CCL:G: van der Waals parameters for Ti

 Sent to CCL by: "Lukasz Cwiklik" [cwiklik[]]
 Dear Saurabh,
 I do not know any tutorial, however, I will try to describe the
 general procedure.
 The way to estimate vdW parameters with ab initio (eg. Gaussian) is to
 scan a potential energy surface for interactions between your atom
 (rutile) and the atoms/molecules which rutile will interact with in
 your simulation. In the simples case, if you have just rutile-rutile
 interactions, make a scan of energy in Gaussian  for changing the
 rutile-rutile distance. Then assume the parameters for rutile (for
 starters, take radius and  well depth for a similar atom) and make the
 same scan in your MD code. Then the hard part starts - you need to fit
 (just by hand or write a code) the radius and  well depth to have an
 agreement between ab initio and the empirical energy vs. distance
 curve (in a range of method's error which sometimes can be as high as
 2 kcal/mol).
 One should remember about the basis set superposition error, however,
 there is no ideal method to get rid of bsse in this case.
 Lukasz Cwiklik
 On Mon, Nov 24, 2008 at 4:02 PM, Saurabh Agrawal
 Saurabh.Agrawal^^ <> wrote:
 > Sent to CCL by: "Saurabh  Agrawal"
 > Dear CCl users,
 > I am trying to get non bonded parameters (van der Waals' radius and  well
 depth) for Ti atom (rutile). I appreciate if some one could provide these
 parameters to me or let me know the way (tutorial) to calculate these parameters
 using Gaussian.
 > Thanking you in advance.
 > Best,
 > Saurabh Agrawal,
 > PhD student,
 > University College Dublin,
 > Ireland>