# Role of multiorbital effects in the magnetic phase diagram of iron-pnictides

## TL;DR

This paper demonstrates that multiorbital bandstructure effects are crucial in determining the magnetic order in iron-pnictides, explaining experimental observations and highlighting the role of orbital content and Fermi surface features.

## Contribution

The study introduces a multiorbital itinerant Landau approach with realistic bandstructures to accurately predict magnetic phases in iron-pnictides, emphasizing the importance of orbital effects.

## Key findings

- Realistic multiorbital models reproduce the experimentally observed magnetic phase.
- Presence of a hole pocket at the M-point favors magnetic stripe order.
- Leading magnetic instabilities belong to the {A_{1g}, B_{1g}} irreducible representations.

## Abstract

We elucidate the pivotal role of the bandstructure's orbital content in deciding the type of commensurate magnetic order stabilized within the itinerant scenario of iron-pnictides. Recent experimental findings in the tetragonal magnetic phase attest to the existence of the so-called charge and spin ordered density wave over the spin-vortex crystal phase, the latter of which tends to be favored in simplified band models of itinerant magnetism. Here we show that employing a multiorbital itinerant Landau approach based on realistic bandstructures can account for the experimentally observed magnetic phase, and thus shed light on the importance of the orbital content in deciding the magnetic order. In addition, we remark that the presence of a hole pocket centered at the Brillouin zone's ${\rm M}$-point favors a magnetic stripe rather than a tetragonal magnetic phase. For inferring the symmetry properties of the different magnetic phases, we formulate our theory in terms of magnetic order parameters transforming according to irreducible representations of the ensuing D$_{\rm 4h}$ point group. The latter method not only provides transparent understanding of the symmetry breaking schemes but also reveals that the leading instabilities always belong to the $\{A_{1g},B_{1g}\}$ subset of irreducible representations, independent of their C$_2$ or C$_4$ nature.