Photoacoustic imaging for physical unclonable functions

Physically unclonable functions (PUFs) use inherent physical randomness for secure authentication and key generation. Conventional electronic PUFs face challenges like environmental sensitivity and rigid substrates. We propose a photoacoustic (PA) PUF (PA-PUF) using a CuO/SnO₂ nanoparticle composite, where laser-induced thermal expansion generates ultrasonic waves encoding nanoscale randomness. The PA images are digitized into unique binary fingerprints. Testing across ten spatially separated regions showed bit uniformity, inter-device Hamming distance, and entropy that approach ideal values. Machine learning attacks achieved close to 50% accuracy, indicating robustness. The PA-PUF maintains performance on flexible substrates, demonstrating practical potential for wearable and embedded security applications.