[gmx-users] RE: Coul-14, LJ-14 and RF-excl definitions (2)

pascal.baillod at epfl.ch pascal.baillod at epfl.ch
Wed Mar 12 16:26:05 CET 2008

Dear developers,

I would again like to thank David, Xavier and Berk for their very informative
explanations on RF-excl and 1-4 interactions. I am still a bit confused, though,
with Berk's very last statement on this issue. If I quote it:

"The reaction field is not applied to pair (1-4) terms. Therefore there are no
issues with fudgeQQ."

As far as I understood, the reaction field correction IS applied to (1-4) pairs,
but added to the RF-excl term, and not to Coul-14 term. At least this is what I
understood from another explanation from Berk, sent on the mailing list sometime
around September 2007:

"The reaction-field correction applies to ALL atom pairs that are within the
cut-off distance(or more accurately: atom pairs for which their charge group
centers are within the cutoff distance). So all "noraml non-bonded interaction
pairs, as well as all excluded pairs including self -pairs. The only issue is to
which energy term wich contribution is added. In old Gromacs versions the RF
correction for 1-4 pairs was added to the 1-4 energy term. In the newer version
its added to the RF-excl term."

I would also like understand why the reaction field correction is applied to
excluded atom pairs. The corrected coulomb potential described in the manual
reads (please open the attached pdf for the compiled equations):

V_{crf} = f *q_i *q_j [ \frac{1}{r_{ij}} + k_{rf}r^2_{ij} - c_{rf} ] (1)

This is equivalent to

V_{crf} = V_c  + V_{rf} (2)

where V_c is the usual Coulomb potential energy, and V_{rf} is the  
reaction field correction.

For excluded atom pairs, there is no Coulomb interaction, except for 1-4 pairs,
where there is a reduced Coulomb interaction. However, in agreement with what I
state above, the code dose seem to compute a reaction field correction for all
excluded atom pairs. In the RF_excl_correction routine of mdlib/rf_util.c, this
correction reads:

V_{rf} = f q_i  q_j   ( k_{rf}r^2_{ij} - c_{rf} )

That means we apply a potential that is a function of r^2, yielding a force
whose magnitude increses with r and whose direction is opposed to the Coulomb
force !!! What is the physical justification to V_{rf}  in the absence of V_c
for excluded pairs?

I thank you very much for your help!


Pascal Baillod (PhD student) 
Swiss Federal Institute of Technology EPFL	        Tel: +41-(0)21-693-0322
Institute of Chemical Sciences and Engineering ,	Fax: +41-(0)21-693-0320
Laboratory of Computational Chemistry and Biochemistry	pascal.baillod at epfl.ch
Room BCH 4121, Avenue Forel,	                        http://lcbcpc21.epfl.ch
CH-1015 Lausanne	
-------------- next part --------------
A non-text attachment was scrubbed...
Name: eqRF.pdf
Type: application/pdf
Size: 9150 bytes
Desc: not available
URL: <http://maillist.sys.kth.se/pipermail/gromacs.org_gmx-users/attachments/20080312/84835ffc/attachment.pdf>

More information about the gromacs.org_gmx-users mailing list