Haerter-Shastry Antiferromagnet

• Haerter-Shastry antiferromagnet is a kinetically-induced 120° antiferromagnetic state in the triangular Hubbard model, emerging from interference in hole hopping.
• The mechanism leverages the triangular lattice geometry where kinetic frustration stabilizes the magnetic order even at finite hole doping.
• Numerical diagnostics using DMRG reveal phase regimes from pure HS order to stripe and paramagnetic metal, demonstrating coexistence of robust magnetism and metallicity.

Searching arXiv for papers on the Haerter–Shastry antiferromagnet and related triangular Hubbard kinetic magnetism. The Haerter–Shastry antiferromagnet denotes a form of antiferromagnetic order generated by kinetic frustration in the triangular Hubbard model, rather than by explicit exchange interactions. In the formulation summarized in "Haerter-Shastry kinetic magnetism and metallicity in the triangular Hubbard model" (Sherif et al., 21 Oct 2025), it arises in the infinite-$U$ triangular Hubbard model at finite hole density, where a $120^{\circ}$ antiferromagnetic state is stabilized by the hopping term alone. The same work studies its stability away from the single-hole limit, identifies an intermediate multimer-stripe regime, and reports evidence of gapless charge excitations throughout the finite-doping phase diagram (Sherif et al., 21 Oct 2025).

1. Definition and conceptual scope

The term refers to the kinetically induced $120^{\circ}$ antiferromagnetic state of the triangular Hubbard model in the $U\to\infty$ limit. The underlying Hamiltonian is the single-band Hubbard model on the triangular lattice,

$$H_{\rm Hubb}\;=\;-\,t\sum_{\langle i,j\rangle,\sigma}\bigl(c_{i\sigma}^\dagger c_{j\sigma}+\mathrm{h.c.}\bigr)\;+\;U\sum_i n_{i\uparrow}n_{i\downarrow}\;-\;\mu\sum_i n_i\,,$$

with $t>0$ the nearest-neighbor hopping amplitude, $U$ the on-site repulsion, and $\mu$ the chemical potential (Sherif et al., 21 Oct 2025). In the infinite-$U$ limit, double occupancy is forbidden by the projector $\mathcal P=\prod_i(1-n_{i\uparrow}n_{i\downarrow})$, yielding

$120^{\circ}$0

At large but finite $120^{\circ}$1, the same setting may be expressed through the $120^{\circ}$2–$120^{\circ}$3 expansion with $120^{\circ}$4 (Sherif et al., 21 Oct 2025).

Within this framework, the Haerter–Shastry mechanism is explicitly distinguished from superexchange-driven magnetism. In a non-frustrated bipartite lattice, a single hole promotes ferromagnetism through Nagaoka’s theorem, whereas on the triangular lattice a single hole in an otherwise half-filled, $120^{\circ}$5 system gains maximal kinetic energy when the background spins form a $120^{\circ}$6 Néel pattern (Sherif et al., 21 Oct 2025). The antiferromagnetic selection is therefore kinetic in origin and occurs in the absence of any explicit $120^{\circ}$7-exchange (Sherif et al., 21 Oct 2025).

This terminology should be separated from the Shastry–Sutherland antiferromagnet, which conventionally refers to Heisenberg or Hubbard models on the Shastry–Sutherland lattice. The literature in the supplied corpus includes several such Shastry–Sutherland studies [(Nakano et al., 2018); (Xi et al., 2021); (Chen et al., 26 Mar 2025); (Liu et al., 2014); (Janson et al., 2011)], but the Haerter–Shastry antiferromagnet itself is defined in the triangular-lattice Hubbard setting (Sherif et al., 21 Oct 2025).

2. Kinetic-frustration mechanism

The mechanism was formulated in the context of a single hole moving on the triangular lattice. According to the summary in (Sherif et al., 21 Oct 2025), a hole can hop around a triangle through two different two-step paths whose amplitudes interfere. In a ferromagnetic background, these two paths acquire a relative sign leading to destructive interference and a suppression of the hopping energy gain. If two of the three spins are antiparallel, the two paths no longer interfere destructively, the hole delocalizes more effectively, and the energy is lowered (Sherif et al., 21 Oct 2025).

This is the central content of kinetic frustration: the geometry of the triangular lattice, together with the sign structure of hole motion under the no-double-occupancy constraint, selects antiferromagnetic correlations. The resulting order is the three-sublattice $120^{\circ}$8 state, with sublattice spin directions parameterized as

$120^{\circ}$9

in the classical description quoted in (Sherif et al., 21 Oct 2025).

A plausible implication is that the Haerter–Shastry mechanism is best understood as a non-bipartite analogue of kinetic magnetism in which lattice frustration reverses the usual Nagaoka tendency. That inference is consistent with the explicit contrast drawn in (Sherif et al., 21 Oct 2025) between ferromagnetic selection on bipartite lattices and $120^{\circ}$0 antiferromagnetic selection on the triangular lattice.

3. Order parameter and numerical diagnostics

The long-range $120^{\circ}$1 state is diagnosed through the static spin structure factor

$120^{\circ}$2

which exhibits peaks at the Brillouin-zone $120^{\circ}$3 points $120^{\circ}$4 and symmetry-related wavevectors (Sherif et al., 21 Oct 2025). A finite-size antiferromagnetic order parameter is defined as

$120^{\circ}$5

with the extrapolation

$120^{\circ}$6

For a perfect classical $120^{\circ}$7 state, $120^{\circ}$8 (Sherif et al., 21 Oct 2025).

The cited study used the density matrix renormalization group on cylinders up to width $120^{\circ}$9 and length $U\to\infty$0 with “XC” boundary condition, together with additional $U\to\infty$1 and $U\to\infty$2 parallelogram geometries. The bond dimension reached $U\to\infty$3 with truncation error $U\to\infty$4, and calculations were performed in the $U\to\infty$5 sector for even holes and $U\to\infty$6 for odd holes (Sherif et al., 21 Oct 2025). Two bulk-averaging procedures were used for the spin structure factor: removing boundary rings and using a single reference-site correlator (Sherif et al., 21 Oct 2025).

These diagnostics matter because the Haerter–Shastry regime is not defined solely by local correlations. In the reported phase analysis, it is identified by the dominance of $U\to\infty$7 over $U\to\infty$8 at any other momentum at low hole density (Sherif et al., 21 Oct 2025).

4. Phase structure at finite hole density

The principal extension beyond the original single-hole mechanism is the finite-doping phase diagram of the infinite-$U\to\infty$9 triangular Hubbard model. By tracking the bulk $$H_{\rm Hubb}\;=\;-\,t\sum_{\langle i,j\rangle,\sigma}\bigl(c_{i\sigma}^\dagger c_{j\sigma}+\mathrm{h.c.}\bigr)\;+\;U\sum_i n_{i\uparrow}n_{i\downarrow}\;-\;\mu\sum_i n_i\,,$$0 as a function of hole density $$H_{\rm Hubb}\;=\;-\,t\sum_{\langle i,j\rangle,\sigma}\bigl(c_{i\sigma}^\dagger c_{j\sigma}+\mathrm{h.c.}\bigr)\;+\;U\sum_i n_{i\uparrow}n_{i\downarrow}\;-\;\mu\sum_i n_i\,,$$1, three regimes are reported (Sherif et al., 21 Oct 2025).

Regime	Hole-density range	Reported characterization
Pure HS $$H_{\rm Hubb}\;=\;-\,t\sum_{\langle i,j\rangle,\sigma}\bigl(c_{i\sigma}^\dagger c_{j\sigma}+\mathrm{h.c.}\bigr)\;+\;U\sum_i n_{i\uparrow}n_{i\downarrow}\;-\;\mu\sum_i n_i\,,$$2 AF	$$H_{\rm Hubb}\;=\;-\,t\sum_{\langle i,j\rangle,\sigma}\bigl(c_{i\sigma}^\dagger c_{j\sigma}+\mathrm{h.c.}\bigr)\;+\;U\sum_i n_{i\uparrow}n_{i\downarrow}\;-\;\mu\sum_i n_i\,,$$3	$$H_{\rm Hubb}\;=\;-\,t\sum_{\langle i,j\rangle,\sigma}\bigl(c_{i\sigma}^\dagger c_{j\sigma}+\mathrm{h.c.}\bigr)\;+\;U\sum_i n_{i\uparrow}n_{i\downarrow}\;-\;\mu\sum_i n_i\,,$$4 at any other $$H_{\rm Hubb}\;=\;-\,t\sum_{\langle i,j\rangle,\sigma}\bigl(c_{i\sigma}^\dagger c_{j\sigma}+\mathrm{h.c.}\bigr)\;+\;U\sum_i n_{i\uparrow}n_{i\downarrow}\;-\;\mu\sum_i n_i\,,$$5
Intermediate multimer/stripe phase	$$H_{\rm Hubb}\;=\;-\,t\sum_{\langle i,j\rangle,\sigma}\bigl(c_{i\sigma}^\dagger c_{j\sigma}+\mathrm{h.c.}\bigr)\;+\;U\sum_i n_{i\uparrow}n_{i\downarrow}\;-\;\mu\sum_i n_i\,,$$6	Weight spreads along one pair of $$H_{\rm Hubb}\;=\;-\,t\sum_{\langle i,j\rangle,\sigma}\bigl(c_{i\sigma}^\dagger c_{j\sigma}+\mathrm{h.c.}\bigr)\;+\;U\sum_i n_{i\uparrow}n_{i\downarrow}\;-\;\mu\sum_i n_i\,,$$7-points
Paramagnetic metal	$$H_{\rm Hubb}\;=\;-\,t\sum_{\langle i,j\rangle,\sigma}\bigl(c_{i\sigma}^\dagger c_{j\sigma}+\mathrm{h.c.}\bigr)\;+\;U\sum_i n_{i\uparrow}n_{i\downarrow}\;-\;\mu\sum_i n_i\,,$$8	$$H_{\rm Hubb}\;=\;-\,t\sum_{\langle i,j\rangle,\sigma}\bigl(c_{i\sigma}^\dagger c_{j\sigma}+\mathrm{h.c.}\bigr)\;+\;U\sum_i n_{i\uparrow}n_{i\downarrow}\;-\;\mu\sum_i n_i\,,$$9 is featureless

The crossing points can be estimated from $t>0$0 or $t>0$1 (Sherif et al., 21 Oct 2025). The first regime is the finite-density Haerter–Shastry antiferromagnet proper: a kinetically stabilized $t>0$2 state that persists up to substantial hole doping, specifically to $t>0$3 in the phase-diagram summary and $t>0$4 in the more detailed regime definition (Sherif et al., 21 Oct 2025).

Beyond this region, the system enters an intermediate multimer-stripe phase. This state is characterized by local “diamond” or multimer motifs inherited from the exactly three-fold-degenerate one-hole solution on a $t>0$5 triangular plaquette. Real-space correlations show antiferromagnetic nearest-neighbor $t>0$6 with modulated strength concentrated on columns parallel to one lattice direction, while the correlator $t>0$7 about a central site decays rapidly along the cylinder length but remains strong along the width, producing a quasi-1D stripe correlation (Sherif et al., 21 Oct 2025). The stripe periodicity is set by the underlying $t>0$8 diamond motif, and correlation lengths along stripes exceed those across stripes (Sherif et al., 21 Oct 2025).

At still larger hole density, the magnetic structure factor becomes featureless and the system is identified as a paramagnetic metal (Sherif et al., 21 Oct 2025).

5. Metallic character and charge-sector gaplessness

A notable result of (Sherif et al., 21 Oct 2025) is that the finite-doping phase diagram remains metallic throughout. Two diagnostics are presented.

The first is the quasiparticle residue

$t>0$9

where

$U$0

Although $U$1 at small $U$2, it rises monotonically with $U$3 and remains nonzero in all three phases, which is described as consistent with a heavy Fermi liquid (Sherif et al., 21 Oct 2025).

The second is the charge structure factor

$U$4

In a metal, $U$5 for $U$6, whereas an insulator would have $U$7. The DMRG line-cuts along the cylinder’s long direction fit excellently to the linear form, yielding a nonzero metallic weight $U$8 with a maximum in the intermediate phase (Sherif et al., 21 Oct 2025). The ground-state energy is also reported to be well explained by

$U$9

which is interpreted in the source as reflecting hole-density or quasiparticle renormalization of the kinetic bandwidth (Sherif et al., 21 Oct 2025).

This combination of robust magnetic ordering and gapless charge dynamics is central to the subject. The Haerter–Shastry antiferromagnet, as developed in (Sherif et al., 21 Oct 2025), is not a Mott-insulating ordered phase but part of a metallic finite-doping phase diagram.

6. Relation to superexchange and finite-$\mu$0 crossover

At finite $\mu$1, the triangular $\mu$2–$\mu$3 model contains two antiferromagnetic tendencies: kinetic Haerter–Shastry antiferromagnetism when $\mu$4, and Heisenberg-type $\mu$5 antiferromagnetism from superexchange $\mu$6 (Sherif et al., 21 Oct 2025). The cited DMRG study reports that, at fixed hole density such as $\mu$7, nearest-neighbor $\mu$8 and $\mu$9 initially grow in magnitude as $U$0 increases from zero, reach a maximum around a crossover $U$1, and then weaken toward the Haerter–Shastry limit where $U$2 vanishes (Sherif et al., 21 Oct 2025). The inflection in the on-site double occupancy $U$3 marks the same smooth crossover (Sherif et al., 21 Oct 2025).

The resulting $U$4 phase diagram distinguishes two antiferromagnetic sectors (Sherif et al., 21 Oct 2025):

Phase	Reported regime
AF$U$5 (HS-$U$6)	$U$7 and $U$8
AF$U$9 ($\mathcal P=\prod_i(1-n_{i\uparrow}n_{i\downarrow})$0 via superexchange)	$\mathcal P=\prod_i(1-n_{i\uparrow}n_{i\downarrow})$1 and $\mathcal P=\prod_i(1-n_{i\uparrow}n_{i\downarrow})$2, extending in a small $\mathcal P=\prod_i(1-n_{i\uparrow}n_{i\downarrow})$3 window
Stripe/multimer phase	$\mathcal P=\prod_i(1-n_{i\uparrow}n_{i\downarrow})$4 at $\mathcal P=\prod_i(1-n_{i\uparrow}n_{i\downarrow})$5, shrinking as $\mathcal P=\prod_i(1-n_{i\uparrow}n_{i\downarrow})$6 lowers
Paramagnetic metal	$\mathcal P=\prod_i(1-n_{i\uparrow}n_{i\downarrow})$7 at all $\mathcal P=\prod_i(1-n_{i\uparrow}n_{i\downarrow})$8

The paper characterizes the connection between the two antiferromagnetic mechanisms as a smooth crossover rather than a sharp phase boundary, with the two regimes meeting at a ridge $\mathcal P=\prod_i(1-n_{i\uparrow}n_{i\downarrow})$9 (Sherif et al., 21 Oct 2025). This is important for experimental interpretation because realistic platforms are expected to lie at large but finite $120^{\circ}$00, not at strictly infinite $120^{\circ}$01.

7. Relation to broader frustrated-magnet literature and experimental relevance

The Haerter–Shastry antiferromagnet belongs to the broader class of frustration-stabilized antiferromagnetic phenomena, but its microscopic origin is distinct from the exchange-frustration physics emphasized on the Shastry–Sutherland lattice. The latter supports dimer, plaquette, Néel, and valence-bond phases in Heisenberg and Hubbard realizations [(Nakano et al., 2018); (Xi et al., 2021); (Chen et al., 26 Mar 2025); (Liu et al., 2014); (Janson et al., 2011)]. By contrast, the Haerter–Shastry case is a triangular-lattice kinetic-magnetism problem in which antiferromagnetic order emerges “without any underlying magnetic interactions,” as stated in the abstract of (Sherif et al., 21 Oct 2025).

The experimental settings proposed in (Sherif et al., 21 Oct 2025) are cold-atom triangular lattices and moiré materials. For cold atoms, the discussion highlights tunable $120^{\circ}$02 via Feshbach resonance, the regime $120^{\circ}$03 where double occupancy is suppressed, and hole doping $120^{\circ}$04 as accessible by local spin-flip spectroscopy, with $120^{\circ}$05 measurable by Bragg scattering (Sherif et al., 21 Oct 2025). For moiré materials, the reported target regime is effective $120^{\circ}$06 with doping $120^{\circ}$07 reachable by gate tuning (Sherif et al., 21 Oct 2025). The same source notes that an enhanced Weiss constant with doping may reflect additive contributions of superexchange and Haerter–Shastry kinetic antiferromagnetism (Sherif et al., 21 Oct 2025).

A common source of confusion is the similarity of names involving “Shastry.” In the supplied arXiv corpus, the Shastry–Sutherland model and its decorated variants concern a geometrically frustrated square-lattice-derived network with orthogonal dimers [(Nakano et al., 2018); (Janson et al., 2011)], while the Haerter–Shastry antiferromagnet is defined on the triangular lattice and is rooted in kinetic interference (Sherif et al., 21 Oct 2025). This suggests that the most precise use of the term should reserve it for the triangular Hubbard phenomenon, even when comparing it to other frustrated antiferromagnets.

In that sense, the Haerter–Shastry antiferromagnet is a paradigmatic instance of kinetic-frustration-driven $120^{\circ}$08 Néel order that survives to finite hole density, crosses over smoothly to superexchange-driven antiferromagnetism at large but finite $120^{\circ}$09, and remains embedded in a metallic charge sector for all $120^{\circ}$10 studied (Sherif et al., 21 Oct 2025).