[Paper Review] Alleviating Projection-Space Sensitivity in DFT+U via Renormalized U

Manjula Raman, Kenneth Park|arXiv (Cornell University)|Mar 7, 2026

Heusler alloys: electronic and magnetic properties0 citations

TL;DR

The paper shows that U_eff depends on the Hubbard projection space; renormalizing U_eff for each projection size stabilizes lattice, electronic, and magnetic properties in TiO2 and MnO2, reducing projection-space sensitivity.

ABSTRACT

Although the DFT+U method significantly improves the description of correlated electronic systems, its accuracy is known to depend strongly on the input parameters including local projection space used for the Hubbard correction. As a result, calculations performed with different projection sizes can yield quantitatively different -- and sometimes divergent -- results. In this work, we investigate the dependence of the effective Coulomb interaction $U_{\mathrm{eff}}$ on projection size using constrained DFT calculations for rutile TiO$_2$ and $β$-MnO$_2$. We find that as the projection size increases, the self-consistently calculated $U_{\mathrm{eff}}$ decreases significantly -- by as much as $33$\%. This trend is attributed to renormalization of the Coulomb interaction through orbital relaxation and enhanced screening. When $U_{\mathrm{eff}}$ values recalculated for each projection size are employed, the results for lattice parameters, electronic structure, and relative phase stability become consistent across different projection sizes. These findings can provide a practical route to alleviate projection-size dependence in DFT+U calculations.

Motivation & Objective

• Investigate how the effective Coulomb interaction U_eff depends on the size of the Hubbard projection space.
• Demonstrate that renormalizing U_eff for each projection size yields consistent structural and electronic properties.
• Apply the approach to rutile TiO2 and β-MnO2 to assess lattice, electronic, and magnetic outcomes.
• Compare fixed-U_eff and renormalized-U_eff schemes across projection sizes to establish practical guidelines.

Proposed method

• Perform DFT+U calculations with APW+lo (WIEN2k) in the APW+lo framework.
• Systematically vary the muffin-tin radius R_MT to change the local projection space.
• Compute U_eff self-consistently via constrained DDFT (cDFT) as the energy cost to transfer a localized d electron (Eq. 3).
• Compare results obtained with a fixed U_eff versus a renormalized U_eff recalculated for each projection size.
• Analyze lattice parameters, electronic structure (PDOS), crystal field splitting, and total energies for TiO2 and MnO2.
• Discuss how U_eff renormalization accounts for orbital relaxation and screening effects on projection-size sensitivity.

Experimental results

Research questions

• RQ1 How does U_eff vary as the Hubbard projection space (R_MT) is changed for TiO2 and MnO2?
• RQ2 Does using a projection-size–dependent (renormalized) U_eff remove or reduce inconsistencies in lattice parameters, band structure, and phase stability across projection spaces?
• RQ3 What are the impacts of fixed vs renormalized U_eff on crystal field splitting and Mn–O/Ti–O hybridization trends?
• RQ4 Can the AFM ground state be consistently preserved across projection sizes when using renormalized U_eff in MnO2 and TiO2?

Key findings

• U_eff decreases with increasing projection size, e.g., TiO2 from 4.5 eV to 3.0 eV as R_MT grows from 1.91 to 2.30 a_B (approx. 33% reduction).
• Using renormalized U_eff recalculated at each projection size yields lattice parameters that are largely invariant across projection spaces (differences ≤ ~0.006 Å in TiO2).
• Renormalized U_eff leads to more consistent crystal field splitting and PDOS trends across projection sizes, avoiding artificial shifts seen with fixed U_eff.
• For MnO2, U_eff drops from 5.6 eV to 4.2 eV when R_MT changes from 1.90 to 2.20 a_B, and AFM ground state is preserved under renormalized U_eff across R_MT variations (where fixed U_eff may predict FM at larger R_MT).
• The fixed-U_eff approach often exhibits stronger projection-size sensitivity in total energies and magnetic ordering, whereas the renormalized scheme stabilizes these properties.