From owner-chemistry@ccl.net Sun Mar 2 19:20:01 2008 From: "David Hose anthrax_brothers:hotmail.com" To: CCL Subject: CCL:G: G03: Non-aqueous pKa calculation (PCM methods) Message-Id: <-36391-080302191828-22355-EKT7UR9pzV3Aqt1aNgAwMg|server.ccl.net> X-Original-From: "David Hose" Date: Sun, 2 Mar 2008 19:18:24 -0500 Sent to CCL by: "David Hose" [anthrax_brothers]^[hotmail.com] Hi Ya Guys, I would appreciate your advice on a problem I have. Firstly, I'm a new user to Gaussian '03 and still learning my way around the package. The first serious project that I want to run using it, is the calculation of non-aqueous pKa's of a series of nitrogen containing bases. From posts over the last couple of months to this list, I'm familiar with the general approach; optermise the geometry in the gas phase, determine the solvation energy at the same geometry, and then apply the appropriate change of standard state (i.e. from 1 atm to 1 mol / L). The substrates that I'm interested in are relatively small, the largest of which contains 13 heavies (C and N only). On that basis I don't mind pushing the boat out and using a composite method like CBS-Q to obtain the most accurate gas phase result I can. The problem I have is how to calculate the solvation energy. Of the methods available in Gaussian '03, what do people recommend as being the 'best'? What base method should I consider (B3LYP/6-31G* or increase the basis set size to something comparable to the largest basis set used in the CBS-Q method)? The final thing is, I want to do these calculations for DMF and NMP. > From the Gaussian manual it states that four parameters are required EPS (static dielectric constant), EPSINV (dielectric constant at infinite frequency [can someone recommend a source for this?]), density and RSOLV (solvent radius). It is latter parameter that I'm particularly interested in. Is this the radius of the longest axis, shortest axis, average of the two? Should is be multiplied by a scaling factor or do I add say 0.5 A to the answer and use that? Any help or references greatly appreciated. Regards, Dave.