Project
DFT-based evaluation of dinuclear Mn-X (X=Ti, Al, B) complexes with regard to the feasibility to hydrogenate carbon dioxide and other challenging substrates
The hydrogenation of carbon dioxide (CO2) to methanol with hydrogen (H2) is an interesting, though challenging chemical transformation wich requires inevitably the use of sophisticated molecular catalysts. Catalysts which are either as yet not avaiable or do not show convincing turn over parameters – such as turn over frequency and/or turn over number. If such a transformation could be established using a molecular catalyst it might in principle be possible to open the door into an industry synthesis alternative for the generation of methanol. And this, in turn, could be interesting for establishing a sustainable methanol based economy.
The traditional way to design molecular catalysts has over decades been an experience based experimentally conducted tedious process with many failures, as in pure experimental work it is often not very clear if a desired/given catalyst will be able to enable the reaction. The reasons for this often are the thermodynamic and kinetic challenges associated with a given reaction. For instance CO2 is a thermodynamically very stable molecule and in any transformation high activation barriers must be exptected. It is therefore an enourmous advantage that such barriers can nowadays be computed and evaluated to a sufficient degree of excatness before any experimental work is undertaken. This is especially helpful in view of the fact that many catalyst syntheses are very challenging in itself. As such High Performance Computing is an enourmously helpful tool, as all kinds of catalysts can be designed computationally and evaluated before any experimental work is done.
In this project dinuclear catalysts containing both manganese and a second metal serving as a Lewis acid were investigated to answer the question if complete catalytic cycles for the CO2 hydrogenation to methanol can be localized with an overall activation barrier low enough to envision experimental realization. This project is part of our research approach in which catalysts for chemically challenging transformation first are developed in silico and then -when successfull- are transferred to experimental work.
In this study we focused on manganese and titanium as active catalyst components built into one single catalysts molecule to hydrogenate CO2 with H2.
Project Details
Project term
July 1, 2025–June 30, 2026
Affiliations
RWTH Aachen University
Institute
Institute of Technical and Macromolecular Chemistry
Principal Investigator
Methods
The project was conducted on the RWTH Aachen University NHR partition by using static density functional theory (DFT) computations employing the Gaussian 16 suite of programs. We chose state of the art density functionals ( for instance MN12-L, B97D3BJ) and basis sets of TZVP quality to ensure that activation barriers were obtained which closely would resemble experimentally expectable barriers. The computations were carried out in the gas phase ininitally and then all local minima and transition states were reoptimized in condensed phase using implicit treatment of solvents by the Polarizable Continuum Method (PCM) applying the SMD radii formalism developed by the Truhlar group.
Results
For the reaction of CO2 with H2 to methanol we were able to design a catalyst system which -after having computed complete catalytic cycles including undesired side reactions- generates an overall activation barrier low enough to justify experimental work. That particular catalyst system has now been chosen to be synthesized experimentally and then tested for catalytic performance. Other substrates such as methyl benzoate and 1,3-diphenylurea which also would be interesting to hydrogenate with such a catalyst system, could as yet not be checked as the work to design the aforementioned catalyst for the CO2 hydrogenation took much longer than initially anticipated.
Discussion
A variety of catalysts for the transformation of CO2 to methanol with H2 was studied and – retaining the core catalyst structure – subtle changes in the ligand periphery made it possible to localize a catalyst system which shows an overall activation barrier of slightly above 30 kcal/mol. This is indicative of a catalyst which at elevated temperatures might indeed catalyze this reaction. It might certainly turn out that further fine tuning of the catalyst would yield structures generating an even lower activation barrier. We will in the future continue to study this and similar systems also for other hydrogenations of challenging substrates and we are currently undertaking experimental studies to synthesize that computationally identified catalyst to check its catalytic performance.
Additional Project Information
DFG classification: 302-03 Chemical Solid State and Surface Research, Theory and Modelling
Software: Gaussian16
Cluster: CLAIX