Tessellated Inc., in collaboration with the University of Notre Dame (ND), proposes to perform extensive, multi-mode characterization of polymer composite films to understand and provide critical guidance to improve the thermal conductivity of such materials to above 70 W/mK. In addition, our cost analysis shows our composite films to be <$5/kg. It can also have high electrical conductivity of up to 0.1 S/cm, and a tensile strength up to 3 GPa. This proposal targets Topic 9b with a focus on characterization. In this Phase I project, the team will leverage its patent-pending roll-to-roll (R2R) manufacturing process for polymer composite films. This process consists of four major steps (Fig. 1): (a) filler-matrix mixing; (b) high shear rate extrusion; (c) drying; and (d) hot drawing. We will focus on ultrahigh molecular weight polyethylene (UHMWPE) as the matrix and graphite, graphene and hexagonal boron-nitride (h-BN) as the filler, with the hypothesis that the microstructures of the fillers and the UHMWPE matrix depends on the manufacturing process and the filler-matrix interaction, and by optimizing them, the thermal conductivity of the composites can be improved. Our preliminary results show that an UHMWPE/graphite composite (30wt% graphite) film led to a thermal conductivity of 50-70 W/mK with a drawing ratio of ~60×. The undrawn sample has a high electrical conductivity of ~0.1 S/m, improved from 10-5 S/m. In addition, the drawn samples mechanical tensile strength is as high as 3 GPa. In this project, we will systematically tune the process parameters and fillers to fabricate the films, and use techniques like XRD, Raman, SEM and TEM at ND, and GIWAXS/SAXS at ANL APS (via beam time application), to characterize the morphology, with the focus on the overall polymer matrix crystallinity and that in the vicinity of filler surfaces, micro-voids and the distribution and exfoliation state of the fillers. For these films, we will extensively test their thermal conductivity and electrical conductivity, as well as mechanical properties. We have access to all needed equipment for these characterizations and tests. The collected data will be analyzed quantitatively to establish a process-structure-property relation. Since we expect such a relation to be high-dimensional and non-linear, we propose to use machine learning (ML) techniques, such as Gaussian Process (GP), to model this relation. Finally, we will use the ML model with BO to identify the optimal process parameter combinations that would enable the consistent fabrication of high thermal conductivity polymer films with ideal electrical and mechanical properties. The collaboration between Tessellated, who has licensed the manufacturing patent, and Prof. Luo at ND, who is a leader in studying thermal transport in polymers and their composites, make the team uniquely qualified to carry out the proposed work.
