Unit-Cell Doubling Magnetism

• Unit-cell doubling magnetism is a phenomenon where the magnetic order extends beyond the basic crystallographic unit cell, typically driven by propagation vectors like (π,0) and (0,π).
• It manifests in various materials including cuprates, van der Waals heterostructures, and complex oxides, and is observed via techniques such as polarized neutron diffraction and moiré pattern analysis.
• Understanding this magnetism provides insights into pseudogap formation, Fermi surface reconstruction, and the interplay between electronic and structural instabilities in correlated systems.

Searching arXiv for recent and foundational papers relevant to unit-cell doubling magnetism. Unit-cell doubling magnetism denotes magnetic order whose periodicity exceeds that of the crystallographic unit cell, typically through propagation vectors such as $(\pi,0)$, $(0,\pi)$, or by superlattice modulations that require an enlarged magnetic cell. In the literature summarized here, the term spans several distinct but related regimes: translational-symmetry-breaking magnetism in cuprates, moiré-induced magnetic textures in van der Waals heterostructures, multi-sublattice ferrimagnetism in complex oxides, odd-parity antiferromagnetism in nonsymmorphic settings, and structural-electronic instabilities that produce doubled periodicity and couple to magnetism. A crucial contrast is with intra-unit-cell magnetism in cuprates, which is a $q=0$ order that breaks time-reversal symmetry without doubling the unit cell (Bourges et al., 2018). The resulting research landscape is therefore defined as much by distinctions between $q=0$ and finite-$q$ order as by the common motif of enlarged magnetic periodicity.

1. Conceptual scope and relation to intra-unit-cell order

In superconducting cuprates, intra-unit-cell magnetism refers to an order in which time-reversal symmetry is broken without net magnetization per unit cell and without unit-cell doubling. It is detected by polarized neutron diffraction at specific Bragg positions and is usually interpreted in terms of loop currents; the order is at $Q=0$, so unit-cell doubling is not required (Bourges et al., 2018). The same cuprate literature also provides a contrasting case of lattice-translation-breaking magnetism: in $\rm YBa_{2}Cu_{3}O_{6.6}$, polarized neutron diffraction revealed magnetic correlations with planar propagation vector $(\pi,0)\equiv(0,\pi)$, yielding a doubling or quadrupling of the magnetic unit cell (Bounoua et al., 2021).

This distinction is central. Intra-unit-cell magnetism preserves lattice translational symmetry and appears at Bragg peaks, whereas unit-cell doubling magnetism breaks lattice translational symmetry and gives magnetic peaks away from the Bragg positions (Bounoua et al., 2021). In the cuprate context, the two phenomena can coexist: the hidden magnetic texture reported in $\rm YBa_{2}Cu_{3}O_{6.6}$ combines previously reported IUC magnetism with short-range $(\pi,0)$ or $(0,\pi)$0 magnetism, producing a magnetic texture of the $(0,\pi)$1 unit cells that forms large supercells (Bounoua et al., 2021).

Outside cuprates, the phrase also applies to situations where the magnetic ground state requires a larger magnetic unit cell because of symmetry-inequivalent sites, stacking-dependent exchange, or intrinsic crystallographic doubling. In moiré heterobilayers, enlarged moiré unit cells host AFM, FM, and ferromagnetic-chain domains; in double-double perovskites, multiple inequivalent transition-metal sites require a five-sublattice magnetic description; in odd-parity antiferromagnets, commensurate unit-cell doubling is tied to effective time-reversal symmetry and odd-parity spin polarization (Keskiner et al., 2024, Dhawan et al., 2022, Lee et al., 8 Aug 2025).

2. Cuprates: from $(0,\pi)$2 intra-unit-cell magnetism to $(0,\pi)$3 doubling

Polarized neutron diffraction established IUC magnetism in four cuprate families: YBa₂Cu₃O₆₊ₓ, HgBa₂CuO₄+δ, La₂₋ₓSrₓCuO₄, and Bi₂Sr₂CaCu₂O₈+δ. The onset temperature closely matches the pseudogap temperature $(0,\pi)$4 and correlates with hole doping, while the signal is observed at specific Bragg peaks such as $(0,\pi)$5 and $(0,\pi)$6 but not at $(0,\pi)$7, consistent with the expected form factor of loop currents (Bourges et al., 2018). The measured observable is the normalized spin-flip intensity,

$(0,\pi)$8

and the magnetic signal is a small increase of $(0,\pi)$9 of order $q=0$0–$q=0$1 at relevant Bragg peaks (Bourges et al., 2018).

The technical assessment by Bourges et al. addressed a controversy over null results in lower-statistics neutron measurements. Their central conclusion was that all reported polarized neutron diffraction experiments in superconducting cuprates are compatible with the existence of IUC magnetism once polarization inhomogeneities, temperature drift in the bare inverse flipping ratio, and sample-size limitations are treated self-consistently (Bourges et al., 2018). In that framework, unit-cell doubling is explicitly not a requirement for the pseudogap magnetic order.

A different magnetic component was later reported in $q=0$2: elastic polarized neutron diffraction with XYZ polarization analysis observed magnetic peaks at $q=0$3 and $q=0$4 in reciprocal lattice units, corresponding to $q=0$5 and $q=0$6 (Bounoua et al., 2021). These vectors imply a doubling in the uniaxial case or a biaxial $q=0$7 quadrupling of the magnetic unit cell. The intensity appears at the pseudogap onset temperature $q=0$8 K, and the out-of-plane component dominates, with moments mainly oriented perpendicular to the $q=0$9 planes (Bounoua et al., 2021).

The correlations are short-ranged. The intrinsic linewidth $q=0$0–$q=0$1 r.l.u. yields an in-plane correlation length

$q=0$2

corresponding to clusters of approximately $q=0$3–$q=0$4 unit cells, while along $q=0$5 the correlation length is $q=0$6 Å, or about one unit cell (Bounoua et al., 2021). This short-range LT-breaking magnetism is therefore distinct from both the parent antiferromagnet at $q=0$7 and the IUC $q=0$8 order. The authors proposed magnetic structure models including a $q=0$9 loop-current supercell carrying an anapole and a diagonal loop-current pattern akin to a rotated $q$0 configuration, and argued that such LT-breaking magnetism can reconstruct the Fermi surface and provide a mechanism for opening a pseudogap (Bounoua et al., 2021).

3. Microscopic motifs and representative material classes

Several microscopic routes to enlarged magnetic periodicity emerge across the cited works. In twisted Mott insulator–semimetal heterobilayers, the starting point is an extended Kondo lattice Hamiltonian with stacking-dependent, non-local Kondo coupling,

$q$1

with $q$2 for $q$3 (Keskiner et al., 2024). In the perturbative regime, stacking-dependent RKKY interactions generate AFM order in AA regions, FM order in AB/BA regions, and ferromagnetic chains coupled antiferromagnetically in saddle-point regions. The moiré pattern thereby produces magnetic domains within the enlarged moiré unit cell, and the structure factor has peaks at $q$4 for FM, $q$5 for AFM, and $q$6 for the chain state, directly reflecting enlarged or doubled periodicity (Keskiner et al., 2024).

In the double-double perovskite CaMnCrSbO₆, unit-cell doubling is tied to crystallographically non-equivalent Mn sites. The tetragonal unit cell contains four formula units and two inequivalent Mn sites, Mn(1) in tetrahedral coordination and Mn(2) in square-planar coordination, while Cr occupies a distorted octahedral site (Dhawan et al., 2022). The exchange network is not that of a simple two-sublattice ferrimagnet: Mn(1)-O-Mn(2), Mn(1)-O-Cr, and Mn(2)-O-Cr superexchange are antiferromagnetic, whereas Cr-O-O-Cr super-superexchange is ferromagnetic (Dhawan et al., 2022). The magnetic unit cell must therefore include four Mn sublattices plus one Cr sublattice. Mean-field calculations found that a five-sublattice model yields $q$7 K, much closer to the experimental value $q$8 K than the two-sublattice estimate of about $q$9 K (Dhawan et al., 2022).

In iron pnictides, unit-cell doubling can be intrinsic rather than interaction-generated. The As sublattice doubles the Fe square-lattice unit cell, so the correct low-energy description contains two Fe atoms per cell. Within the $Q=0$0 effective model, the $Q=0$1 or $Q=0$2 collinear antiferromagnetic order does not simply couple $Q=0$3 and $Q=0$4; it mixes the set $Q=0$5 and