Structural and Electrical Variability of Functional Materials for Cryptographic Key Generation in Quantum-Resistant Hardware

Abstract

Security into quantum-resistant hardware needs to be built more than just cryptographic algorithms and must take advantage physically hard to duplicate properties. This work was aimed at investigating the potential for using the structure and electric properties of functional materials for hardware-based generation of cryptographic keys. The structural properties were grain size, film thickness, surface roughness, defect density, porosity, crystallinity, lattice parameter and phase purity, and the electrical properties were conductivity, resistance, dielectric constant, leakage current, switching voltage and impedance. These properties were converted into hardware fingerprints of the devices and tested with uniqueness, reproducibility, bit balance, Hamming distance, entropy, stable-key rate and bit-error rate. The results revealed significant structural and electrical device variations in functional materials with high cryptographic performance. The uniqueness of the Mean PUF approached 50%, the maximum inter-device Hamming distance was greater than 97%, and the Shannon entropy was almost 1 bit per bit. Uniqueness, reproducibility and entropy were positively correlated with structural variability, while key bit-error rate was negatively correlated with both structural and electrical variability. The entropy, uniqueness, reproducibility and stable-key performance reported by devices that met the hardware-readiness requirements was far more robust than that of non-ready devices. The findings suggest that the functional-material variability could be a valuable physical entropy source and hardware root of trust for cryptographic key generation. The method can be used for non-quantum cryptographic designs to enhance secure embedded hardware against cloning, key derivation, and device authentication.