From owner-chemistry -A_T- ccl.net Sun Nov 25 13:30:01 2007 From: "John McKelvey jmmckel:+:gmail.com" To: CCL Subject: CCL: error while i change basis set Message-Id: <-35716-071125130348-9109-evSruPNsywUXhpMfYYkt0w-*-server.ccl.net> X-Original-From: "John McKelvey" Content-Type: multipart/alternative; boundary="----=_Part_40046_526555.1196010481744" Date: Sun, 25 Nov 2007 12:08:01 -0500 MIME-Version: 1.0 Sent to CCL by: "John McKelvey" [jmmckel:_:gmail.com] ------=_Part_40046_526555.1196010481744 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Content-Disposition: inline Hello.. This is caused by collinearity in the overlap matrix. There are basis functions that are too similar, and when the overlap matrix is diagonalized there a re eigenvalues that are too close to being zero, and this leads to large values when the inverse square roots are taken of the eigenvalues in the process of computing the necessary S**(-1/2) matrix, which is then used in computing the MO's in the original non-orthogonal basis set. This problem can be minimized by using smaller basis sets. There are no magic formulas to predict this problem other than the lowest eigenvalue of the overlap matrix which is probably printed in the output. Perhaps start with 6-311G++ and add basis functions in steps and check on the lowest overlap eigenvalue. Cheers, John McKelvey On Nov 25, 2007 10:08 AM, Aleksandra n Rudnitskaya aleksandra.rudnit001]![ umb.edu wrote: > > Sent to CCL by: "Aleksandra n Rudnitskaya" [aleksandra.rudnit001*o*umb.edu > ] > Dear all, > I faced with another interesting problem. When I optimized structure on > b3lyp/6-31(d) everything was fine, but when I did optimization on > b3lyp/6-311++g(3df,3pd) there appeared error massage. > Initial input file is > %nproc=8 > %mem=3500MB > # ub3lyp/6-311++g(3df,3pd) opt=calcFc freq=noraman > > Title > > 0 2 > Molecular specification > Error massage at the end of the file is > > **** Warning!!: The largest alpha MO coefficient is 0.15386734D+03 > > > **** Warning!!: The largest beta MO coefficient is 0.14838413D+03 > > Symmetrizing basis deriv contribution to polar: > IMax=3 JMax=2 DiffMx= 0.00D+00 > G2DrvN: will do 17 centers at a time, making 1 passes doing > MaxLOS=3. > FoFDir/FoFCou used for L=0 through L=3. > DoAtom=TTTTTTTTTTTTTTTT > Differentiating once with respect to electric field. > with respect to dipole field. > Differentiating once with respect to nuclear coordinates. > Integrals replicated using symmetry in FoFDir. > MinBra= 0 MaxBra= 3 Meth= 1. > IRaf= 0 NMat= 51 IRICut= 51 DoRegI=T DoRafI=T ISym2E= 2 > JSym2E=2. > There are 51 degrees of freedom in the 1st order CPHF. > 48 vectors were produced by pass 0. > AX will form 48 AO Fock derivatives at one time. > > What does it mean and how to fix it? > > I tried to increase %mem=5500MB it didnt work. > > Thanks in advance. > > Aleksandra> > > > ------=_Part_40046_526555.1196010481744 Content-Type: text/html; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Content-Disposition: inline Hello..

This is caused by collinearity in the overlap matrix.  There are basis functions that are too similar, and when the overlap matrix is diagonalized there a re eigenvalues that are too close to being zero, and this leads to large values when the inverse square roots are taken of the eigenvalues in the process of computing the necessary S**(-1/2) matrix, which is then used in computing the MO's in the original non-orthogonal basis set.  This problem can be minimized by using smaller basis sets.  There are no magic formulas to predict this problem other than the lowest eigenvalue of the overlap matrix which is probably printed in the output.
Perhaps start with 6-311G++ and add basis functions in steps and check on the lowest overlap eigenvalue.

Cheers,

John McKelvey


On Nov 25, 2007 10:08 AM, Aleksandra n Rudnitskaya aleksandra.rudnit001]![umb.edu <owner-chemistry^ccl.net> wrote:

Sent to CCL by: "Aleksandra n Rudnitskaya" [aleksandra.rudnit001*o*umb.edu]
Dear all,
I faced with another interesting problem.  When I optimized structure on b3lyp/6-31(d) everything was fine, but when I did optimization on b3lyp/6-311++g(3df,3pd) there appeared error massage.
Initial input file is
%nproc=8
%mem=3500MB
# ub3lyp/6-311++g(3df,3pd) opt=calcFc freq=noraman

Title

0  2
Molecular specification
Error massage at the end of the file is

**** Warning!!: The largest alpha MO coefficient is   0.15386734D+03


 **** Warning!!: The largest beta MO coefficient is  0.14838413D+03

 Symmetrizing basis deriv contribution to polar:
 IMax=3 JMax=2 DiffMx= 0.00D+00
 G2DrvN: will do   17 centers at a time, making    1 passes doing MaxLOS=3.
 FoFDir/FoFCou used for L=0 through L=3.
 DoAtom=TTTTTTTTTTTTTTTT
         Differentiating once with respect to electric field.
               with respect to dipole field.
         Differentiating once with respect to nuclear coordinates.
 Integrals replicated using symmetry in FoFDir.
 MinBra= 0 MaxBra= 3 Meth= 1.
 IRaf=       0 NMat=  51 IRICut=      51 DoRegI=T DoRafI=T ISym2E= 2 JSym2E=2.
         There are  51 degrees of freedom in the 1st order CPHF.
   48 vectors were produced by pass  0.
 AX will form  48 AO Fock derivatives at one time.

What does it mean and how to fix it?

I tried to increase %mem=5500MB it didnt work.

Thanks in advance.

Aleksandra



E-mail to subscribers: CHEMISTRY^ccl.net or use:
     

E-mail to administrators: CHEMISTRY-REQUEST^ccl.net or use
     

Subscribe/Unsubscribe:
     http://www.ccl.net/chemistry/sub_unsub.shtml

Before posting, check wait time at: http://www.ccl.net

Job: http://www.ccl.net/jobs
Conferences: http://server.ccl.net/chemistry/announcements/conferences/

Search Messages: http://www.ccl.net/htdig  (login: ccl, Password: search)
     http://www.ccl.net/spammers.txt

RTFI: http://www.ccl.net/chemistry/aboutccl/instructions/




------=_Part_40046_526555.1196010481744--