Hi all,
I have a polar organic compound, typically found in the human extra-cellular matrix. I would like to make a quantifiable measure of whether certain ions (Fe II/III, Cu II, Mg, Co, Zn, Ca) have a binding affinity for -OH functional groups and an imidazole ring. Using DFT/ab initio techniques in Gaussian 09. What would be the best approach?
I have experience with this particular molecule and its reaction pathway, but I lack in the experience of including transition metals. What type of basis set and functional should I consider (I understand this is quite an open question) and how does one go about measuring ionic binding affinity?
Many thanks
Anthony
Hello Anthony,
For this case I suggest two options.
If the system that you use can have dispersion forces, the Minnesota functional (M06-2x).
The other option, and I think is the best, is use the B3PW91 (B3LYP only shoud be used for small organic molecules only and can't explain metals properly).
Also think that can be some problems with spin states where you will see a very different trend in energy or some problem. If this success, hybrid method could stabilize the high spin of metal. Then you must use BHLYP... Even so, I think B3PW91 will give good results, but a MP2 calculus to compare energy could be a goo idea to detect some general problem.
About basis, I prefer the Dunning ones. Are little expensive but I think that are better than traditional Pople. An option is do the OPT with DZV and energy with TZV.
I hope it helps you,
Joaquim
A few years back we did a study on Zn binding to amino acid models: http://www.ncbi.nlm.nih.gov/pubmed/17924152 . I guess the methodology is still sound, but you should obviously replace B3LYP with some other modern functional known to describe metals properly (as Joaquim suggested).
Hi both,
Thank you for that information. I should at least be able to build some models and use an appropriate functional/basis set. Thanks.
Don't forget that DFT and ab initio methods give "vacuum" (gas-phase) values of the interaction. Modelling solvation effects is still a challenge.
Hi Jean-François,
I'm currently modelling the reaction pathway of the organic compound (minus metals) at the moment. I've been working on it for months, and I've come to the conclusion that the greatest difficulty is modelling with a solvation. I've switched to gas - which is a shame.
Well, it's a first step, as for al of us... You may try adding a few water molecules to the cation to have a first guess. There are also continuous models which are much less efficient to reproduce specific solvation.
I have a question following the discussion here. What basis sets approach shall I consider for simulating ionic bonding seen in cement constituents like dicalcium silicate and tricalcium silicate?
Any other special parameters I need to consider?
- Badges
- Science topic
- Similar topics
- Engineering
- Chemical Engineering