Frustration-free free fermions and beyond

Rintaro Masaoka, Seishiro Ono, Hoi Chun Po, Haruki Watanabe · 2025

Read Frustration-free free fermions and beyond by Rintaro Masaoka, Seishiro Ono, Hoi Chun Po, Haruki Watanabe (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2503-12879.txt.

Introduction

Rintaro Masaoka, Seishiro Ono, Hoi Chun Po, and Haruki Watanabe build a general framework for quantum systems whose ground state satisfies every local rule at once. Such frustration-free Hamiltonians are prized models: their ground states are exact and clean. Their 2025 paper works with free fermions, particles that ignore each other, and derives a necessary and sufficient condition for frustration-freeness. In orderly, repeating systems, any touching of the filled and empty energy bands must be at least quadratic. Extending a known inequality to fermions, they also bound how fast the energy gap closes as such systems grow.

These results police the border between gapped and gapless phases, a map every materials theorist uses. The gap law holds even with interactions and disorder, under a stated condition. The authors move from a tidy solvable case to a general statement.

Catalog record

CREATORRintaro Masaoka, Seishiro Ono, Hoi Chun Po, Haruki Watanabe
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE123 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2503-12879

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Frustration-free free fermions and beyond Rintaro Masaoka, Seishiro Ono, Hoi Chun Po, Haruki Watanabe Abstract Frustration-free Hamiltonians provide pivotal models for understanding quantum many-body systems. In this paper, we establish a general framework for frustration-free fermionic systems. First, we derive a necessary and sufficient condition for a free fermion model to be frustration-free. In the case of translation-invariant, noninteracting systems, we show that any band touching between the valence and conduction bands is at least quadratic. Furthermore, by extending the Gosset-Huang inequality to fermionic systems, we demonstrate that even in interacting and non-translation-invariant cases, the finite-size gap of gapless excitations scales as O ⁡ ( ( log ⁡ L ) 2 / L 2 ) O((\log L)^{2}/L^{2}) , provided the ground-state correlation function exhibits a power-law decay. Our results provide a foundation for studying frustration-free fermionic systems, including flat-band ferromagnetism and η \eta -pairing states. Preprint: RIKEN-iTHEMS-Report-25 I Introduction In condensed matter physics, frustration —the impossibility of minimizing all terms in the Hamiltonian simultaneously—is a key concept. One of the most fundamental examples is geometrical frustration in magnets. For example, in the antiferromagnetic Heisenberg model on a triangular lattice, no spin configuration can simultaneously minimize all antiferromagnetic exchange interactions. In contrast, we say that a Hamiltonian is frustration-free if the ground state of the total Hamiltonian is also a ground state of each local term in the Hamiltonian. Although frustration-free systems may appear to form a narrow class, they contain representative models of a diverse range of quantum phases of matter, such as quantum critical points, spontaneously symmetry breaking (SSB) phases, symmetry protected topological (SPT) phases, and topologically ordered phases. For instance, the Rokhsar-Kivelson model [ 1 ] , the Majumdar–Ghosh model [ 2 , 3 ] , the Affleck–Kennedy–Lieb–Tasaki model [ 4 , 5 ] , and the toric code model [ 6 ] are frustration-free models that represent a quantum critical point, an SSB phase, an SPT phase, and a topologically ordered phase, respectively. More generally, the parent Hamiltonians of matrix product states and projected entangled pair states are also frustration-free [ 7 , 8 ] . Frustration-free gapless systems exhibit features substantially different from those of frustrated systems [ 9 , 10 , 11 , 12 , 13 ] . Recently, it has been conjectured that when a frustration-free gapless Hamiltonian is translationally invariant, the dispersion relation of low-energy excitations above the ground state is generally quadratic or softer [ 12 , 14 ] . This implies that the low-energy effective theory of gapless frustration-free systems cannot be Lorentz invariant. Remarkably, such quadratic dispersions allow SSB of continuous symmetries at zero temperature even in one dimension [ 12 ] , which is typically forbidden by the Hohenberg-Mermin-Wagner theorem [ 15 , 16 , 17 ] . Also, when the ground state correlation function shows a power-law decay, the finite-size gap of gapless excitations scales as O(\log L)^{2}/L^{2}) [ 18 ] 1 1 1 Under the open boundary condition, a tighter bound O((L^{-2}) is derived without assuming power-law decay of correlation functions [ 78 , 79 , 80 , 81 , 82 ] . . To date, these intriguing properties have been studied primarily in frustration-free spin systems. It is therefore natural to ask whether fermionic frustration-free systems have similar characteristics. However, despite some studies on frustration-free fermionic systems [ 20 , 21 , 22 , 23 , 24 ] , their properties remain less comprehensively explored compared to those of spin systems. In this work, we study frustration-free fermionic systems. First, we focus on frustration-free free fermionic systems, where Hamiltonians are expressed as quadratic forms of fermionic creation and annihilation operators. We derive a necessary and sufficient condition for free fermionic systems to be frustration-free. Such a condition has been previously investigated for 1D one-band models in class BDI [ 23 ] and 1D two-band models in class AIII [ 24 ] from a different approach. Our results are more general in that no assumptions are made regarding the dimensionality of the system, the number of bands, or the symmetry classes. We find a deep connection to tight-binding models with flat bands [ 25 , 26 , 27 , 28 , 29 , 30 , 31 , 32 , 5 ] . Leveraging this condition and band theory, we prove that for gapless systems with translation symmetry, the excitation energy disperses as O(|\bm{k}-\bm{k}_{0}|^{2}) around some \bm{k}_{0} . We also investigate the topological properties of frustration-free free fermionic models. We establish that the valence bands in these systems always admit an (over)complete set of compactly supported Wannier-type functions, provided that the valence band energy is strictly negative. The classification of topological phases with such Wannier-type functions [ 28 ] implies that stable topological phases, including Chern insulators and \mathbb{Z}_{2} topological insulators, cannot be realized in frustration-free free fermionic models. In contrast, we demonstrate via a concrete model that fragile topology [ 33 ] can be realized. We then extend our results to more general fermionic systems, including cases with interactions and without translation symmetry, by generalizing the Gosset-Huang inequality [ 34 ] to fermionic cases. With this extension, we show that even in interacting and/or non-translationally invariant fermionic systems, the finite-size gap of gapless excitations scales as O((\log L)^{2}/L^{2}) , provided the ground state correlation function exhibits a power-law decay. The main results of this work are presented in conjunction with a companion letter [ 35 ] . While the companion letter includes separate discussions on a frustration-free model on the honeycomb lattice and the effects of interactions in that model, the present paper establishes the comprehensive general framework. Here, we provide rigorous mathematical derivations, extend the analysis to \text{U}(1) symmetry-breaking systems, detail the topological properties, and prove the finite-size gap scaling in interacting systems. The remainder of this paper is organized as follows. In Sec. II , we derive a necessary and sufficient condition for frustration-freeness in free fermionic systems. We also show that this condition implies that the low-energy excitation is quadratic or softer for frustration-free free fermionic systems with translation symmetry. In Sec. III , we present several illustrative examples. In Sec. IV , we discuss the case of \text{U}(1) -symmetry breaking. In Sec. V , we examine the possible stable topology realizable within frustration-free free-fermion models. In Sec. VI , we discuss the Gosset–Huang inequality for general fermionic systems. II U(1)-symmetric free fermionic systems In this section, we consider noninteracting fermions described by a tight-binding model with U(1) symmetry. II.1 Setting In general, the Hamiltonian \hat{H}=\sum_{\bm{R}\in\Lambda}\hat{H}_{\bm{R}} is given by a sum of local terms \hat{H}_{\bm{R}} , which take the quadratic form \displaystyle\hat{H}_{\bm{R}} \displaystyle=\sum_{\bm{R}^{\prime},\bm{R}^{\prime\prime}\in\Lambda}\sum_{\tau,\tau^{\prime}=1}^{N_{\bm{R}}}\hat{c}_{\bm{R}^{\prime}\tau}^{\dagger}(t_{\bm{R}})_{\bm{R}^{\prime}\tau,\bm{R}^{\prime\prime}\tau^{\prime}}\hat{c}_{\bm{R}^{\prime\prime}\tau^{\prime}}+C_{\bm{R}}. (1) Here, C_{\bm{R}} is a constant that will be chosen later, \bm{R}\in\Lambda is the unit cell label, and the indices \tau,\tau^{\prime}=1,2,\cdots,N_{\bm{R}} include all internal degrees of freedom. The creation/annihilation operators satisfy the anti-commutation relations \{\hat{c}_{\bm{R}\tau},\hat{c}_{\bm{R}^{\prime}\tau^{\prime}}^{\dagger}\}=\delta_{\bm{R},\bm{R}^{\prime}}\delta_{\tau,\tau^{\prime}} and \{\hat{c}_{\bm{R}\tau},\hat{c}_{\bm{R}^{\prime}\tau^{\prime}}\}=0 . The Hamiltonian is local in the sense that (t_{\bm{R}})_{\bm{R}^{\prime}\tau,\bm{R}^{\prime\prime}\tau^{\prime}}=0 whenever |\bm{R}^{\prime}-\bm{R}|>r or |\bm{R}^{\prime\prime}-\bm{R}|>r . This r is called the range of the hopping, which is assumed to be finite and independent of the system size. Hence, we can regard the hopping amplitude (t_{\bm{R}})_{\bm{R}^{\prime}\tau,\bm{R}^{\prime\prime}\tau^{\prime}} in Eq. ( 1 ) as a matrix with indices \bm{R}^{\prime}\tau and \bm{R}^{\prime\prime}\tau^{\prime} and \hat{c}_{\bm{R}^{\prime\prime}\tau^{\prime}} as a vector with an

Read the complete text