Running COSMO for acetone as the solvent


Clicked A Few Times
I'm a junior chemistry major working on computational research for my adviser. My project will ultimately study beta effects in basic organic reactions, both in vacuum and in different solvents. Due to the involvement of calculations under solvent conditions, I have been trying to learn more about using COSMO with NWChem (in ecce) for different solvents by attempting to calculate molecular energies for dichloromethane in water, acetonitrile, and acetone. I have found a few sites identifying keywords to use for coding for the different solvents. I have run calculations successfully with water and acetonitrile using the corresponding dielectric constants, radii of the atoms in the molecules, and the indices of refraction. However, when trying to use the same parameters to run an energy calculations with acetone as the solvent, the calculations is always incomplete.

The parameters I have found to use include the dielectric constant, the index of refraction, the van der waals radius of the solvent, the effective molecular radius of the solvent, the Distance threshold for using two-point interaction approximation, and the number of segments per atom

Another site I found also lists the radii of the atoms of the molecule of the solvent and the solvent's accessible surface as additional parameters.

I do not completely understand all of these parameters, particularly the difference radii that can be used as parameters. Would someone be able to explain these parameters to me and help me learn how to run a COSMO energy calculation with acetone as the solvent?

Forum Vet
I would recommend you read the original article of Klamt, which describes all the parameters and their function. See the citation at the end of http://en.wikipedia.org/wiki/COSMO_Solvation_Model .

A little surprised you were able to run acetonitrile and not acetone. They have similar dielectric constants, and the only difference in input is really that.

Did you check your output file (not just what ecce tells you). For example, the energy convergence or the optimization (if you're doing a geometry optimization) might need a few more iterations to complete. If this is the case, you can increase those in ecce too.

Bert


Forum >> NWChem's corner >> Running NWChem