Magneto-Ionic Hardware Security Primitives: Embedding Data Protection at the Material Level

Abstract

The Big Data revolution has heightened the demand for robust, energy-efficient security hardware capable of withstanding increasingly sophisticated cyber threats. Conventional encryption schemes, reliant on complex algorithms, are resource-intensive and remain vulnerable. To fortify sensitive information, society needs innovative anti-hacking and anti-counterfeiting technologies that exploit new materials and designs. Here, we present a magneto-ionic strategy for hardware-level security based on fully selective voltage-controlled N3- ion migration within pre-defined, initially paramagnetic FeCoN dots. This process generates ferromagnetic sublayers of tuneable thickness, resulting in either deterministic (single-domain or vortex) or probabilistic states (with coexisting magnetic configurations and voltage-adjustable probabilities), each exhibiting stochastic orientation and chirality, thereby providing a rich platform for magnetic fingerprinting. This approach enables self-protected primitives, including true random number generators, physical unclonable functions, and in-memory probabilistic inference. The resulting reconfigurable architecture combines tamper resistance, low energy consumption, and scalability, marking a significant leap toward next-generation hardware security rooted in emergent magnetic phenomena.