CCL:G: TD-DFT opt failing in G09

I have consulted this question with Gaussian Technique Support, and following are their answers. Please notice the last paragraph.
Also, you can try use the last structure to restart the optimization or add the direct options to TDDFT.

In the case of "No map to state **, you need to resolve more vectors" messages, this is usually an indication that one did not include enough excited states in the TD or CIS calculation. The "States=N" option to the "TD" or "CIS" keywords tells how many excited states to include in an excited state energy calculation. If this is not specified, the default value will be "States=3". The recommended value is to include a minimum of 2 or 3 more states than the state of interest. Thus, if you want to perform a geometry optimization for excited state 5, for example, I would recommend at least using "States=7" or "States=8". The geometry optimization will be done for one excited state M, selected with "Root=M", and one has to make sure that enough states are included in the CIS or TD expansion by having "States=N" where N is larger than M.

It is possible that, at some point during an optimization of an excited state, the order of the excited states changes and the CIS or TD expansion might need to include more states in order to be able to follow correctly the state of interest. This is essentially what that message about including more vectors mean, that is that at that point, the number of states that you originally specified with "States=N" was not enough in order to solve for the state of interest, so a larger number "N" will need to be used for "States=N".

Other times, the problem is that the ground state wavefunction becomes unstable, that is one of the states that was an excited at the initial geometry now becomes lower in energy than the state that was the ground state at the initial geometry. This kind of situation, unfortunately, cannot be modeled properly with single determinant expansions such as CIS or TD, and one would need to use CAS in order to be able to deal with the conical intersection or avoided crossing of states.

Another thing to note is that one should be much more careful with geometry optimizations on excited states than for the ground state. Typically the energy differences among excited states are smaller than between the ground state and the first excited state. Thus, one can afford to perform larger geometry optimization steps when optimizing the ground state than in the case of optimizing an excited state.

A "bad" geometry optimization step in the optimization of the ground state, may take you a bit off track but in following steps the optimization might find the way back and approach the converged structure. In the case of an optimization of an excited state, a "bad" geometry optimization step will also take you off track but, since other electronic states are close in energy, it is possible that at the new geometry the order of the excited states change and now the geometry optimization follows a different electronic state.

This is not only a problem because the optimization could be pursuing a different state than the one you were interested in, but also because, if several of these changes occur during a geometry optimization, it may even be hopeless to continue with the optimization because the gradient information and the estimated hessian could be useless (since not all the previous points in the geometry optimization where points from the same potential energy surface).

As a first measure to increase the reliability of the geometry optimization of excited states, I recommend to reduce the maximum allowed step size during geometry optimizations. Try "Opt=(MaxStep=10)" to set this value to 0.10 Bohr, or a smaller value if you still have problems. The default value is typically 0.30 Bohr. Reducing the maximum allowed step size will result in the geometry optimization taking more steps to reach convergence than with the default value. This will be true obviously for well-behaved geometry optimizations, but for problematic cases it will be the other way around, i.e. it will take fewer steps (and may even be impossible with the default step size) because it will be easier for the optimizer to follow a particular electronic state if the changes from step to step are not very drastic.

2010/4/1 Jamin Krinsky jamink_-_berkeley.edu <owner-chemistry_-_ccl.net>

Sent to CCL by: Jamin Krinsky [jamink(~)berkeley.edu]
Dear forum,

I have a user who is getting a mysterious failure message while
attempting TD-DFT optimizations in G09. It's related to link 914 but
it doesn't make sense. Here is the route section:

#p opt td=(singlets,nstates=6,root=1) rb3lyp/6-31+g(d) nosymm
int=ultrafine scf(xqc,maxconventionalcycles=60)

The calculation runs for 6 geometry steps but the first excited state
energy is quite oscillatory. At the 7th step, it quits with the
following error:

No map to state      1
You need to solve for more vectors in order to follow this state.
Error termination via Lnk1e in /usr/software/gaussian/g09.revA02/l914.exe

If he's following the 1st excited state then he shouldn't need more
states (increasing "nstates" to 20 does not help). I've never seen
this although my experience with this algorithm is minimal, and
although that error has come up a couple of times in these threads I
haven't seen a conclusive explanation. Any help with this would be
appreciated.

Regards,
Jamin

--
Jamin L Krinsky, Ph.D.
Molecular Graphics and Computation Facility
175 Tan Hall, University of California, Berkeley, CA 94720
jamink~~berkeley.edu, 510-643-0616
http://glab.cchem.berkeley.edu



-= This is automatically added to each message by the mailing script =-
E-mail to subscribers: CHEMISTRY_-_ccl.net or use:
     http://www.ccl.net/cgi-bin/ccl/send_ccl_message

E-mail to administrators: CHEMISTRY-REQUEST_-_ccl.net or use
     http://www.ccl.net/cgi-bin/ccl/send_ccl_message

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/chemistry/searchccl/index.shtml
     http://www.ccl.net/spammers.txt

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





--
========================================
Panwang Zhou
State Key Laboratory of Molecular Reaction Dynamics
Dalian Institute of Chemical Physics
Chinese Academy of Sciences.
Tel: 0411-84379195 Fax: 0411-84675584
========================================