Machine-learning-assisted development of alternatives to diarylide pigments

The goal of the proposed effort is to develop polychlorinated biphenyls (PCB)-free yellow pigments and synthetic pathways that do not unintentionally generate PCB as byproducts. This effort will harness Kebotix’s existing AI-powered platform, which includes a workflow that controls high?throughput computational modeling of molecules and materials, proprietary machine learning methods for process control and optimization, deep-learning algorithms for prediction of molecular properties, generative models, as well as automated high-throughput preparation and characterization capabilities. The proposed closed-loop approach to design and discovery of pigments is uniquely efficient because it will intentionally include molecular properties (e.g., optical and thermal) and materials compatibility in the generation of virtual libraries. In turn, this approach will translate into a reduced cost of pigment design and discovery. Kebotix has already generated preliminary results and engaged a number of companies all of whom have expressed interest in testing and buying new halogen- and heavy metal-free pigments that our technology discovers and develops. Increasing demand from various industries, such as textiles, plastics, paint and coatings, and construction, is expected to drive growth. Data suggest that the use of pigments and dyes is significantly contributing to the global distribution of PCBs, which are synthetic chemicals that do not occur naturally. Thus, there is an increasing need to actively use advance technologies to discover new materials and manufacturing routes that do not produce PCBs and other toxic by?products. Intentional manufacture of PCBs has been banned throughout the developed world. However, they are still generated during the production of certain classes of pigments.