Phenomenology of orbital torque, pumping and mixing conductance in metallic bilayers

Xiaobai Ning, Henri Jaffrès, Weisheng Zhao, Aurélien Manchon · 2024

Read Phenomenology of orbital torque, pumping and mixing conductance in metallic bilayers by Xiaobai Ning, Henri Jaffrès, Weisheng Zhao, Aurélien Manchon (2024) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2412-08340.txt.

Introduction

Xiaobai Ning, Henri Jaffrès, Weisheng Zhao, and Aurélien Manchon model in 2024 a newer kind of torque in magnetic electronics. Beyond the familiar spin currents, electrons also carry orbital motion. Converting between the two in layered metals produces an orbital torque that can flip a magnet, plus the reverse effect, orbital pumping.

Their phenomenological model treats a bilayer: a light metal that sources orbital current, joined to a magnet with strong spin-orbit coupling. Conversions at the interface and inside the magnet compete, and the torque emerges from that balance. The model gives experimenters a map for this young corner of spintronics.

Catalog record

CREATORXiaobai Ning, Henri Jaffrès, Weisheng Zhao, Aurélien Manchon
YEAR2024
COLLECTIONPAPERS
FORMATTXT
FILE SIZE35 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2412-08340

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Phenomenology of orbital torque, pumping and mixing conductance in metallic bilayers Xiaobai Ning, Henri Jaffrès, Weisheng Zhao, Aurélien Manchon Abstract The conversion between spin and orbital currents is at the origin of the orbital torque and its Onsager reciprocal, the orbital pumping. Here, we propose a phenomenological model to describe the orbital torque in magnetic bilayers composed of an orbital source (i.e., a light metal such as Ti, Ru, CuOx…) and a spin-orbit coupled magnet (i.e., typically Ni, (Co/Pt) n , etc.). This approach accounts for spin-to-orbit and orbit-to-spin conversion in the ferromagnet and at the interface. We show that the orbital torque arises from a compromise between orbital current injection from the orbital source to the ferromagnet and spin current backflow from the ferromagnet back to the orbital source. We also discuss the concept of orbital-mixing conductance and introduce the ”orbit-spin-” and ”spin-orbit-mixing” conductances that govern the orbital torque and orbital pumping, respectively. Preprint: APS/123-QED I INTRODUCTION Spin-orbit torque (SOT) induced by an electrical current has garnered significant attention for effectively manipulating the magnetization in spintronic heterostructures [ 1 , 2 ] , standing out as a promising candidate for next-generation memory and logic devices [ 3 , 4 ] . In a bilayer composed of a nonmagnetic metal (N) and a ferromagnet (F), it encompasses a two-step process: the generation of a spin current via spin Hall effect (SHE) [ 5 , 6 ] or spin Rashba-Edelstein effect (SREE) [ 7 , 8 ] in layer N, followed by its absorption by the adjacent layer F. Because both SHE and SREE arise from spin-orbit coupling (SOC), efficient SOT devices require the use of heavy metals (Pt, W, etc.). An alternative to SOT can be achieved by using currents of orbital angular momentum rather than spin currents. Such orbital currents can be efficiently generated in light metals (Ti, Al, Zr, oxidized Cu) upon charge-orbit interconversion mechanisms such as the orbital Hall effect (OHE) [ 9 , 10 ] and the orbital Rashba-Edelstein effect (OREE) [ 11 , 12 ] . In this case, an orbital current is generated in N in the absence of SOC, and is subsequently absorbed in F through orbit-to-spin conversion mediated by SOC [ 13 ] . This orbital torque and its parent effect, the orbital (Hall or Rashba) magnetoresistance, have been reported recently [ 14 , 15 , 16 , 17 , 18 ] . A hallmark of the orbital torque is its dependence as a function of F’s thickness. Since the orbital current is more likely to perform long-range transport than the spin current in F [ 19 ] , the strength of the torque exhibits a steady enhancement with increasing F’s thickness, while the conventional SOT quickly reaches saturation [ 20 , 21 , 22 , 23 , 17 , 18 ] . A practical way to model SOT in an N/F bilayer is to combine a drift-diffusion model that describes the generation of the spin current in N, and the spin-mixing conductance boundary condition that describes the spin current absorption in F [ 24 , 25 ] . This approach is valid only when the SOC of both F and the N/F interface is neglected. For orbital torque, the charge-to-orbital conversion has been computed from first principles [ 26 , 27 , 28 , 29 , 30 ] , and the resulting torque in N/F bilayers have been obtained using, e.g., tight-binding modeling [ 13 ] . Such an approach provides an instructive microscopic description but lacks transparency. Therefore, a phenomenological approach similar to the one used to model SOT is highly desired. An attempt has been made to extend the spin drift-diffusion theory to orbital transport [ 31 , 17 ] by introducing a coupling length representing the conversion between spin and orbital moments in N, while the absorption of the spin and orbital currents is modeled by an effective mixing conductance at the N/F interface. This approach provides reasonable scaling but assumes that the orbital current is fully absorbed at the interface, which is incorrect as pointed out by Ref. [ 19 ] : since the orbital moment does not directly couple to the magnetization, it does not experience dephasing, in contrast with the spin moment. Recent microscopic theories [ 32 , 33 ] have pointed out additional features that enrich the physics of orbital transport. First, since the orbital momentum operator does not commute with the Hamiltonian in general, it is not a conserved quantity (even in the absence of SOC). This has two important consequences: (i) the orbital moment exhibits a much smaller diffusion coefficient than the spin or the charge [ 33 ] . In addition, an orbital moment penetrating a normal metal experiences a precession around the local crystal field [ 32 ] , similar to the precession of a spin current penetrating a ferromagnet. Second, in the presence of SOC, spin and orbital currents are coupled through a so-called spin-to-orbit and orbit-to-spin polarization [ 33 ] . This coefficient is different from the spin-orbit length introduced in Ref. [ 31 ] , as further discussed in the present Article. Although additional microscopic investigations are necessary to clarify the nature of these effects at interfaces and the impact of disorder and defects, these features are remarkable and must be accounted for to properly model orbital transport. In the present work, we develop a drift-diffusion model that describes the spin and orbital transport induced by SHE and OHE in N/F bilayers. This approach accounts for spin and orbital diffusion inside F, as well as spin-to-orbit conversion in F and at the N/F interface. It enables us to clarify the parameters that control the orbital torque, demonstrating the interplay between orbital current injection and spin current backflow. Finally, we discuss the concept of orbital-mixing conductance and its connection to orbital pumping. II THEORY We intend to describe the phenomenology of orbital torque in an N/F bilayer, as depicted in Fig. 1 . The theory of spin transfer torque has been studied in great detail over the past two decades, pointing out the central role of spin precession and dephasing in F [ 34 , 35 , 36 , 37 ] . Since the spin dephasing length is usually short in most transition metal ferromagnets (typically, \approx 1 nm), the physics governing the absorption of an incoming spin current in a ferromagnet is often modeled by the so-called interfacial spin-mixing conductance [ 38 , 39 , 40 ] , i.e., a boundary condition that connects the interfacial spin current with the drop in spin chemical potential across the interface. It is tempting to develop a similar concept for orbital torque, the ”orbital-mixing conductance”, that connects the impinging orbital current with the drop in orbital chemical potential. Nonetheless, since orbital currents diffuse over much longer distances than spin currents in ferromagnets [ 20 , 21 , 22 , 23 , 17 , 18 ] , a comprehensive model needs to properly account for the actual spin precession and dephasing inside F. II.1 Two-channel Spin-Orbit Transport We assume that N accommodates both OHE and SHE. The spin diffusion equations in the presence of both SOC and magnetic exchange have been

Read the complete text