This is an old revision of the document!


CalculationRestrictions for 3d-TM systems

asked by Saki IMADA (2024/10/09 04:59)

Dear Developers,

I have a question about CalculationRestrictions for 3d-TM systems. I found a related question, 'XAS M4,5 partial excitations' (2021/09/21 18:20). It was for a 4f (Ce) system. In your answer,

NConfigurations = 3 CalculationRestrictions = {NF, NB, {'0000000000 00000000000000 11111111111111', 14 - (NConfigurations - 1), 14}}

My question: do I need to define the CalculationRestrictions for both 3d-TM and ligand in the case of a 3d-TM system?

For NiO in the input file '02_XAS_L23.Quanty', the “StartRestrictions” has two {} sets as follows,

StartRestrictions = {NF, NB, {“000000 1111111111 0000000000”,8,8}, {“111111 0000000000 1111111111”,16,16}}

I would like to know if I need to set the “CalculationRestrictions” on each {}.

I aim to understand how the XAS, PFY, and RIXS spectra depend on the number of configurations for 3d-TM oxides and nitrides in the ligand field calculations.

Thank you for your help.

Saki

Answers

, 2024/10/09 07:06, 2024/10/10 13:19

Dear Saki,

The function Eigensystem starts with a set of random states that fulfil the StartRestrictions $| \psi_i \rangle$. We then iteratively converge to the ground-state by looking at the states $H^n | \psi_i \rangle$. This is a Lanczos routine. The states $H^n | \psi_i \rangle$ can contain determinants that do not fulfil the StartRestrictions.

An example. For a $d^0$ compound the StartRestrictions could be

NF = 10 + 10
NB = 0
StartRestrictions = {NF, NB, {"1111111111 0000000000",0,0}, {"0000000000 1111111111",10,10}}

The state $| \psi_i \rangle$ then is the state $d^{0}L^{10}$. The state $| H \psi_i \rangle$ contains determinants that are part of the $d^{0}L^{10}$ configuration, but also of the $d^{1}L^{9}$ configuration. The state $| H^2 \psi_i \rangle$ contains determinants that are part of the $d^{0}L^{10}$ configuration, the $d^{1}L^{9}$ configuration and the $d^{2}L^{8}$ configuration. The state $| H^{10} \psi_i \rangle$ then contains determinants from the full Hilbertspace.

The restrictions given in the options of Eigenstystem restrict the set of allowed determinants in $| H^n \psi_i \rangle$, given the starting states $| \psi_i \rangle$ that fulfil the StartRestrictions.

In the case of the $d^0$ example the following 2 Restrictions are the same

Nconfig = 2
Nd      = 0
-- Nd + Nconfig must be <= 11
-- Nconfig must be > 0
CalculationRestrictions = {NF, NB, {"1111111111 0000000000",Nd,Nd+Nconfig-1}}
CalculationRestrictions = {NF, NB, {"0000000000 1111111111",10-Nconfig+1,10}}

For $Nconfig = 1$ you only look at states in the $d^N$ configuration. For $Nconfig=2$ you look for states that are a superposition of states in the $d^N$ and $d^{N+1}$ configuration etc. (Please make sure that $0<=Nd<=10$ and $0<=Nd+Nconfig-1<=10$ and $0<10-Nconfig+1<=10$. Quanty should give error messages if this is not the case, but we do not capture all errors.

You are allowed to apply both restrictions at the same time, that will not change anything. The electrons can only hop from the Ligand to the $d$-shell, so if you know the number of holes on the ligand you know the number of electrons on the $d$ shell.

Hope this helps, best wishes, Maurits

, 2024/10/10 12:18, 2024/10/11 01:45

Dear Maurits,

Thank you for your answer. I've already gotten the configuration dependence of XAS spectra for the 3d0 system with the commands!!

Best regards,

Saki

You could leave a comment if you were logged in.
Print/export