Stochastic Orientational Encoding via Hydrogen Bonding Driven Assembly of Woven-Like Molecular Physically Unclonable Functions.

Abstract

The prevention of counterfeiting and the assurance of object authenticity require stochastic encoding schemes based on physically unclonable functions (PUFs). There is an urgent need for exceptionally large encoding capacities and multi-level responses within a molecularly defined, single-material system. Herein, a novel stochastic orientational encoding approach is demonstrated using a facile ambient-atmosphere solution processing of a molecular thin film based on the rod-shaped oligo(p-phenyleneethynylene) (OPE) π-architecture. The nanoscopic film, derived from the small molecule 2EHO-CF₃PyPE with donor, acceptor, and π-spacer building units, is designed for energetically favorable uniaxial molecular assembly and crystal growth via directional multiple hydrogen-bonding motifs at the molecular termini and short C─H···π contacts at the center. A facile solvent vapor annealing induces concurrent dewetting and microscopic 1D random crystallization, yielding a woven-textured random features. Using convolutional neural networks, the rich variations in microcrystal domain properties and stochastic encoding of 1D crystal orientations generate artificial coloration, achieving an encoding capacity reaching (6.5 × 10⁴)^{(2752 × 2208)}. The results demonstrate an effective strategy for achieving ultrahigh encoding capacities in a thin film composed of a single-material. This approach enables low-cost, solution-processed fabrication for mass production and broad adoption, while opening new opportunities to explore molecular-PUFs through structural design and engineering noncovalent interactions.