Interfacial - Dzyaloshinskii - Moriya interaction in quasi-perpendicularly magnetized Ta/Pt/CoFeB/Pt ultra-thin films

Abstract

The physics of Ferromagnetic-Metal/Heavy-Metal(FM/HM) ultra-thin film heterostructures is not only rich enough to investigate the interplay of magnetic interactions but also relevant enough to harness a new paradigm of spintronics memory. While these interactions essentially manifest into four ingredients –

• Perpendicular Magnetic Anisotropy (PMA),

• Spin Hall Effect (SHE),

• interfacial-Dzyaloshinskii-Moriya Interaction (interfacial-DMI), and

• Proximity-Induced Magnetization (PIM),

– that characterize these heterostructures, the ability to switch using Spin-Orbit Torque (SOT) and host sub-10 nm homo-chiral Néel domain walls can be harnessed respectively for the energy-efficient Magnetic Random Access Memory (MRAM) and the scaling-efficient domain-wall-based racetrack architectures. However, these technologies suffer respectively from the requirement of an external magnetic field for deterministic switch- ing and Gilbert-damping-limited Walker breakdown.

The current thesis addresses these challenges by the introduction of thickness gradient using an oblique-angle sputter deposition technique in the constituent layers and using a low damping material CoFeB for the FM layer. The thickness gradient tilts the anisotropy away from the surface normal, leading to quasi-PMA. The tilt is characterized by the method of shifted hysteresis loops in the presence of an in-plane bias field. The tilt in the anisotropy further facilitates field-free switching. The Spin Hall angle has been characterized by the method of shifted threshold current density for switching in the presence of the in-plane field. To characterize the interfacial-DMI, we study the bubble domain expansion in the presence of the in-plane field. The presence of tilt leads to an elliptical bubble domain profile that is attributed to an additional step anisotropy developed due to o= blique-angle sputtering. Additionally, a new more sensitive technique to ascertain the tilt has been developed. Finally, the effect of annealing on the interfacial-DMI has been investigated. Annealing is found to be detrimental to the interfacial-DMI, possibly due to the in= termixing of the layers.

In Chapter 1, we describe the progress in various architectures of Magnetic Random Access Memory (MRAM). While the Spin Transfer Torque (STT) based two-terminal MRAM architecture offers scaling up to 40 nm, the high current density leads to significant barrier damage. Furthermore, STT based mediation suffers from high incubation delays. Spin-Orbit Torque (SOT) based MRAM architecture addresses these challenges using a three-terminal memory cell where the high current density required for writing is sourced via the FM/HM layers and the low current density is sourced through the Magnetic Tunnel Junction (MTJ) at the third terminal. However, scaling in both these architectures is limited by their 2D architecture. Recently proposed SOT-based racetrack magnetic memory architecture aims to drive domain walls across 3D nanowires. It is, therefore, imperative to study the effect of underlying spin-orbit interaction on the domain wall.

In Chapter 2, we describe the modes of domain walls based on the wall energy model. The interplay between exchange, spin-orbit, and Zeeman interactions determines the statics and dynamics of the domain wall. In the context of ultra-thin films, the spin-orbit interaction manifests as perpendicular magnetic anisotropy. Further, in the context of HM/FM heterostructures, the spin-orbit interaction manifests as Dzyaloshinskii-Moriya interaction that plays a key role in stabilizing Néel modes of specific chirality.

In Chapter 3, we describe the characterization tools used for the study of perpendicular magnetized ultra-thin films. The Magneto-optic Kerr effect and Vibrating Sample Magnetometer characterize the static magnetic properties. X-ray reflectometry measures the structural properties, whereas the domain wall dynamics is studied by wide-field Kerr microscopy. These interactions are described using a phenomenological model under continuum limit.

In Chapter 4, we engineer a tilt in the perpendicular anisotropy of Ta (3 nm)/Pt (3 nm)/CoFeBWR (0.5 nm)/Pt (1 nm) thin films using an obliquely angled sputter deposition technique. The tilt is characterized using the hysteresis shift observed in MOKE loops of the quasi-PMA films in the presence of the in-plane magnetic field. The tilt is optimized by varying the thickness of the Ta buffer layer, the bottom Pt layer, and the CoFeB thickness. We observe that the tilt is sensitive to the Pt thickness and interface roughness at the Pt/ferromagnet interface.

In Chapter 5, we determined the effect of tilt in the anisotropy on the current-induced deterministic magnetized reversal. To achieve this, we fabricate a microchannel using optical lithography and ion milling. Two parameters are extracted by the effect of the in-plane field on the current-induced magnetization reversal studied. While the tilt affects a deterministic reversal, it also enables to calculation of the spin Hall angle from the spin torque efficiency.

In Chapter 6, we estimate the DMI in the quasi-PMA stack. Here, we used field-induced domain wall motion in the creep regime. Due to the tilt, there is an elliptical profile for the bubble domain. We fit the scaling profile for the edges of the domain wall to the wall energy and simultaneously extract the tilt and effective DMI field. This newly developed analysis technique enables sensitive measurement of tilt due to the exponential nature of the creep dynamics.

In Chapter 7, we use thermal annealing as a method to control the interfacial-DMI. Due to the intermixing of the layers, there is a reduction in the DMI with annealing. Trends in static magnetometry parameters with annealing are also discussed.

In Chapter 8, we conclude our results and delineate possible future directions for extending the current work. One direction is to study the PIM and its relation to other ingredients of FM/HM systems. Also, the nature of domain wall as a topological defect was studied in the present thesis; we propose to study skyrmions that are higher-order topological defects for their promise in next-generation robust memory architectures.