Synthesis Planning and Reaction Discovery For Photochemistry and Chemistry in Novel Environments
Abstract
This MURI program aims to develop an automated toolset for chemical synthesis planning involving the novel chemistries that will be required to produce myriad types of nextgeneration molecular systems. In Task 1, project PI Todd Mart??nez (Stanford) will extend the recently-developed ???nanoreactor??? concept for abitio discovery of chemical reactions to perform reaction discovery in solvent, in molten salt, and in photochemical transformations, supported by co-PI Steven Lopez (Northeastern), who will develop next-generation reaction rate methodology for use in the refinement stage of the nanoreactor. In Task 2, co-PI Alan Aspuru- Guzik (Harvard) will develop machine learning approaches for synthesis planning on a known but incomplete reaction graph. The approaches in Task 2 will use a reaction graph that comes from a mixture of existing experimental knowledge and directed nanoreactor predictions, and will iteratively feed back into both of these domains. In Task 3, we will rigorously benchmark the theoretical methodology of Tasks 1 and 2 via a series of experimental studies, and will also directly explore novel chemistries that might lead to synthetic breakthroughs. Co-PI Matthew Kanan (Stanford) will perform a series of experiments in molten alkali salt environments, which have recently been shown to demonstrate an astonishing range of novel acid-base reactions. This chemistry will be used to validate the extended nanoreactor technology of Task 1. Co-PI Noah Burns (Stanford) will explore a broad range of photochemical approaches to the synthesis of saturated strained ring molecules to test the new photochemical capabilities of the nanoreactor and the automated synthesis planning tools. Co-PI Stephen Bradforth (USC) will perform ultrafast experiments to elucidate and test the fine details of the photochemical reaction mechanisms predicted by the theoretical approaches, and will explore the use of high-throughput screening schemes to accelerate synthetic searches from the experimental perspective. The work proposed herein has the potential to transform the way that synthetic chemistry is performed, both by providing chemists with a powerful automated toolset for reaction discovery and synthesis planning, and by significantly expanding our knowledge of molten salt and photochemical reactions. The tight and two-way coupling between theory and experiment being pursued in this work is part of what we view as a much larger paradigm shift that is presently blurring the lines between strictly observational or predictive approaches to chemistry and physics. We firmly believe that the day is near when it will be unthinkable for a chemical physics experiment to be performed without a theoretical component, and that the lessons learned in the present work will accelerate progress toward this eventuality. Moreover, the classes of reactions being targeted in this work have significant potential to provide synthetic breakthroughs for a number of important target systems important to DOD???s mission. One of our direct targets in this work is to investigate photochemical pathways to highly strained ring systems, which have obvious applications in explosives and propellants. Beyond this, the toolset and domain knowledge gained during this project will likely lead to breakthrough synthetic pathways with applications in photomechanical and mechanochemical systems.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Jul 26, 2018
- Source ID
- N000141812659
Entities
People
- Todd Martinez
Organizations
- Office of Naval Research
- Stanford University
- United States Navy