Calculating hyperfine structure constants for first EXCITED state of a diatomic

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
ievutec
Posts: 5
Joined: 29 Aug 2018, 09:29
First name(s): Ieva
Last name(s): Cepaite
Affiliation: CQT
Country: Singapore

Calculating hyperfine structure constants for first EXCITED state of a diatomic

Post by ievutec » 02 Sep 2018, 04:55

Dear all,

Apologies for being a little inexperienced in quantum computational chemistry, but I need a little guidance to make sure I'm not just fumbling my way into any random computation and getting exactly what I need.

I want to calculate the hyperfine structure constants for the LiK molecule in its first excited singlet state (the A singlet sigma +). This includes getting the electric field gradients for both atoms as well as everything required for the constants defined as c1-c4 (spin-spin, spin-rotation etc.) I would also like the electronic g-factor if possible.

I understand enough to run these calculations for the ground-state (although perhaps not the output for the electric field gradient) but I am stumped as to how to re-define the calculation to do it for the A-singlet-sigma+.

Any help is greatly appreciated!

hjaaj
Posts: 282
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: Calculating hyperfine structure constants for first EXCITED state of a diatomic

Post by hjaaj » 02 Sep 2018, 13:44

The hyperfine structure constants and the electronic g-factor are all zero for singlet (S=0) states.

Anyway, you can calculate properties of excited states with MCSCF by specifying the desired state in the wave function input.

-- Hans Jørgen.

ievutec
Posts: 5
Joined: 29 Aug 2018, 09:29
First name(s): Ieva
Last name(s): Cepaite
Affiliation: CQT
Country: Singapore

Re: Calculating hyperfine structure constants for first EXCITED state of a diatomic

Post by ievutec » 03 Sep 2018, 05:45

hjaaj wrote:
02 Sep 2018, 13:44
The hyperfine structure constants and the electronic g-factor are all zero for singlet (S=0) states.

Anyway, you can calculate properties of excited states with MCSCF by specifying the desired state in the wave function input.

-- Hans Jørgen.
Thanks! Though I'm not sure we're talking about the same thing in this case. I'm interested in finding the coupling constants as in the Hamiltonian I've attached below.
Image
Image

Just for the sake of time and simplicity, what is the best way of defining the necessary wavefunctions in MCSCF with dalton? Say for the A-singlet-sigma or the b-triplet-pi.

hjaaj
Posts: 282
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: Calculating hyperfine structure constants for first EXCITED state of a diatomic

Post by hjaaj » 03 Sep 2018, 11:16

Aha (wrt what you want to calculate).

An MCSCF calculation will by default converge to the lowest state of the specified symmetry and spin multiplicity. To converge to e.g. the second state (as for A singlet SIGMA +) you need to add

Code: Select all

*OPTIMIZATION
.STATE
 2
For PI states one must select either PI_x or PI_y of the degenerate pair under .SYMMETRY, typically symmetry 2 or 3. If you do not want a symmetry broken solution (with pi_x orbitals different from pi_y orbitals) you must also specify .SUPSYM, which will force the pi_x and pi_y orbitals to be not symmetry broken.

-- Hans Jørgen.

ievutec
Posts: 5
Joined: 29 Aug 2018, 09:29
First name(s): Ieva
Last name(s): Cepaite
Affiliation: CQT
Country: Singapore

Re: Calculating hyperfine structure constants for first EXCITED state of a diatomic

Post by ievutec » 04 Sep 2018, 10:50

Alright brilliant! Although I wonder if MCSCF calculations are any good in terms of accuracy or calculating the properties I need. Say I wanted electric field gradients, nuclear shielding constants and the rotational g-factor. From what I've read previously, most of these are best determined using DFT rather than ab initio approaches? Or are they both equally valid?

I can't find much on TDDFT in the manual, is that an option in Dalton? If so, how would I approach excited states using TDDFT instead?

The properties I'd like to investigate include:

Code: Select all

**PROPERTIES 
.MAGNET
.QUADRU
.NQCC
.SHIELD
.MOLGFA
.SPIN-SPIN
.SPIN-R
I'm sorry for the barrage of questions, feel free to point me to any example files that might help, I'm just having trouble navigating the myriad of options and non-options to extract what might be best.

Thank you again!

hjaaj
Posts: 282
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: Calculating hyperfine structure constants for first EXCITED state of a diatomic

Post by hjaaj » 04 Sep 2018, 12:48

As your problem is essentially a two-electron problem I think you should be able to do quite OK with MCSCF, maybe a RASSCF with 2 electrons in the 8 valence orbitals in RAS2, some subvalence orbitals in RAS1 and correlating orbitals for those in RAS3. Some testing would be a good idea to find a suitable MCSCF model.

Yes, TDDFT is implemented in the **RESPONS module (called linear response) and quadratic and cubic response is also implemented for some functionals. However, even though quadratic and cubic response in principle can give you excited state properties, it will definitely not work for all your desired properties.

ievutec
Posts: 5
Joined: 29 Aug 2018, 09:29
First name(s): Ieva
Last name(s): Cepaite
Affiliation: CQT
Country: Singapore

Re: Calculating hyperfine structure constants for first EXCITED state of a diatomic

Post by ievutec » 06 Sep 2018, 03:17

That is amazing help, thank you again!

ievutec
Posts: 5
Joined: 29 Aug 2018, 09:29
First name(s): Ieva
Last name(s): Cepaite
Affiliation: CQT
Country: Singapore

Re: Calculating hyperfine structure constants for first EXCITED state of a diatomic

Post by ievutec » 06 Sep 2018, 13:25

And after that lovely thank you, I have one final question as a complete newbie: how would you recommend me specifying the INACTIVE ORBITALS and the CAS SPACE for the LiK molecule? I've tried a number of combinations but I'm completely unsure as to which one is best. Again, we're considering some of the lower excited states.

What is the general approach to these variables?

hjaaj
Posts: 282
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: Calculating hyperfine structure constants for first EXCITED state of a diatomic

Post by hjaaj » 09 Sep 2018, 09:35

I would first run a Hartree-Fock plus MP2 and look at the MP2 natural orbital occupation numbers to see how important it is to correlate some of the subvalence orbitals. The Hartree-Fock output also shows you how the occupied orbitals are distributed over symmetries, a little easier that doing the analysis yourself.

taylor
Posts: 525
Joined: 15 Oct 2013, 05:37
First name(s): Peter
Middle name(s): Robert
Last name(s): Taylor
Affiliation: Tianjin University
Country: China

Re: Calculating hyperfine structure constants for first EXCITED state of a diatomic

Post by taylor » 15 Sep 2018, 04:35

You also need to be aware that there are very considerable "core polarization" and core correlation effects in alkali-metal dimers, especially with heavier alkali metals. I am less optimistic that MCSCF will do well for these properties because such effects must be important near the nuclei. This is not an argument against trying, of course! But accounting for these effects is not easy and places heavy demands on the basis sets as well as the many-electron treatment.

By the way, I am of the opinion that any agreement between observation and (TD)DFT for these properties is simply sheer luck. The functionals are not parametrized for this sort of thing.

Best regards
Pete

Post Reply

Who is online

Users browsing this forum: No registered users and 6 guests