Re: 回复： [CP2K:4205] SCF Convergence

[SeokYun Kim]

Thank you very much.

I really appreciate it.

I'll try to modify and rerun my job.


On Wed, Nov 21, 2012 at 10:36 PM, shoutian sun
<sunshou... at yahoo.com.cn>wrote:

> Hi SeokYun,
> I have some suggestions<http://cn.mg20.mail.yahoo.com/search?hl=zh-CN&safe=strict&tbo=d&biw=1397&bih=740&spell=1&q=suggestions&sa=X&ei=X5etUM2zKvGOiAfbpYGABA&ved=0CCsQBSgA>for your input file.
>
> 1. The conformer you used for optimization is very poor, even using atomic
> unit,
>     the C-C distance of the ring is about 3.33 anstrom (kinds be wrong) or
> 3.33 bohr
>     (about 1.76 angstrom, also a bit larger, isn't it?).  So, my
> suggestion<http://cn.mg20.mail.yahoo.com/search?hl=zh-CN&safe=strict&tbo=d&biw=1397&bih=740&spell=1&q=suggestions&sa=X&ei=X5etUM2zKvGOiAfbpYGABA&ved=0CCsQBSgA> is
> rebuilding
>     your conformer.
>
> 2. use the POISSON_SOLVER (I think your systerm is a cluster, right?)
>     &POISSON
>          PERIODIC none
>          POISSON_SOLVER wavelet
>     &END POISSON
> 3. use  MINIMIZER CG for OT sec. (If SCF doesn't converge, sometimes, it
> works)
>
> 4. XC_FUNCTIONAL Pade is not a good choice. (BLYP, I always used.)
>
> 5. sometime, SCF doesn't converge, you have to use smooth method in the
> &XC section
>     , as follows:
>       &XC_GRID
>          XC_SMOOTH_RHO NN10
>          XC_DERIV SPLINE2_SMOOTH
>       &END XC_GRID
>
> 6. cell parameter:      ABC 40.000 40.000 40.000
>     I think the default unit is angstrom, right?
>     40.0 angstrom for 35 atoms conformer is quite big.
>     And very time consuming, Don't you think so?
>     (PERIODIC none for clusters.)
>
> 7. The coordinates of your conformer. I think your comformer
>     is build in atomic unit. And you do not modify it and just put
>     it in the &COORD.....&END COORD section. Am I right?
>     However, the problem is the default unit of atom coordinates
>     for cp2k is angstrom, not bohr. So, please modify it, or add
>     key word: UNIT Bohr in the following sec.
>
>     &COORD
>         C    15.02103029        21.43572257     20.18000314
>         C    12.13797029        19.86029257     20.10217314
>             ..................
>        &END COORD
>
> 8. using POISSON_SOLVER wavelet, you have to use the following key words
>     to put the cluster in the center of the box.
>
>     &TOPOLOGY
>         &CENTER_COORDINATES
>         &END
>     &END
>
> I rebuild your comformer and do the optimization with my input file, it
> converge well.
> For detail, please refer to cp2k input reference.
>
> Hope, it works for you.
>
> Regards,
> sun
> Chemistry, CUHK, Hong Kong
>
>
>
>
>
> --
> You received this message because you are subscribed to the Google Groups
> "cp2k" group.
> To post to this group, send email to cp... at googlegroups.com.
> To unsubscribe from this group, send email to
> cp2k+uns... at googlegroups.com.
> For more options, visit this group at
> http://groups.google.com/group/cp2k?hl=en.
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20121122/66feb34c/attachment.htm>