test	basis	scale	ecutwfc_ry	f_electrons	f_moment_ub	energy_ev_per_atom
isolated Gd atom	plane waves		100			-4837.5190821658
isolated Gd atom	plane waves		150			-4851.3848108726
isolated Gd atom	primitive jy 60 Ry		100			-4701.9605871044
isolated Gd atom	primitive jy 100 Ry		150			-4839.8543408625
isolated Gd atom	plane waves		200			-4851.4468655564
isolated Gd atom	primitive jy 120 Ry		150			-4848.2955078327
isolated Gd atom	primitive jy 150 Ry		200			-4851.1463589033
hcp Gd E(V)	plane waves	0.94	150			-4860.892472
hcp Gd E(V)	CSW-NAO 4s3p3d3f (shipped)	0.94	150	7.4781	7.0919	-4860.713754
hcp Gd E(V)	plane waves	0.97	150			-4860.989775
hcp Gd E(V)	CSW-NAO 4s3p3d3f (shipped)	0.97	150	7.4356	7.0973	-4860.80303
hcp Gd E(V)	plane waves	1	150			-4861.01847

Sign in to see the full dataset

Sep 29

Gd basis validation: plane waves vs atomic orbitals, same PP

Data Views Docs Logs

Overview

00

Attribution

Connections

1

Referenced by

Ouro DFT now holds gadolinium's 4f shell: what was wrong, and how we built a basis that fixes it
Rare-earth runs on Ouro DFT converged cleanly and were wrong: hcp Gd came out at 0.06 μB instead of 7.6. The pseudopotential was fine; the orbital basis had no room for 4f electrons. We generated a matched basis, validated it against plane waves, and Gd now gives 7.71 μB, 2.65 T, and the right ferrimagnetic order in GdCo₅.

Attribution

ARR license

Validation data for the 4f-in-valence Gd basis now served by Ouro DFT (ABACUS). Isolated-atom total energies vs plane-wave and primitive spherical-wave cutoffs, and hcp Gd E(V), moment and Mulliken 4f occupation for plane waves against three candidate bases. All with the jingslaw PseudoDojo FR Gd pseudopotential (4f in valence), PBE, collinear.

test	basis	scale	ecutwfc_ry	f_electrons	f_moment_ub	energy_ev_per_atom
isolated Gd atom	plane waves		100			-4837.5190821658
isolated Gd atom	plane waves		150			-4851.3848108726
isolated Gd atom	primitive jy 60 Ry		100			-4701.9605871044
isolated Gd atom	primitive jy 100 Ry		150			-4839.8543408625
isolated Gd atom	plane waves		200			-4851.4468655564
isolated Gd atom	primitive jy 120 Ry		150			-4848.2955078327
isolated Gd atom	primitive jy 150 Ry		200			-4851.1463589033
hcp Gd E(V)	plane waves	0.94	150			-4860.892472
hcp Gd E(V)	CSW-NAO 4s3p3d3f (shipped)	0.94	150	7.4781	7.0919	-4860.713754
hcp Gd E(V)	plane waves	0.97	150			-4860.989775
hcp Gd E(V)	CSW-NAO 4s3p3d3f (shipped)	0.97	150	7.4356	7.0973	-4860.80303
hcp Gd E(V)	plane waves	1	150			-4861.01847

Sign in to see the full dataset

0

Stats

Rows: 32
Columns: 9
Size: 32 KB
Downloads: 0

Join to comment

Connections

1

Referenced by

Ouro DFT now holds gadolinium's 4f shell: what was wrong, and how we built a basis that fixes it
Rare-earth runs on Ouro DFT converged cleanly and were wrong: hcp Gd came out at 0.06 μB instead of 7.6. The pseudopotential was fine; the orbital basis had no room for 4f electrons. We generated a matched basis, validated it against plane waves, and Gd now gives 7.71 μB, 2.65 T, and the right ferrimagnetic order in GdCo₅.