A catalyst is a substance that increases the rate at which a reaction occurs without being consumed. In this project organometallic catalysts (organic catalysts that contain a metal centre) and biomolecular catalysts (e.g. enzymes) are of particular interest. These two classes of catalysts have complementary characteristics and the aim of the current project is to combine these by embedding the metal center into an enzyme, resulting in a hybrid catalyst. The challenge that this idea invites is to find an appropriate enzyme with a binding site that can accommodate an organometallic complex and subsequently to fine-tune the catalytic ability of the hybrid catalyst for the desired reaction. In general, the binding sites of the enzymes we wish to investigate have not evolved to contain metals and thus the inclusion of the metal center will have unforeseeable implications on the structure and reactivity of both the enzyme and the organometallic catalyst. The role of this project is to predict what these changes in structure and reactivity will be and then to modify the original system to obtain the desired reactivity. This structure-based approach of designing molecules is known as rational design, whereby the surrounding residues (and consequently the molecular architecture) are mutated based on their known properties and orientation. The knowledge of the chemical and structural properties of the system is gained through computational modelling, which allows us to visualize the system and predict how it will react to certain chemical modifications. For this purpose we use a range of computational methods, including low-level methods (molecular mechanics) in order to analyse structural changes; and high-level methods (such as quantum mechanics) in order to analyse changes in the chemical properties of the system, such as activation barrier heights. Large systems, such as enzymes, are far too expensive, computationally, to be treated with quantum mechanical methods, thus a hybrid quantum mechanical/molecular mechanical (QM/MM) method is employed to study the reactivity of the resulting hybrid catalyst.Organometallic catalysts are used in a wide range of industrial processes; most notably in the production of finechemicals and pharmaceuticals. However, compared to enzymes, the rate at which these compounds catalyse reactions and the corresponding turnover numbers (i.e. amount of product produced per catalyst) is quite small. One of the advantages of using hybrid catalysts will be in the increase of the control of the reaction mechanism, which leads to faster rates. Furthermore, in the hybrid catalyst one may be able to obtain greater control over the specificity of the reaction; this implies that protecting groups (used to stop side reactions occurring in organometallic catalysis) are no longer necessary. By reducing the chance of side reactions we decrease the amount of starting material required to produce a given quantity of the actual product. Thus hybrid catalysts offer the possibility of performing chemically challenging syntheses in a much more efficient manner than is possible with traditional organometallic catalysts. This greater efficiency is achieved through a reduction in the amount of catalyst required, a reduction in the number of reactions steps required and an increase in the percentage yield of the desired product. With hybrid catalysts we get more product for less - which makes sense both economically and environmentally.
In this project we sort to understand the way in which compounds containing transition metals are able to catalyse (speed up) challenging chemical transformations. In order to do this we investigated a range of transition metal catalyzed reactions including:
- Rhenium based compounds (MTO) which can be used to produce dihydrogen trioxide (a compound which plays an important role in understanding the chemistry of the atmosphere).
- Synergic Catalysts that have two different metals involved in the chemical reactions (e.g., Potassium-Magnesium and Zinc-Lithium).
- Nickel based compounds that are able to carry out extremely challenging reductions of organic substrates.
- The ability of Tungsten to form stable N-Heterocyclic Carbene complexes in high oxidation states.
These various computational studies that were carried out in collaboration with our experimental partners helped reveal the important role that the transition metal played in carrying out the desired chemistry. The results of these studied have been published in internationally-leading peer reviewed journals.
Understanding the important role of transition metals was the first step for our ability to incorporate these metals into biochemical systems in order to enhance their ability to carry out novel chemistry. However, before we could include any transition metals into a biochemical setting (proteins and peptides in our case) it was necessary to understand the factors that govern the structure and dynamic nature of proteins and peptides. Therefore, the second main goal that was achieved through this work was the development of methods that could be used to study at a hybrid quantum mechanical and molecular mechanical level of theory the dynamic nature of proteins and peptides. To this end we developed the OMx-D methods. OMx-D is a semi-empirical method which we showed was able to be effectively combined with a molecular mechanics method (CHARMM) in order to study properties such as ligand binding and peptide conformations during relatively long (up to 100 ps) timescale QM/MM MD simulations. The length of the QM/MM MD simulations was sufficient to observe small perturbations in the binding site of a protein, or to observe minor conformational changes in small chain peptides. However, in order to study spontaneous self-assembly processes (such as those that we would like to observe in the formation of an active site around a metal in the hybrid catalysts) it was necessary to extend our time-scale regime by several orders of magnitude, up to microseconds. That is, 10,000 times longer than our longest QM/MM MD simulation. In order to access such timescales we needed a different method that was able to rapidly screen the self-assembly propensities of various biochemical systems (e.g., peptides) in order to assess their ability to form ordered structures capable of supporting a catalytic active site. We managed to achieve this goal through the use of coarse grained methods. These methods were found to be capable of accurately describing, not only the final structure of a self-assembled system, but also the process by which disordered peptides are able self-assemble in to structures such as tubes and fibres.