(628e) Optimal Reaction Pathway Design Using Computer-Aided Process Chemistry

Chemical products should be designed according to the principles of green process chemistry for a sustainable society. To ensure that the materials produced are environmentally friendly, renewable, and hazard-free, important intrinsic properties of the chemicals must be considered from the design stage using sustainable performance indicators. One way to achieve this is to use computer-aided chemical synthetic planning, which has recently received much attention and is becoming increasingly important in organic chemistry. However, because there are numerous feasible routes to choose from in synthesizing the target compound from starting materials, selecting which pathway is optimal according to process chemistry criteria is challenging because of the high dimensionality of the chemistry and reaction search space. This work identifies the best synthetic route to a given drug substance by formulating an optimization problem to minimize green process chemistry principles in large chemical reaction networks. We consider multiple factors including, the number of reaction steps, cost of materials, and atom economy in selecting the optimal reaction network. The atomic economy is a measure of the efficiency of the synthesis process defined by the ratio of the desired product's total mass to the products' total mass. In the developed workflow, after retrosynthesis analysis is performed using computer-aided synthesis planning (CASP) with ASKCOS (Coley et al. Science 2019) the optimizer takes in the generated reaction pathways as a tree builder graph. Then, optimization is run to solve a multi-objective function that is formulated as a mixed-integer programming problem based on the chosen process metrics, and finally, the optimal combination of pathways to the target is found. As a result, solving the multi-objective optimization problem in viable reaction pathway networks can effectively reduce the number of steps and materials costs while increasing atom economy. Incorporating green chemistry considerations in computer-assisted synthesis planning can improve chemical synthesis processes, aid in experimental design, and consequently facilitate process development for fine chemicals or pharmaceuticals. This work can be further expanded by including other useful metrics encompassing process mass intensity, impurity prediction score, product yield, and solvent score.