We critically examine the exchange bias (EB) effect reported in ${\mathrm{Y}}_3\mathrm{F}{\mathrm{e}}_5{\mathrm{O}}_{12}\left(\mathrm{YIG}\right)/\mathrm{G}{\mathrm{d}}_3\mathrm{G}{\mathrm{a}}_5{\mathrm{O}}_{12}/\mathrm{G}{\mathrm{d}}_3\mathrm{G}{\mathrm{a}}_5{\mathrm{O}}_{12}$ heterostructures [Kumar et al., Phys. Rev. B 103, 064421 (2021)]. We first replicated the original experimental conditions and successfully reproduced the reported results. We then examined the dependence of the EB field on YIG thickness but observed no systematic variation, which challenges the interpretation of EB as an interfacial effect. A Cross comparison using electromagnet- and superconducting magnet-based setups further showed that the EB-like signal appears only in the latter. These findings suggest that, while the EB effect reported in the original study could exist under certain conditions, it is highly susceptible to artifacts arising from trapped magnetic fields in superconducting magnet systems. Our results highlight the necessity of rigorous cross validation when evaluating EB in low-coercivity systems such as YIG.

Close

PrevNext

View figure in article

• Exchange bias
• Spintronics
• Interfaces
• Magnetic insulators
• Thin films
• Laser ablation
• Magnetization measurements
• X-ray diffraction