vibrational calculations

[Noam Bernstein]

Hi - I'm trying to do a vibrational calculation for a molecule using CP2K
and QS.  I've relaxed the geometry (LBFGS, RMS_FORCE 0.000001).
Forces do indeed seem to be that small in the output.  However, I keep
getting negative frequencies in the cp2k NORMAL_MODES output.
I've tried with the default DX, and with larger ones (up to 0.3 a0, in
case I was
getting noise in the forces).  I've increased the cutoff (up to 450),
decreased the
SCF tolerance (1e-8) turned off MT non-periodic solver (although I do actually
want a non-periodic calculation for the real answer, I just thought
I'd see if it
was causing problems), and CENTER_COORDINATES off, but keep on
getting negative frequencies.  I've attached the input and output.  It's 30
atoms, (20 A)^3, so I don't know that anyone wants to reproduce my results, but
if anyone has any suggestions for what I might be doing wrong, or what to check
next, I'd appreciate it.  My only guess is some subtle egg-crate effect, but
I don't know what parameter could cause that.

        thanks,
        Noam
-------------- next part --------------
A non-text attachment was scrubbed...
Name: cp2k_input.1mer_VIB.inp
Type: application/octet-stream
Size: 2310 bytes
Desc: not available
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20100614/37b9dffd/attachment.obj>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: 1mer_VIB.stdout.DX_0.30.new.no_MT
Type: application/octet-stream
Size: 768024 bytes
Desc: not available
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20100614/37b9dffd/attachment-0001.obj>