| Chemical Networks | |
|
Part of our "philosophical" fascination with self-organization derives from the fact that self-assembly and self-organization pertain not only to things which we, humans, control but also to ourselves. In our work, we examine one such area of human activity -- synthesis of new chemicals. Here, our group was first to look at chemistry en large as a network/graph of directed reaction "edges" connecting molecule "nodes" (Angew. Chem. 2005, 2006). Out network comprises all (yes, all) organic reactions reported in literature from the times of Lavoisier and the present. By analyzing this monstruous construction with the tools of network theory and statistical physics, we discovered the laws that govern all synthetic transformations carried out to date or to be carried out in the future. The amazing regularity embodied in these laws not only poses intriguing questions as to how the collective work of apparently autonomous agents (here, chemists) gives rise to a well defined, higher-order structure, but is also of practical value since it enables identification of the most useful chemicals, prediction of the efficiencies of new chemical transformations, the properties of the most likely products, prices of specific chemicals, and more. Currently, we are actively working on applying network models to predict reactivities of individual molecules and on algorithms that optimize multistep chemical syntheses. These project are a rather unique mix of high-end graph theory and organic synthesis -- as such, they are highly collaborative and we currently have two physicists and two chemists working on them. Bartosz thinks that the fact that these theoretical physicists and organic chemists can collaborate and communicate with one another is a scientific miracle in itself.
|
|
|
grzybowski-group.net |