Final Activity Report Summary - REACTIVE MD (Reactive molecular dynamics for Organometallic reactions relevant to asymmetric hydrogenation reactions)
Asymmetric hydrogenation is a very useful chemical reaction for the synthesis of pharmaceuticals and natural products in an enantiopure form; that is, without being contaminated by the mirror image of the desired compound, which in the best case represents a 50 % waste, and in the worst case can have undesirable biological activity such as side effects. However, this reaction depends on the availability of a suitable catalyst. Catalysts are usually organometallic compounds (molecules containing bonds between carbon and a metal atom), but they all have a limited scope and it is very difficult to predict which catalyst will be best for a given reaction without carrying out expensive experiments.
The goal of this project was to develop efficient theoretical methods that can eventually be used to predict the selectivity of a catalyst by means of computer simulations, thus saving time and effort. A key part of this process is to be able to calculate the energies of the many possible intermediate species that are involved in the reaction. While there are methods based on quantum mechanics that are often used for this purpose, such methods have practical limitations because they are slow and expensive to use. We developed a method based on molecular mechanics, a much faster approach, by incorporating a function based on atomic orbital overlap that accounts for the 'trans influence', a remarkable electronic interaction between atoms bound on opposite sides of a metal atom.
By adjusting the parameters in our function to best fit a set of reference data obtained from quantum mechanics on small molecules, we were able to develop a method that can predict which configuration of an organometallic compound will be most stable, even for large molecules, within a timeframe of seconds using a typical desktop computer. For comparison, quantum mechanical methods on similar molecules can take days and may require a supercomputer. Our method can be incorporated into sophisticated simulations of reactions in solution, and promises to become a powerful tool for the study of chemical reactions involving organometallic compounds.
The goal of this project was to develop efficient theoretical methods that can eventually be used to predict the selectivity of a catalyst by means of computer simulations, thus saving time and effort. A key part of this process is to be able to calculate the energies of the many possible intermediate species that are involved in the reaction. While there are methods based on quantum mechanics that are often used for this purpose, such methods have practical limitations because they are slow and expensive to use. We developed a method based on molecular mechanics, a much faster approach, by incorporating a function based on atomic orbital overlap that accounts for the 'trans influence', a remarkable electronic interaction between atoms bound on opposite sides of a metal atom.
By adjusting the parameters in our function to best fit a set of reference data obtained from quantum mechanics on small molecules, we were able to develop a method that can predict which configuration of an organometallic compound will be most stable, even for large molecules, within a timeframe of seconds using a typical desktop computer. For comparison, quantum mechanical methods on similar molecules can take days and may require a supercomputer. Our method can be incorporated into sophisticated simulations of reactions in solution, and promises to become a powerful tool for the study of chemical reactions involving organometallic compounds.