Sep 29

Nd and Sm basis validation: plane waves vs atomic orbitals, same PP

Data Views Docs Logs

Overview

22

Attribution

Connections

1

Referenced by

Nd and Sm on Ouro DFT: matched 4f bases, SmCo₅ gets the right spin order, and why +U isn't ready yet
Neodymium and samarium now have validated 4f-in-valence bases on Ouro DFT. SmCo₅ converges to the Hund's-rule state, with Sm spin against Co and 0.71 eV/f.u. lower, once the SCF is damped. DFT+U on an open 4f shell still wanders between states; this post covers what occupation-matrix control fixed and what it didn't.

Attribution

ARR license

E(V), moment and Mulliken 4f occupation on 2-atom hcp models of Nd and Sm at 150 Ry, plane waves against the shipped CSW-NAO 4s3p3d3f bases, all with the jingslaw PseudoDojo FR pseudopotentials (4f in valence), PBE, collinear FM. The Sm plane-wave point at 1.09 failed and is omitted. For Nd, PW and LCAO land in different near-degenerate 4f^3 states at 1.06-1.09; compare 0.94-1.03.

2

Stats

Rows: 27
Columns: 7
Size: 32 KB
Downloads: 0

Join to comment

Connections

1

Referenced by

Nd and Sm on Ouro DFT: matched 4f bases, SmCo₅ gets the right spin order, and why +U isn't ready yet
Neodymium and samarium now have validated 4f-in-valence bases on Ouro DFT. SmCo₅ converges to the Hund's-rule state, with Sm spin against Co and 0.71 eV/f.u. lower, once the SCF is damped. DFT+U on an open 4f shell still wanders between states; this post covers what occupation-matrix control fixed and what it didn't.