Giant orbital magnetoresistance in the antiferromagnet CoO driven by dynamic orbital angular momentum interaction

Recent predictions of orders of magnitude larger orbital current effects compared to spin currents have attracted significant interest. However, the full potential of giant orbital currents remains to be fully harnessed, since so far, the orbital currents need to be converted into spin currents before they can interact with the static magnetization that is dominated by spin angular momentum in conventional magnets. By using a magnet dominated by orbital angular momentum, we demonstrate a more than fifty-fold enhancement in orbital Hall magnetoresistance in CoO/Cu*, compared to conventional CoO/Pt. This is found to be driven by a unique interaction between dynamic orbital angular momentum from surface oxidized Cu* (i.e., the orbital current) and the static orbital angular momentum which constitutes the magnetic moments in the antiferromagnetic insulator CoO. A distinctive scattering mechanism for orbital currents at the CoO interface leads to a sign reversal in orbital magnetoresistance in CoO/Cu* compared to CoO/Pt. Our results show how by using orbital angular momentum-dominated materials such as CoO, we can harness the benefits of giant orbital currents that have not been possible using conventional spin-dominated magnets, for orbitronics-based devices, offering unprecedented energy efficiency for operations of antiferromagnets that combine ultimate stability with THz dynamics.

💡 Research Summary

In this work the authors demonstrate a strikingly large orbital‑Hall magnetoresistance (OMR) in antiferromagnetic CoO when paired with a surface‑oxidized copper layer (Cu*), and they compare it directly with the conventional spin‑Hall magnetoresistance (SMR) observed in CoO/Pt bilayers. The key idea is to replace the usual spin‑dominant magnetic layer with a material whose static magnetization is carried predominantly by orbital angular momentum (OAM). CoO is an insulating collinear antiferromagnet in which each Co²⁺ ion retains a sizable orbital moment (~2.05 μB), unlike typical 3d ferromagnets where crystal‑field quenching reduces the orbital contribution to <0.1 μB.

Cu* (naturally oxidized Cu) is known to generate strong orbital currents via the bulk orbital Hall effect (OHE) and/or the interfacial orbital Rashba‑Edelstein effect (OREE). When a charge current is driven through the Hall‑bar device, an orbital accumulation μ_O builds up at the CoO/Cu* interface. Because the CoO magnetic order is defined by the Néel vector, the absorption or reflection of the orbital current depends on the Néel orientation, giving rise to a transverse resistance change analogous to SMR but rooted in orbital physics.

Experimentally, 5 nm CoO/2 nm Pt and 5 nm CoO/6 nm Cu* heterostructures were grown epitaxially on MgO(001). In the Pt‑based sample, a conventional SMR of 0.0078 % is measured at 150 K, consistent with a spin‑Hall angle of ≈3.5 % for the 2 nm Pt layer and a spin‑mixing conductance G_r ≈5 × 10¹⁴ Ω⁻¹ m⁻². In stark contrast, the Cu*‑based sample exhibits an OMR of 0.28 %—about 36 times larger—and, importantly, the sign of the resistance change is opposite to that of the Pt sample. This sign reversal signals a fundamentally different scattering mechanism: the orbital current couples to the unquenched orbital moment of CoO rather than to a spin moment.

Temperature‑dependent measurements reveal that the OMR peaks at 150 K (0.28 %) and remains sizable up to 275 K, while the SMR stays an order of magnitude smaller and retains the opposite sign throughout. The OMR/SMR ratio grows from ~35 at 200 K to ~59 at 100 K, suggesting that phonon‑mediated angular‑momentum relaxation, which strongly damps spin currents, has a much weaker effect on orbital currents.

To pinpoint the origin of the orbital current, the authors performed a systematic Cu* thickness study. Reducing the Cu* layer from 5.5 nm to 3.7 nm leaves the OMR essentially unchanged (~0.32 %), whereas further thinning to 3.2 nm causes a modest drop, indicating that the bulk OHE dominates over any interfacial OREE contribution. As a control, an α‑Fe₂O₃/Cu* bilayer—where the static orbital moment is quenched—shows virtually no OMR, confirming that the giant effect requires a magnetic layer with robust orbital angular momentum.

The authors interpret the results in terms of an orbital‑orbital exchange interaction that can be comparable in magnitude to the conventional spin‑spin (s‑d) exchange. Recent theoretical work predicts such strong orbital‑orbital coupling, and the present data provide the first experimental validation. The observed sign reversal is attributed to the presence of an orbital quadrupole moment in CoO, whose dynamics differ from the usual dipolar spin dynamics of ferromagnets.

Overall, the study establishes three important points: (1) orbital currents generated by OHE can be orders of magnitude larger than spin currents generated by SHE; (2) when the adjacent magnetic material possesses a sizable, unquenched orbital moment, the orbital‑orbital exchange yields a magnetoresistive response far exceeding conventional SMR; (3) the orbital‑based effect is less susceptible to temperature‑dependent phonon scattering, offering superior performance at low temperatures.

These findings open a new pathway for “orbitronics” devices that exploit orbital angular momentum directly, bypassing the inefficient spin‑to‑orbit conversion step required in spin‑orbit torque technologies. By leveraging antiferromagnetic insulators like CoO, which combine THz‑scale dynamics, negligible stray fields, and strong orbital magnetism, future memory and logic architectures could achieve unprecedented energy efficiency and speed. The work also suggests further exploration of other orbital‑rich antiferromagnets (e.g., MnO, FeO) and detailed spectroscopic studies to disentangle OHE versus OREE contributions, paving the way toward practical orbital‑current‑driven spintronic applications.