Measures low-field response of MTJs, revealing improved detectability in small magnetic fields, suggesting enhanced evaluation methods.
Magnetic tunnel junction (MTJ) sensors are attractive for detecting extremely small magnetic fields, yet their performance is often quantified using sensitivities derived from major loops or larger-field minor loops which inherently include hysteretic contributions. As a result, these conventional metrics significantly overestimate detectability by including irreversible processes that vanish in the small-field limit. Here, we systematically measure the low-field response of uniaxial MTJs under both ac and dc magnetic excitation and demonstrate a distinct transition from hysteretic to fully reversible behavior as the field amplitude is reduced. By adapting the Jiles–Atherton model to analyze MTJ conductance, we establish a unified framework that captures magnetization processes down to nanotesla range, yielding quantitative parameters that resolve reversible and irreversible dynamics with excellent fidelity. In the reversible regime, the sensitivity converges to a constant value of 0.55%Oe − 1 , defining an intrinsic, history-independent response of the MTJ. Importantly, this is less than half of the 1.2%Oe − 1 obtained from higher-field minor loop estimates but is physically representative of the sensing response in low-field operation. The constant and intrinsic sensitivity corresponds to the true value of detectability, confirmed by noise spectral density measurements under low-field ac excitation. These results establish a quantitative framework for MTJ evaluation that emphasizes intrinsic, hysteresis-free performance, providing both a realistic basis for calculating field detectability and a general methodology for probing magnetization in micron- and nano-scale ferromagnets.
No takes yet. Share an insight, caveat, or question.
Brown et al. (2026) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: