Re: Calculation of the electronegativity of a periodic system ( No.1 ) 
 Date: 2023/05/30 14:13
 Name: T. Ozaki
 Hi,
Do you mean by the electronegativity, A, the following definition? A = E(N) − E(N+1)
If so, in the GGA and LDA functionals the quantity for bulk must be equivalent to the conduction band minimum (CBM) given by the KS eigenvalue.
The way you metioned: "only consider charge addition and subtraction by scf.system.charge and spin polarization" does not work, since an additonanal energy term is included by the interaction with the compesation background charge.
The method: "the deltaSCF and exact Coulomb cutoff methods" also does not work, since the method is valid for a localized state. An electron added to a bulk tends to be delocalized, and the situation violates the condition that "the deltaSCF and exact Coulomb cutoff methods" is valid. Even if we apply the method to a super cell being large enough, we will see that E(N+1)  E(N) is equivalent to the CBM.
The issue is closely related to the underestimation of the fundamental gap by LDA and GGA. Note that the fundamental gap is given by Egap = I  A, where I = E(N−1) − E(N), A = E(N) − E(N+1).
Regards,
TO

Re: Calculation of the electronegativity of a periodic system ( No.2 ) 
 Date: 2023/05/30 14:29
 Name: Masanobu Miyata <mmiyata@jaist.ac.jp>
 Dear Ozakisensei,
Thank you very much for your prompt reply. Sorry.
The definition of electronegativity was ambiguous. The electronegativity I am talking about is Mulliken's electronegativity (average value of electron affinity and ionization energy).
First ionization energy: E(N+1)  E(N) Electron affinity: E(N)  E(N1) E(N): Nuetral, scf.SpinPolarization off E(N+1): scf.system.charge = 1, scf.SpinPolarization on E(N1): scf.system.charge = 1, scf.SpinPolarization on
Then, the electronegativity of Mulliken is as follows. {E(N+1)E(N1)}/2
In this case, is the guarantee of the background charge still not cancelled?
Best regards, Miyata

Re: Calculation of the electronegativity of a periodic system ( No.3 ) 
 Date: 2023/05/31 11:24
 Name: T. Ozaki
 Hi,
Even if the calculations of E(N+1) and E(N1) are properly performed, we must obtain that (E(N+1)  E(N) + E(N)  E(N1))/2 = (ep_VBM  ep_CBM)/2 in LDA and GGA, where ep_VBM and ep_CBM are the KohnSham eigenvalues for the VBM and CBM, respectively.
Regards,
TO

Re: Calculation of the electronegativity of a periodic system ( No.4 ) 
 Date: 2023/05/31 17:37
 Name: Masanobu Miyata <mmiyata@jaist.ac.jp>
 Dear Ozakisensei,
Thank you for your prompt reply.
Best regards, Miyata

