Competing collinear magnetic structures in superconducting FeSe by first-principles quantum Monte Carlo calculations

Brian Busemeyer, Mario Dagrada, Sandro Sorella, Michele Casula, Lucas K. Wagner

Research output: Contribution to journalArticlepeer-review


Resolving the interplay between magnetic interactions and structural properties in strongly correlated materials through a quantitatively accurate approach has been a major challenge in condensed-matter physics. Here we apply highly accurate first-principles quantum Monte Carlo (QMC) techniques to obtain structural and magnetic properties of the iron selenide (FeSe) superconductor under pressure. Where comparable, the computed properties are very close to the experimental values. Of potential ordered magnetic configurations, collinear spin configurations are the most energetically favorable over the explored pressure range. They become nearly degenerate in energy with bicollinear spin orderings at around 7 GPa, when the experimental critical temperature Tc is the highest. On the other hand, ferromagnetic, checkerboard, and staggered dimer configurations become relatively higher in energy as the pressure increases. The behavior under pressure is explained by an analysis of the local charge compressibility and the orbital occupation as described by the QMC many-body wave function, which reveals how spin, charge, and orbital degrees of freedom are strongly coupled in this compound. This remarkable pressure evolution suggests that stripelike magnetic fluctuations may be responsible for the enhanced Tc in FeSe and that higher Tc is associated with nearness to a crossover between collinear and bicollinear ordering.

Original languageEnglish (US)
Article number035108
JournalPhysical Review B
Issue number3
StatePublished - Jul 5 2016

ASJC Scopus subject areas

  • Electronic, Optical and Magnetic Materials
  • Condensed Matter Physics


Dive into the research topics of 'Competing collinear magnetic structures in superconducting FeSe by first-principles quantum Monte Carlo calculations'. Together they form a unique fingerprint.

Cite this