Low-current spin-transfer switching and its thermal durability in a low-saturation-magnetization nanomagnet

Abstract

A spin-transfer magnetization switching technique is a promising candidate as a writing mechanism for a high-density magnetic random access memory because of its scalability. The required switching current Ic, however, is still too large for this technique to be applied to MRAM using tunneling magnetoresistive devices. Here, it is demonstrated that reducing the saturation magnetization Ms of magnet cells is an effective way to decrease Ic. Use of a CoFeB film with μ0Ms of 0.75T as a magnet cell reduced Ic measured with a continuous current by an order of magnitude. We changed the duration of a writing current pulse from 1μs to 5s to investigate thermal effects on the switching process, and predicted that CoFeB magnet cells with low Ic can be compatible with the thermal durability required for MRAM applications.

Current-Excited Magnetization Dynamics in Narrow Ferromagnetic Wires

“…2(a). This behavior is reminiscent of thermally-assisted spin torque induced magnetization reversal observed in pillar structures, 20,21) supporting that the spin torque plays an essential role to drive the domain wall in the present case and that thermal agitation effectively lessens the pinning potential energy barrier for the domain walls.…”

Hybrid CMOS/Magnetic Memories (MRAMs) and Logic Circuits

“…One of the first attempts to reduce the critical current density for writing was proposed by Yagami et al who proposed to replace the 2.5 nm CoFe 25 storage layer by a 2 nm thick CoFeB layer the M s of which is only 40 % of the CoFe magnetization [60]. The study was performed on pseudo-spin valve structures of the form CoFe(10 nm)/Cu(6 nm)/storage.…”

Method for Evaluating the Thermal Stability Parameter by Distribution of the Switching Current in Spin-Torque Switching in Nano-magnets

“…The larger Δ leads to longer retention time and the narrower distribution of I c . Until now, Δ in STS has been evaluated from the dependence of I c on the width of applied pulse current (t p ) 11) from several μsec to several seconds. Hereafter we call this method as the t p dependence method.…”