Effect of SIL magnetic design on STO optimization in MAMR

Abstract

Uniform and stable oscillation of magnetization is one of the most important factors for the spin-torque oscillator (STO) in microwave assisted magnetic recording (MAMR) [1], [2]. Impact of magneto-static interaction between STO and recording head have been discussed often [3], [4], but the interaction including spin torque effect is not understood well even just between the field generation layer (FGL) and the spin injection layer (SIL) in STO. In this paper, we focused upon an effect of SIL magnetic design, i.e., the saturation magnetization and layer thickness, on STO optimization for uniform and stable oscillation and tried to understand the magnetic interaction with spin torque. A commercial micromagnetics software (Fujitsu Examag v2.1.1) was used, in which the effective field due to spin transfer torque is considered by the equation shown in Fig.1. In this analysis, a uniform alternative field was applied to STO and magneto-static interaction between STO and head was neglected as shown in Fig.1. The dimensions and magnetic properties are as the followings; FGL&SIL width and height $=30 \mathrm {x}30$ nm, FGL thickness $=6$ nm, FGL Bs $=2$ or 1 T, FGL&SIL Hk $=1$ Oe, Polarization $=0.5$, Exchange $=1.0\mathrm {E} -6$ erg/cm, Damping $=0.02$, SIL thickness and SIL Bs are variables. An alternative rectangular field with 1 GHz frequency and 14 kOe magnitude was applied and the injection current density was set at $4.0\mathrm {E} +8\mathrm {A} /$cm 2 with a direction to SIL from FGL. Results are shown in Fig.2 for (a) FGL Bs $=2\mathrm {T}$ and (b) FGL Bs $=1\mathrm {T}$, respectively. Here, since the reduced crosstrack component (My/Ms) of averaged FGL magnetization alternates at a microwave frequency, the root mean square (rms) value was calculated and multiplied by $\surd 2$ to get 1 at the perfect sine-wave oscillation. In Fig.2(a) for the FGL Bs $=2\mathrm {T}$, we can see that the optimum Bs of SIL is around at 1.5, 0.8, 0.6 and 0.4 T for the SIL thickness of 6, 3, 2 and 1, respectively. It's very interesting that the optimum Bs of SIL decreases simply according as SIL thickness increases, and these SIL design impacts significantly to FGL oscillation. In the case for FGL Bs $=1\mathrm {T}$, as shown in Fig.2(b), all optimum points have moved to larger Bs positions compared to Fig.2(a). Mechanism for these characteristics is not fully understood yet, but it should relate to both the magneto-static effect and the spin transfer effect, which interact each other. From these figures, once FGL Bs is given, the optimum Bs and the thickness for SIL can be designed. We also confirmed that these optimum positions for the SIL doesn't change even if FGL thickness was changed, and it depends only upon the FGL Bs.