Skip to main content

Controlling domain wall dynamics for functional devices

Final Report Summary - WALL (Controlling domain wall dynamics for functional devices)

Final report: publishable summary

The WALL consortium and the Early-Stage Researchers
The WALL consortium (www.itn-wall.eu) was constituted to train the next generation of researchers in the area of spintronics, focusing on domain walls as fundamental constituents of nanoscale magnetic structures. Domain walls (DWs) are the interfaces separating magnetic domains, and their high speed manipulation promised to be used in the next generation of low power functional devices for computation and communication.The consortium consisted of world-leading experts on condensed matter physics and leading private companies, along with a range of associated partners spanning basic research, instrumentation development, electronic products, and technology policy. Eleven Early-Stage Researchers (ESR) and one Experienced Researcher (ER) coming from France, Italy, Iran, India, Bangladesh and Canada, were hired within the first year of the project. Within the WALL bended learning1 plan, they attended 17 courses on “Training through Research“, and 13 courses on “Training for Life”, both on-line (14) and on-site (16). They also attended local courses (>70 courses) as complementary training, and had extended periods of secondments within the partners, and the associated ones. They showed an high capability to work together, and collaborate on different aspects of the project. This also enhanced the collaborations within the WALL partners. In particular, they successfully organized the 2 nd Marie Curie School on Domain Wall Dynamics and Spintronics in Spetses Island, Greece, from 12 to 16 Sept. 2016. It was open to external participants free of charge and offered 15 invited talks all coming outside the WALL project. It hosted in total 37 students with 26 outside the WALL project coming from 12 countries. They also successfully attended the Collaborative Device Concept training programme where they formed 3 different groups with the commitment to investigate possible devices based on DW for recording, logic circuits, and sensing. In a final meeting they presented their Claim I for a patent about their devices. The ESRs published 30 papers (15 as first authors) and other 11 have been submitted or close to submission to International Journals. They presented their activity in 41 oral talks and 46 posters at Conferences, Schools and Workshops. Two of them also won a special award for their presentations. All of them had many outreach experiences, having courses and meeting with high school students and general public. Particular moments were the 10th Researchers' night in Torino on Sept. 2015, and ‘K’fet des sciences' in Orsay on June 2016 where they presented their activities to about 90 high school students. We believe this experience prepared them to successfully work on both academia and in European industries.

Scientific results
In the original WALL plan was to have three distinct Work Packages with the aim to 1) determinate controlled and reproducible spin structures and tailor their properties, 2) investigate and determine the influence of spin orbit torques to improve the efficiency of current-induced domain wall (DW) motion, and 3) to obtain a local and reversible control of the domain wall propagation using electrical gate. This original plan was boosted by the experimental observation of topologically non-trivial spin textures, such as homochiral DWs, which were reported to be very stable against annihilation, thus promising for technological applications. The origin of these chiral spin structures lies in the interfacial Dzyaloshinskii-Moriya interaction (DMI), an antisymmetric exchange interaction originating from the interface between a heavy metal and a ferromagnet. It was therefore necessary to adapt the first two WPs and focus more on the characterisation and optimisation of the DMI, together with other magnetic properties such as the Perpendicular Magnetic Anisotropy (PMA) and the Spin Hall angle, using new architectures and material stacks. Theoretical studies focused on different aspects of DW motion in these systems, in particular its mobility and damping, the effect of Joule heating, also to get a general description of the dynamics using a restricted number of coordinates or introducing a realistic description of the disorder. Devices based on nonlocal spin valve geometry and DW based sensors were studied in detail to explore new architectures able to improve their performances. Thin films with various underlayers (W, Ta, Pd, Pt and Ir) and different ferromagnetic (FM) alloys were first investigated. PMA was found to be tuned by 1) changing the bottom seed layer, the FM thickness and its composition, 2) varying the annealing temperature, and 3) the number of repetition layers. To optimize the DMI, the interface was modified by changing the deposition conditions of the bottom seed layer. It was found that a reduction in the DW pinning occurs for seed layers deposited at higher growth powers. In general we found that the choice of the heavy metal layer sets the strength of the DMI. New architectures where also explored: i) ferromagnet/antiferromagnet (FM/AFM) systems ii) a FM sandwiched within two layers having large and opposite DMI/Spin Hall angle, iii) fully epitaxial layers of Pt/Co/Pt1-xAux at different Au concentrations. Multilayers of Pt/Co/IrMn and Pt/Co/FeMn exhibit perpendicular exchange bias, which enhances the electrical switching through the large spin Hall torque generated by both the heavy metal and the AFM layers. In the second case, a Co layer was surrounded by Pt and Ir layers having opposite DMI values. In the last case, the addition of gold could change the effect of increasing broken inversion symmetry in a very controlled way. Creep experiments were performed to a) determine the value of the DMI by the bubble expansion method where the DW velocity is measured under in-plane fields, b) to study the DW dynamics in film and wires, also after irradiation with atoms. In wires, the shape of propagating DWs is strongly dependent on the size of the channel containing them, having a significant impact on DW velocity at low magnetic fields but also on the maximum velocity the DWs can attain at high magnetic fields. The dynamics of DW were also investigated in the case of spin-transfer torque without Joule heating, using a non-local spin valve geometry based on graphene between two magnetic electrodes of Permalloy. The non local spin valve, with half rings as injector and detector electrodes, offer the best geometry to precisely position DWs at the electrodes and study the depinning of a pure spin current onto the DW. We also high improved the technology the DW based sensors, going beyond the used open-loop concept, by introducing a different geometry, namely a cross of nanowires to allow the intertwining of loops. We explored the DW propagation in this cross-shaped element and developed a syphon structure comprised of a tilted wire to overcome the problems leading to failure events. The third task to control of the DW propagation with electronic gates was first obtained in Ta/CoFeB/MgO/HfO2 films, which showed good stability and large charge accumulation effects. This enables us to prepare the final design of Ta/CoFeB/MgO/HfO2/ionic liquid/ITO films, in which we get for the first time a reversible ionic liquid gating of CoFeB/MgO materials with a anisotropy change from perpendicular to in-plane under an applied voltage. Theoretical studies have been conducted to enlighten the interpretation of the experimental results and to simulate the best conditions for DW dynamics: in particular, efforts are focus on a description of the DW dynamics using a reduced number of variables (the so called 1D model), on the effect of Joule Heating, and in general on the dynamics in the presence of strong spin-orbit coupling.