Bond-making and bond-breaking can also be modeled using free energy perturbation (FEP) simulations in the empirical valence bond (EVB) approach . . . which is superior to QM/MM.
Dear Mr. Mohamed,
Roughly speaking, this a combination of two approaches:
1- QM method; this is mainly for the purpose of increasing the results' accuracy.
2- MM method; this is for reducing time consumption.
This means the QM/MM approach is introduced to study relatively big chemical systems, such a protein case.
Here is the link, which I hope this can help a bit. But still, I am not quite sure what is exactly your question about?
http://anusf.anu.edu.au/~vvv900/qm-mm.html
--Best,
Thank you for all of your helps and enriching explanations
Regards
M. El Hasnaoui
A QM/MM implementation is provided within NWChem. It is decidedly not user-friendly but the developer has been quite helpful in guiding me through the idiosyncracies of the implementation. The advantage here is that one has access to very high-level quantum methods in the QM partition, provided that you have the computational resources at your disposal to perform such calculations. I have successfully run simulations with ~100 quantum atoms with B3LYP/6-31++g** level of theory. In principle, one could also perform coupled-cluster types of simulations, given the processing resources.
Such high-level calculation, I believe, are superior to the EVB method mentioned above. Conducting QM/MM simulations, though, is relatively painful. NWChem only supports the AMBER force field for the MM partition. The implementation requires performing each step in a separate directory and a good deal of copying files back and forth. If you are meticulous, you can study the chemistry taking place within a small volume that is influenced by the electric field produced by the larger environment surrounding the volume. In my work, this enabled me to investigate different chemical pathways in enzymatic catalysis. Although, I will admit that the precision available is meager. One can certainly differentiate between two pathways where the barrier heights differ by 100 kJ/mol but you would probably not be able to distinguish between pathways where the barriers differed by 10 kJ/mol.
There is significant tradecraft required to undertake such simulations. Defining the QM partition almost always requires treating part of one molecule with QM and the other part with MM. How that is done is tricky and can greatly influence the outcome. For the NWChem implementation, you must always break the molecule across a single bond. (Between the Calpha and Cbeta carbons in an amino acid, for example.) As things stand, QM/MM provides the ability to ask and answer questions that other methods cannot but is not something that one can access trivially. You must narrowly define what you want to ask and prepare yourself for exerting signficant effort to accomplish those goals.
Thank you dear Mark,
Does software NWChem free or must buy it.
Regards
M. El Hasnaoui
If you are working on biomolecular systems, you can also consider to use CHARMM+GAMESS-US for ab initio QM/MM calculation and CHARMM+DFTB for semiempirical QM/MM calculation/simulation. They are all free now.
Here are two papers for your reference,
Article X-ray Structure of a Hg2+ Complex of Mercuric Reductase (Mer...
Article Catalytic mechanism and origin of high activity of cellulase...