Biomimetic\nFingerprint-like Unclonable Optical Anticounterfeiting\nSystem with Selectively In Situ-Synthesized Perovskite Quantum Dots\nEmbedded in Spontaneous-Phase-Separated Polymers

Anticounterfeiting\ntechnologies meet challenges in the Internet\nof Things era due to the rapidly growing volume of objects, their\nfrequent connection with humans, and the accelerated advance of counterfeiting/cracking\ntechniques. Here, we, inspired by biological fingerprints, present\na simple anticounterfeiting system based on perovskite quantum dot\n(PQD) fingerprint physical unclonable function (FPUF) by cooperatively\nutilizing the spontaneous-phase separation of polymers and selective\nin situ synthesis PQDs as an entropy source. The FPUFs offer red,\ngreen, and blue full-color fingerprint identifiers and random three-dimensional\n(3D) morphology, which extends binary to multivalued encoding by tuning\nthe perovskite and polymer components, enabling a high encoding capacity\n(about 10<sup>8570000</sup>, far surpassing that of biometric fingerprints).\nThe strategy is compatible with mainstream production techniques that\nare widely used in traditional low-cost printed anticounterfeiting\nlabels including spray printing, stamping, writing, and laser printing,\navoiding complicated fabrication. Macrographical patterns and micro/nanofingerprint\npatterns with multiscale-tailorable inter-ridge sizes can be fused\ninto a single FPUF label, satisfying different levels of anticounterfeiting\nrequirements. Furthermore, a smart fused scheme of enhanced deep learning\nand fingerprint characteristic comparison is leveraged, by which high-efficiency,\nhigh-accuracy authentication of our FPUFs is achieved even for the\nincreasingly huge FPUF databases and imperfectly captured images from\nusers.