Sarah Gorski - Materials Seminar

Identifying covert methods to prevent the production and distribution of counterfeit goods in additive manufacturing (AM) is critical, making metal nanoparticle anticounterfeit tags a highly effective solution. Vat polymerization (VP) techniques use photopolymer resins that are exposed to UV light to 3D print objects layer-by-layer. This dissertation reports the integration of metal nanoparticles with VP to create high-resolution 3D printed structures with security features embedded directly within the object. Optimization of material selection and nanoparticle integration is considered, along with the effects of UV exposure on the embedded nanoparticles. Drawing inspiration from existing anticounterfeit examples, methods for embedding security features into VP processes and their authentication are developed. These concepts are extended to stretchable substrates, where plasmonic nanoparticle-based tags act as tamper-evident features. Embedding the nanoparticles enhances protection compared to surface-deposited tags, reducing susceptibility to tampering. Machine learning was used to authenticate these tags by quantifying deformation-induced changes in prominent optical features. To address the growing risk of counterfeit production in AM, this dissertation reports methods for incorporating metal nanoparticle-based taggants into 3D printed objects. Plasmonic nanoparticle inks were used to create physical unclonable functions (PUFs) integrated into stereolithographic (SLA) processes, including fabrication of a 3D medical device. Multiple nanoparticle ink formulations were evaluated for PUF generation with dark-field optical microscopy. Deep machine learning of PUF images enables lot-level authentication, with four convolutional neural networks being compared for effectiveness. Overall, integrating PUFs with AM enables anticounterfeit enables scalable anticounterfeit strategies using covert taggants and rapid authentication.