How to calculate the Delta G of solvation? This is a question that I get a lot in this blog, so it is about time I wrote a (mini)post on it, and at the same time put an end to this posting drought which has lasted for quite a few months due to a lot of pending work with which I’ve had to catch up. Therefore, this is another post in the series of SCRF calculations that are so popular in this blog. For the other posts on this subjects remember to click here and here.

SMD

SMD is the keyword you want to use when performing a Self Consistent Reaction Field (SCRF) calculation with G09. This keyword was only made available in this last version of the program and it corresponds to Truhlar’s and coworkers solvation model which is recommended by Gaussian itself as the preferred model to calculate Delta G of solvation. The syntax used is the standard way used in any other Gaussian input files as follows:

# 'route section keywords' SCRF=SMD

Separately, we must either perform a gas phase calculation or use the DoVacuum keyword within the same SCRF input, and then take the energy difference between gas phase and solvated models.

# 'route section keywords' SCRF=(SMD,DoVacuum)

No solvation or cavity model should be defined since, by definition, SMD will use the IEFPCM model which is a synonym for PCM.

As opposed to the previous versions of Gaussian, the output energy already contains all corrections, this is why we must take the difference between both values (remember to calculate them both at the same level of theory if calculated separately!). Nevertheless, when using the SMD keyword we get a separate line, just below the energy, stating the SMD-CDS non electrostatic value in kCal/mol.

The radii were also defined in the original paper by Truhlar; I’m not sure if using the keyword RADII with any of its options yields a different result or if it even ends in an error. Its worth the try!

Some calculation variations are not available when using SMD, such as Dis (calculation of the solute-solvent dispersion interaction energy), Rep (solute-solvent repulsion interaction energy) and Cav (inclusion of the solute cavitation energy in the total energy). I guess the reason for this might be that the SMD model is highly parametrized.

Have you found any issue with any item listed above? Pleases share your thoughts in the comments section below. As usual I hope this post was useful and that you all rate it, like it and comment.

References

A. V. Marenich, C. J. Cramer, and D. G. Truhlar, “Universal solvation model based on solute electron density and a continuum model of the solvent defined by the bulk dielectric constant and atomic surface tensions,” J. Phys. Chem. B, 113 (2009) 6378-96.
Advertisement