Session N68: Creation, Control and Manipulation of Chiral Spin Textures

Skyrmion-hosting systems: from novel multilayers to magnetic tunnel junctions

Magnetic skyrmions are fascinating topological textures with a number of intriguing fundamental and applicative properties. In the past years, a lot of efforts have been put into the stabilization and observation of skyrmions at room temperature, and today they can be achieved in a variety of systems spanning from magnetic multilayers¹ to ferrimagnets². However, the exploitation of skyrmion features for real device implementation still requires the search of new and more efficient materials, systems, and manipulation mathods.

Here, we show the existence of skyrmions in different systems. First, we prove the stabilization of two distinct skyrmion phases in a novel hybrid ferromagnetic/ferrimagnetic multilayer^2,3. We identify one skyrmion which goes through the whole thickness of the multilayer – tubular skyrmion – and another one which exists only in the external ferromagnetic multilayers but not in the internal ferrimagnetic material – incomplete skyrmion. The two skyrmions exhibit very distinct MFM contrasts, and this can be used to code the different information bit in a racetrack memory.

Second, we focus on the skyrmions dynamics in ferromagnetic multilayers driven by thermal gradients^4,5. Our experimental measurements show that skyrmions can be thermally manipulated and, particularly, can be moved from the hot region to the cold region in a deterministic way. Many interactions (entropic torque, magnonic torque, thermal spin-transfer torque) can be responsible for such skyrmion motion. To disentangle the different contributions, we perform micromagnetic simulations which suggest that the entropic torque could be the major factor in promoting the hot-to-cold motion.

Eventually, we deal with the achievement of skyrmions in a magnetic tunnel junction (MTJ)⁶. We design an innovative system where a standard ferromagnetic multilayer is coupled with a standard CoFeB-based MTJ. In particular, the multilayer acts as “skyrmion generator” for the MTJ free layer. We can measure a TMR of about 20%, and each skyrmion provide a resistance change of about 50 ?. The existence of skyrmions in MTJ paves the way not only to an efficient detection method, but also to applications, such as skyrmion oscillators, detectors, memristors, which are still only theorized.   

Spin textures in synthetic antiferromagnets

Synthetic antiferromagnets make use of indirect exchange coupling across spacer layers to achieve an antiferromagnetic ground state in a magnetic multilayer. Here we report on studies of synthetic antiferromagnets based on the repeat unit [Pt(8)/ CoB(16)/ Ru(7)/ Pt(8)/ CoFeB(10)/ Ru(7)], with layer thicknesses in Å. The CoB and CoFeB layer thicknesses have been chosen so that their moments are equal and cancel owing to the antiferromagnetic coupling though the Ru spacers. They thus form the sublattices of an antiferromagnetic structure. The Pt layers ensure that the magnetic layers are perpendicularly magnetised and possess an interfacial Dzyaloshinskii-Moriya interaction.

We have studied the magnetisation process of a multilayer with 5 repeat units. Conventional magnetometry shows a low magnetisation regime (SAF alignment) up to ~40 mT, with saturation occurring at ~80 mT. Phase coexistence of antiferromagnetic and ferromagnetic regions is observed in the intermediate regime by MFM. The lack of spin textures in the AF regions was confirmed by PEEM imaging of the top layer only. When they exceed a certain size ~150 nm, the FM regions break up into worm domains and skyrmions, as seen by MFM in real space and small angle soft x-ray scattering in reciprocal space.

Nanoscale devices formed from a similar multilayer were studied by scanning x-ray transmission microscopy. Correlated SAF domain wall motion at current densities as low as 0.3 TA/m² was observed in both sublattices by tuning to the Co and Fe L₃ edges. Motion at velocities up to 40 m/s was observed at current densities ~0.6 TA/m², markedly less than required to drive domain walls in a conventional chiral multilayer where Ir replaces the Ru.

The currents were injected from a finger-shaped electrode with radius of curvature at the tip of 250 nm. Under appropriate conditions, skyrmion nucleation in the SAF state can occur.   

Skyrmionics: Second Transformation of Spin Era

A magnetic skyrmion is a particle-like topological spin texture, where the forming nots are key parameters to hold its topology. Its existence has been proposed in a system with low symmetry since late 80’s and especially MnSi in 2006[1], and the experimental proof of this structure in crystalline form has come out in 2009[2]. After that, isolated magnetic skyrmion has been the main target since each magnetic skyrmion can carry the information in all kinds of devices. Especially, it has drawn a lot of attention for its potential use in spintronic devices. From memory to logic devices, its manipulations(creation, deletion, movement) with pure electric external variables are most demanding. Contrary to its expectation for the appearance of a revolutionary device, even the generation of this magnetic skyrmion at the place aimed has not been possible. Recently we have shown several unique methods for the generation and manipulation of this magnetic skyrmion based on the detailed skyrmion formation process[3-5]. With these methods, we can demonstrate several skyrmionic devices[6-8] such as skyrmion racetrack memory, skyrmion transistor, and skyrmion neuromorphic device. Our next step is leading the quantum effects with magnetic skyrmions. What we need at this stage is to decrease its size. With the consideration of shrinking its size down to the scales of lattice spacing, we can discuss several examples of quantum skyrmions. 

[1] A. N. Bogdanov, D. A. Yablonskii, Zh Eksp Teor Fiz. 95, 178 (1989), U. K. Rößler et al., Nature 442, 797 (2006)

[2] S. Mu¨hlbauer et al. Science, 323, 915(2009). 

[3] K.-W. Moon et al., NPG Asia Materials 13, 20 (2021)

[4] S. Yang et al., JMMM 539, 168381(2021)

[5] S. Yang, Nano Letters accepted (2022)

[6] S. Yang et al., Adv. Mater. 33, 2104406(2021)

[7] M. Song et al., Adv. Mater. 34, 2270277(2022)

[8] under review

    

Pulling Magnetic Skyrmions out of Thin Air

Magnetic skyrmions are topologically nontrivial spin textures with envisioned applications in energy-efficient magnetic information storage. Toggling the presence of skyrmions via writing/deleting processes is essential for spintronics applications, which usually requires a magnetic field, current injection, electric field, laser pulse, or thermal excitation. A key challenge is to directly image such ultrasmall spin textures in order to confirm their topological character. Using spin-polarized low-energy electron microscopy (SPLEEM), we have demonstrated a new, field-free method to write/delete skyrmions at room temperature, via hydrogen chemisorption/desorption cycles on Ni/Co/Pd/W(110) [1]. The written skyrmions are resolved to be Néel-type using SPLEEM. Supported by Monte-Carlo simulations, the skyrmion creation/annihilation is attributed to the hydrogen-induced anisotropy change on ferromagnetic surfaces. The roles of hydrogen and oxygen on anisotropy and skyrmion deletion on other magnetic surfaces are also demonstrated.

In another study, we focus on the interfacial Dzyaloshinskii-Moriya interaction (DMI), which stabilizes chirality with preferred handedness. Experimentally, controlling the handedness is crucial to tune the efficiency of current-induced manipulation of spin texture. This is conventionally achieved by stacking asymmetric multilayers where the thickness of each layer is at least a few monolayers. We observed a chirality switching in (Ni/Co)_n multilayer induced by capping only 0.22 monolayer of Pd [2]. Using SPLEEM, we monitor the evolution of domain walls from left-handed to right-handed Néel walls and quantify the DMI induced by the Pd capping layer. We also observe the chiral evolution of a skyrmion during the DMI switching, where no significant topological protection is found as the skyrmion winding number varies. This corresponds to a minimum energy cost of < 1 attojoule during the skyrmion chirality switching.

These results open up new opportunities for designing energy-efficient skyrmionic and magneto-ionic devices. They also illustrate the effectiveness of SPLEEM in resolving magnetization vector in chiral spin textures with high spatial resolution.   

Magnetic Skyrmions for Quantum Computing

In this talk, I will introduce a new class of quantum logic elements based on nanoscale topological solitons called skyrmions [1]. In a skyrmion qubit, information is stored in the quantum degree of helicity, and the logical states can be adjusted by electric and magnetic fields, offering a rich operation regime. I will discuss the appropriate fields required to generate single-qubit gates for quantum computing, and skyrmion multiqubit schemes for a scalable architecture with tailored couplings. Scalability, controllability by microwave fields, operation time scales, and readout by nonvolatile techniques converge to make the skyrmion qubit highly attractive as a logical element of a quantum processor.

The applicability to quantum operations, however, depends on the viability of macroscopic quantum tunneling, a precondition for a quantum computer. I will discuss how by tuning the external fields, magnetic skyrmions in frustrated magnets offer a platform for the observation of diverse quantum tunneling effects [2], We predict quantum tunneling processes for a typical skyrmion spin texture of a 5 nm radius to occur with an inverse escape rate within seconds below 100 mK. In the absence of an energy bias, quantum tunneling effects give rise to an energy tunnel splitting in the MHz regime. Our studies suggest that quantum mechanics manifests itself at the macroscopic level of skyrmion helicity. We offer the parameter space for a practical recipe in device architectures to explore magnetic helicity in the solid-state for qubit operations.

[1] C. Psaroudaki and C. Panagopoulos, Phys. Rev. Lett. 127, 067201 (2021).

[2] C. Psaroudaki and C. Panagopoulos, Phys. Rev. B 106, 104422 (2022).