Seminar on AI/ML Directed Polymer Materials and 3D Printing

Professor Rigoberto Advincula

Governor's Chair Professor and Group Leader

Oak Ridge National Laboratory and University of Tennessee

This seminar will be chaired by A/P Hortense Le Ferrand.

Background and Objectives: Creating and curating new data appends the way we approach materials science. In additive manufacturing (AM), the fabrication of parts and objects with high complexity and high performance is advantageous over other methods. Using nanocomposites enables highly improved properties even with “commodity polymers” that do not need to undergo high-temperature processes or extensive reformulation. With artificial intelligence and machine learning (AI/ML), optimizing the formulation and manufacturing methods is possible. Using sensors capable of a feedback loop mechanism and the ability to use simulation to create digital twins, optimizing properties in record time is possible.

Methods: Beyond statistical and logic-derived design, including regression analysis, are starting points for designing experiments (DOE) or principal component analysis (PCA) in optimization we have utilized ML to enable better experimental parameters towards high performance. We have demonstrated this in two examples: 1) synthesis protocol for copolymers for 3D printing, and 2) ML Methods for optimization of 2D nanomaterial dispersions in piezoelectric sensors.

Results: We thus demonstrated the approaches toward understanding nanostructuring in composites optimization via AI/ML and other training/learning sets for specific properties and applications. 3D printing enabled quadrupling of the piezoelectric behavior of PVDF. The use of 8 wt% nanomaterial loading resulted in up to 8x improvement in sensitivity as a piezoelectric sensor. Digital workflows and ML in flow chemistry reactions was made possible by introducing more sensors (monitoring instruments) in the optimization processes and real-time ML with online monitoring allows a feedback loop control.

Dr. Advincula is a Governor's Chair Professor and Group Leader at Oak Ridge National Laboratory and University of Tennessee. His research and group interests are: design, synthesis, and characterization of polymers and nanomaterials capable of controlled-assembly, nanostructuring, and self-organization in ultrathin films and monoliths. This includes functional macromolecules, coordination polymerization, polymer brushes, electropolymerization, and preparation of nanoparticles and hybrid materials. Properties include: smart coatings, stimuli-responsive, toughened, conducting, energy translating, optically active, or biocompatible. Surface sensitive spectroscopy and microscopy is systematically utilized to probe materials properties at interfaces. They are also involved in investigating nanoparticles, nanostructured surfaces, nanocomposite materials, 3-D printing, coatings, biomaterials, and process development.