Fortran runtime error: Bad integer for item 5 in list input

Find answers or ask questions regarding Dalton calculations.
Please upload an output file showing the problem, if applicable.
(It is not necessary to upload input files, they can be found in the output file.)

Post Reply
kaushikhatua
Posts: 35
Joined: 10 Jan 2014, 07:05
First name(s): kaushik
Last name(s): hatua
Affiliation: IIEST India
Country: India

Fortran runtime error: Bad integer for item 5 in list input

Post by kaushikhatua » 22 Mar 2014, 06:16

I am calculating excitation energy between pair of states (present case it is first three states) by response module. Here is the input. The Molecule is linear (acetylene)

**DALTON INPUT

.RUN RESPONSE

.DIRECT

**WAVE FUNCTIONS

.DFT

CAMB3LYP

**RESPONSE

*QUADRATIC

.DIPLEN

.DOUBLE RESIDUE

.ROOTS

3

**END OF DALTON INPUT

I get the following error message
At line 1468 of file /home/pkn/DALTON-2013.0-Source/DALTON/rsp/rspmai.F (unit = 11, file = 'DALTON.INP')
Fortran runtime error: Bad integer for item 2 in list input
However when .DOUBLE RESIDUE was not given job runs fine. I guess i cant underestand .ROOTS
Any help will be appreciated

lyzhao
Posts: 74
Joined: 11 Nov 2013, 00:36
First name(s): Youzhao
Last name(s): Lan
Affiliation: Institute of Physical Chemistry
Country: China

Re: Fortran runtime error: Bad integer for item 5 in list in

Post by lyzhao » 22 Mar 2014, 07:14

acetylene has a D2h(Dinfh) symmetry.
you should define the .ROOTS by
.ROOTS
1 1 1 1 1 1 1 1 for eight irreducible representations, respectively.
Of course, the integer number can be larger than 1, and 0 is also allowed.
If you want the first three excited states . I recommmand the input of 2 2 2 2 2 2 2 2 for such a small system.
3 3 3 3 3 3 3 3 would be a best choice. If you are proficient in symmetry, you can only define the symmetry you want.
More infos, please refer to the manual and test files in the Dalton source distribution.
Best.
Lan

kaushikhatua
Posts: 35
Joined: 10 Jan 2014, 07:05
First name(s): kaushik
Last name(s): hatua
Affiliation: IIEST India
Country: India

Re: Fortran runtime error: Bad integer for item 5 in list in

Post by kaushikhatua » 22 Mar 2014, 18:11

thanks it solved the problem. but I would like ask some times .roots has been specified by no of excited only say 10 or 50.

hjaaj
Posts: 338
Joined: 27 Jun 2013, 18:44
First name(s): Hans Jørgen
Middle name(s): Aagaard
Last name(s): Jensen
Affiliation: Universith of Southern Denmark
Country: Denmark

Re: Fortran runtime error: Bad integer for item 5 in list in

Post by hjaaj » 22 Mar 2014, 22:00

Would be nice, however, when you calculate with symmetry, the calculation of excitations is in fact a separate calculation for each symmetry. And even Dalton ;) cannot know beforehand how e.g. the lowest 10 excitations will be distributed among the symmetries, so it is up to the user to specify the number of excitations you want in each symmetry.

You can only just ask for the 10 lowest excitations without regard to distribution over symmetries by running the calculation in C1 -- that is with "NoSymmetry". It will be slower to do so, but you will get the exactly the lowest 10 roots. (However, note that you then will have a harder job to determine the symmetries of the excited states, because Dalton cannot help you with that if you run with "NoSymmetry").

taylor
Posts: 546
Joined: 15 Oct 2013, 05:37
First name(s): Peter
Middle name(s): Robert
Last name(s): Taylor
Affiliation: Tianjin University
Country: China

Re: Fortran runtime error: Bad integer for item 5 in list in

Post by taylor » 22 Mar 2014, 23:49

In addition to Hans Joergen's comments, can I repeat here a couple of things I pointed out some time ago on the old Dalton mailing list?

First, as you seek more excited states it becomes more and more unlikely that they will be dominated by a single excitation from the ground state. The more multiconfigurational the excited-state wave function, the less well it will be described by a response approach based on a single configuration, which is essentially what DFT response does. As a consequence states that are multiconfigurational will have incorrect excitation energies (the classic example is the lowest 1Ag state of s-tetrazine, which is predicted to lie around 10 eV above the ground state by DFT response and in reality is at less than 5 eV: the excited-state wave function is totally dominated by a double excitation from the ground state). It is important to realize that this is not a question of which functional to use --- no method that is not multiconfigurational can handle this situation. Coupled-cluster or multiconfiguration-based perturbation theory would be necessary.

Second, as you seek more excited states, at least for neutral systems, states that have some Rydberg character will appear more and more frequently. To describe these reliably will need basis sets that have very diffuse functions, or the addition of specialized Rydberg basis sets. It should be obvious that if the basis set lacks these functions, Rydberg or mixed valence/Rydberg states will either not appear at all, or will not appear at anything like the right energies.

Hence one should be rather cautious about trying to seek too many excited states. I doubt that for almost any molecule the lowest ten excited states would all be valence-like and dominated by a single excitation: taking 10 roots out in DFT response might give reasonable results if you are very lucky, but I would be surprised... Taking more than 10 out is probably a waste of time without using a method that can describe multiconfigurational excited states well, and a basis set that can handle Rydberg character.

Best regards
Pete

Post Reply

Who is online

Users browsing this forum: Majestic-12 [Bot] and 7 guests