CCL:G: ONIOM with PCM

[Soren Eustis <soreneustis . gmail.com>]

Title: ONIOM with PCM

I have been studying the different behaviors of explicit and implicit solvent models on a given system.  One of my final calculations was to be a QM:MM calculation (1 molecule:200 waters) with IEFPCM (solvent=water).  The previous run (ONIOM without PCM) ran fine and converged after a reasonable time.  However with PCM and after 168 processor hours, the calculation has not left the first MM stage.  See below:

---

INPUT:
%chk=output.chk
%mem=8GB
%nprocshared=8
# opt oniom(mpw1pw91/cc-pvdz:amber) scrf=(iefpcm,solvent=water,> geom=connectivity

22DCDPA in water box with IEFPCM

0 1 0 1 0 1
....

OUTPUT:
....

 ONIOM: generating point  3 -- low level on real system.
 ONIOM-PCM-X: Computing reaction field of low level on real system.
 Standard basis: Dummy (5D, 7F)
 There are  2130 symmetry adapted basis functions of A   symmetry.
 Integral buffers will be    131072 words long.
 Raffenetti 1 integral format.
 Two-electron integral symmetry is turned on.
  2130 basis functions,  2130 primitive gaussians,  2130 cartesian basis functions
  1424 alpha electrons     1424 beta electrons
       nuclear repulsion energy    158651.1978886932 Hartrees.
 NAtoms= 2130 NActive= 2130 NUniq= 2130 SFac= 7.50D-01 NAtFMM=   80 NAOKFM=T Big=T
 ------------------------------------------------------------------------------
 Polarizable Continuum Model (PCM)
 =================================
 Model                : PCM.
 Atomic radii         : UFF (Universal Force Field).
 Polarization charges : Total charges.
 Charge compensation  : None.
 Solution method      : Iterative solution.
 Cavity type          : Scaled VdW (van der Waals Surface) (Alpha=1.100).
 Cavity algorithm     : GePol (No added spheres)
                        Default sphere list used, NSphG= 2130.
                        Lebedev-Laikov grids with approx.  5.0 points / Ang**2.
                        Smoothing algorithm: Karplus/York (Gamma=1.0000).
                        Polarization charges: spherical gaussians, with
                                              point-specific exponents (IZeta= 3).
                        Self-potential: point-specific (ISelfS= 7).
                        Self-field    : sphere-specific E.n sum rule (ISelfD= 2).
 1st derivatives      : Analytical E(r).r(x)/FMM algorithm (CHGder, D1EAlg=3).
                        Cavity 1st derivative terms included.
 Solvent              : Water, Eps=  78.355300 Eps(inf)=   1.777849
 ------------------------------------------------------------------------------
 AMBER calculation of energy and first derivatives.
 CoulSu:  requested number of processors reduced to:   2 ShMem   1 Linda.
 CoulSu:  requested number of processors reduced to:   1 ShMem   1 Linda.
 CoulSu:  requested number of processors reduced to:   2 ShMem   1 Linda.
 CoulSu:  requested number of processors reduced to:   1 ShMem   1 Linda.
 CoulSu:  requested number of processors reduced to:   2 ShMem   1 Linda.
 CoulSu:  requested number of processors reduced to:   1 ShMem   1 Linda.
 CoulSu:  requested number of processors reduced to:   2 ShMem   1 Linda.
 CoulSu:  requested number of processors reduced to:   1 ShMem   1 Linda.
 CoulSu:  requested number of processors reduced to:   2 ShMem   1 Linda.
 CoulSu:  requested number of processors reduced to:   1 ShMem   1 Linda.
 CoulSu:  requested number of processors reduced to:   2 ShMem   1 Linda.
 CoulSu:  requested number of processors reduced to:   1 ShMem   1 Linda.
 CoulSu:  requested number of processors reduced to:   2 ShMem   1 Linda.
 CoulSu:  requested number of processors reduced to:   1 ShMem   1 Linda.
 CoulSu:  requested number of processors reduced to:   2 ShMem   1 Linda.
 CoulSu:  requested number of processors reduced to:   1 ShMem   1 Linda.
 CoulSu:  requested number of processors reduced to:   2 ShMem   1 Linda.
 CoulSu:  requested number of processors reduced to:   1 ShMem   1 Linda.
 CoulSu:  requested number of processors reduced to:   2 ShMem   1 Linda.
 CoulSu:  requested number of processors reduced to:   1 ShMem   1 Linda.
 CoulSu:  requested number of processors reduced to:   2 ShMem   1 Linda.
 CoulSu:  requested number of processors reduced to:   1 ShMem   1 Linda.
 CoulSu:  requested number of processors reduced to:   2 ShMem   1 Linda.
 CoulSu:  requested number of processors reduced to:   1 ShMem   1 Linda.
 CoulSu:  requested number of processors reduced to:   2 ShMem   1 Linda.
 CoulSu:  requested number of processors reduced to:   1 ShMem   1 Linda.
 CoulSu:  requested number of processors reduced to:   2 ShMem   1 Linda.
 CoulSu:  requested number of processors reduced to:   1 ShMem   1 Linda.
 CoulSu:  requested number of processors reduced to:   2 ShMem   1 Linda.
 CoulSu:  requested number of processors reduced to:   1 ShMem   1 Linda.
 CoulSu:  requested number of processors reduced to:   2 ShMem   1 Linda.
 CoulSu:  requested number of processors reduced to:   1 ShMem   1 Linda.
 CoulSu:  requested number of processors reduced to:   2 ShMem   1 Linda.
 CoulSu:  requested number of processors reduced to:   1 ShMem   1 Linda.
 CoulSu:  requested number of processors reduced to:   2 ShMem   1 Linda.
 CoulSu:  requested number of processors reduced to:   1 ShMem   1 Linda.
 CoulSu:  requested number of processors reduced to:   2 ShMem   1 Linda.
 CoulSu:  requested number of processors reduced to:   1 ShMem   1 Linda.
 CoulSu:  requested number of processors reduced to:   2 ShMem   1 Linda.
 CoulSu:  requested number of processors reduced to:   1 ShMem   1 Linda.
 CoulSu:  requested number of processors reduced to:   2 ShMem   1 Linda.
 CoulSu:  requested number of processors reduced to:   1 ShMem   1 Linda.
 CoulSu:  requested number of processors reduced to:   2 ShMem   1 Linda.
 CoulSu:  requested number of processors reduced to:   1 ShMem   1 Linda.
 CoulSu:  requested number of processors reduced to:   2 ShMem   1 Linda.
 CoulSu:  requested number of processors reduced to:   1 ShMem   1 Linda.
 CoulSu:  requested number of processors reduced to:   2 ShMem   1 Linda.
 CoulSu:  requested number of processors reduced to:   1 ShMem   1 Linda.
 CoulSu:  requested number of processors reduced to:   2 ShMem   1 Linda.
 CoulSu:  requested number of processors reduced to:   1 ShMem   1 Linda.
 CoulSu:  requested number of processors reduced to:   2 ShMem   1 Linda.
 CoulSu:  requested number of processors reduced to:   1 ShMem   1 Linda.
 CoulSu:  requested number of processors reduced to:   2 ShMem   1 Linda.
 CoulSu:  requested number of processors reduced to:   1 ShMem   1 Linda.
 CoulSu:  requested number of processors reduced to:   2 ShMem   1 Linda.
 CoulSu:  requested number of processors reduced to:   1 ShMem   1 Linda.
 CoulSu:  requested number of processors reduced to:   2 ShMem   1 Linda.
 CoulSu:  requested number of processors reduced to:   1 ShMem   1 Linda.
 CoulSu:  requested number of processors reduced to:   2 ShMem   1 Linda.
 CoulSu:  requested number of processors reduced to:   1 ShMem   1 Linda.
 CoulSu:  requested number of processors reduced to:   2 ShMem   1 Linda.
 CoulSu:  requested number of processors reduced to:   1 ShMem   1 Linda.
 CoulSu:  requested number of processors reduced to:   2 ShMem   1 Linda.
 CoulSu:  requested number of processors reduced to:   1 ShMem   1 Linda.
 CoulSu:  requested number of processors reduced to:   2 ShMem   1 Linda.
 CoulSu:  requested number of processors reduced to:   1 ShMem   1 Linda.
 CoulSu:  requested number of processors reduced to:   2 ShMem   1 Linda.
 CoulSu:  requested number of processors reduced to:   1 ShMem   1 Linda.
***at which point I abandoned the job

Does anyone have any thoughts on this.  Is this specific to Amber?  Is this only an initial long step, followed by more reasonable iterations later on?  Is this calculation doomed?!




--
Dr. Soren N. Eustis
ETH – Zurich
Institute for Biogeochemistry and Pollutant Dynamics
Universitatstrasse 16
8092 Zurich

+41 44 632 93 48 (office)
+41 44 632 14 38 (fax)

soren++env.ethz.ch