls trapped at the pinning potential provided by the notch is imaged using photoemission electron microscopy. The experimental depinning boundary is compared with those obtained by micromagnetic simulations including the adiabatic and nonadiabatic spin-torque terms. This method allows for the determination of both the nonadiabaticity parameter  and spin current polarization P, which we obtain as  = 0.040Ϯ 0.005 and P = 0.40Ϯ 0.02 at room temperature.Current induced domain-wall movement 1 has been the focus of much recent research due to its promising device applications and interesting underlying physics. Theoretically, the interaction between spin-polarized currents and magnetization is described using additional spin-transfer torque terms added to the Landau-Lifshitz-Gilbert equation, 2,3
͑1͒valid for current applied along the x axis. Here m is the unit vector along the local magnetization direction, ␥ is the gyromagnetic ratio, ␣ is the damping constant, and u = JPg B / 2eM S , where J is the charge current density, P is the spin current polarization, and M S is the saturation magnetization. The last two terms on the right-hand side are the spin-transfer torque terms, namely, the adiabatic 4 and nonadiabatic spin-transfer torque. 2,3 The latter term was introduced to resolve discrepancies between experimental observations and theoretical predictions 4 and its contribution is determined by a dimensionless constant  known as the nonadiabaticity parameter, given by  = ប / J ex sf , 3 where J ex is the s-d exchange interaction energy and sf is the spin-flip time. Although the current theoretical treatment is sufficient to explain the interaction between spin-polarized currents and domain walls, the exact value of  for various materials is still under debate. Zhang and Li 2 have argued that the value of  should be of the order of 10 −2 , as confirmed by studies of current-driven domain-wall motion in permalloy nanowires. 5,6 A number of recent studies have investigated the value of , 2,3,5-13 yielding a range of values. In a recent experiment, 12 the dynamics of domain-wall depinning was studied in permalloy wires by pulsed-current measurements and comparison of experimental results with micromagnetic modeling have yielded values of  = 0.016 and P = 0.6 for ␣ = 0.008. Experimental observations of the current-induced domain-wall propagation in permalloy strips 3 were best reproduced for  = 0.04 and P = 0.4 for ␣ = 0.02. On the other hand, the domain-wall depinning from a notch induced by a pulsed current 12 was reproduced for  = 0.016 and P = 0.6 for ␣ = 0.008. The relationship between  and the damping constant ␣ has also come under debate. Kohno et al. 14 have argued that generally  ␣, while other theoretical considerations have shown that  should equal ␣. 15-17 However, recent comparison of microm...
