Paper-Fiber-Activated Triplet Excitons of Carbon Nanodots for Time-Resolved Anti-counterfeiting Signature with Artificial Intelligence Authentication.

Abstract

The easy-to-imitate character of a personal signature may cause significant economy loss due to the lack of speed and strength information. In this work, we report a time-resolved anti-counterfeiting signature strategy with artificial intelligence (AI) authentication based on the designed luminescent carbon nanodot (CND) ink, whose triplet excitons can be activated by the bonding between the paper fibers and the CNDs. Paper fibers can bond with the CNDs through multiple hydrogen bonds, and the activated triplet excitons release photons for about 13 s; thus, the speed and strength of the signature are recorded through recording the changes in luminescence intensity over time. The background noise from commercial paper fluorescence is completely suppressed, benefiting from the long phosphorescence lifetime of the CNDs. In addition, a reliable AI authentication method with quick response based on a convolutional neural network is developed, and 100% identification accuracy of the signature based on the CND ink is achieved, which is higher than that of the signature with commercial ink (78%). This strategy can also be expanded for painting, calligraphy identification.