Nonsymmorphic two-dimensional Dirac semimetals: experimental realization at $4n+2$ filling, correlated gapped phases, and disorder-induced localization

Young and Kane (Phys. Rev. Lett. 115, 126803 (2015)) showed that two-dimensional (2D) Dirac semimetals whose Dirac points survive spin-orbit coupling are enforced by nonsymmorphic lattice symmetries (glide mirrors and screw axes), are pinned to the boundary between 2D topological and trivial insulators, and require an electron filling of $4n+2$. The construction, however, was established only for a clean, weakly interacting tight-binding model. The open problem is to establish this phase as a real and robust state of matter: (i) identify a 2D material or engineered superlattice with the required layer-group symmetries, $4n+2$ filling, and sizable spin-orbit coupling, with the symmetry-protected Dirac points at the Fermi level (requiring band-structure engineering to undo the particle-hole-breaking second-neighbor hopping $t_2$, and demonstrating both the topological and the trivial insulating distortions); (ii) determine the fate under strong electron-electron interactions — in particular whether a symmetric gapped state must carry nontrivial topological order, and of what type; and (iii) determine the disorder threshold and transport consequences of localization in this strictly 2D semimetal. Progress as of 2026 is substantial but incomplete: symmetry-protected, SOC-robust Dirac crossings have been observed by ARPES in nonsymmorphic alpha-bismuthene and alpha-antimonene, but with Dirac points away from the Fermi level, while the interacting and disorder questions for this 2D class remain essentially untouched.

Background

A Dirac semimetal is a crystal whose conduction and valence bands touch linearly at isolated points (Dirac points), giving low-energy quasiparticles with a relativistic, graphene-like dispersion. Graphene is the canonical 2D example, but its Dirac points are gapped by any realistic spin-orbit coupling, making graphene formally a quantum spin Hall (QSH) insulator — a 2D topological insulator (TI) with helical, symmetry-protected edge channels. A symmetry-protected 2D Dirac semimetal instead requires Dirac points that remain gapless in the presence of significant spin-orbit coupling. Young and Kane (PRL 115, 126803 (2015)) showed this is impossible with ordinary point symmetries and requires nonsymmorphic space-group symmetries. A nonsymmorphic symmetry combines a point operation with a translation by a fraction of a lattice vector — e.g. a glide mirror ${M_z|1/2,1/2}$ (mirror reflection plus half-diagonal translation) or a two-fold screw rotation ${C_{2x}|1/2,0}$. Because such an operation squares to a lattice translation, its Bloch eigenvalues pick up sign factors $e^{i\mathbf{k}\cdot\mathbf{t}}$ that flip between the Brillouin-zone center and edge. Together with time-reversal symmetry, whose Kramers theorem pairs states of opposite spin at momenta $\mathbf{k}$ and $-\mathbf{k}$, this forces 'band sticking' at zone-edge momenta (such as $X_1$, $X_2$, $M$ in their square-lattice model): four bands stick together, so a time-reversal-symmetric crystal with an electron filling of $4n+2$ (an even number of filled Kramers-degenerate bands per primitive cell, leaving the stuck bands half-filled) must be gapless with Dirac points pinned at those momenta. A second-neighbor hopping $t_2$ in their two-site model breaks particle-hole symmetry, shifting the $M$ Dirac point relative to $X_{1,2}$ and producing spurious electron and hole pockets; only with $t_2 \approx 0$ or with band-structure engineering do all Dirac points align at the Fermi level $E_F$. The same analysis shows a single symmetry-protected Dirac point is impossible in 2D, and that the semimetal sits exactly at the phase boundary between a QSH (topological) insulator and a trivial insulator: symmetry-breaking lattice distortions that preserve one glide/screw gap the system into one insulator or the other. Young and Kane identified concrete candidate structures — litharge-type (the layered PbO structure) and $WTe_2$-type monolayers (which have the appropriate layer groups, the 2D analogues of space groups), $(Nb,Ta)Te_2$ with the right filling, and the iridium-oxide superlattices of Chen and Kee (PRB 90, 195145 (2014)), where distortions to the TI phase were already established — and noted that the physically central issues lie beyond the clean, non-interacting model. First, for any candidate material the Dirac points must actually sit at $E_F$ (filling $4n+2$ and $t_2$ control), otherwise one measures symmetry-protected band crossings but not a semimetal. Second, interactions: weak electron-electron interactions leave the band structure intact, but strong interactions could gap the semimetal; if the protecting spatial symmetries are not broken, the general argument of Parameswaran, Turner, Arovas and Vishwanath (Nat. Phys. 9, 299 (2013), for nonsymmorphic crystals at integer filling) implies the gapped state cannot be an ordinary band insulator and must carry nontrivial topological order — long-range quantum entanglement, typically with fractionalized quasiparticles; which orders can actually occur was left undetermined. (A competing modern possibility, symmetric mass generation, is an interaction-driven, symmetry-preserving fermion gap without bilinear condensation, debated as an alternative to topological order for (2+1)D Dirac systems.) Third, disorder: in 2D the trivial and TI phases are not related by a symmetry, so unlike Dirac cones on the surface of a 3D topological phase, which cannot be Anderson-localized (Anderson localization being the confinement of an electron's wave function by interference between randomly scattered partial waves) even by strong disorder, the 2D nonsymmorphic Dirac semimetal lies in the symplectic symmetry class (time-reversal-invariant systems with strong spin-orbit coupling, for which weak disorder causes anti-localization) and can be localized by sufficiently strong disorder. Angle-resolved photoemission spectroscopy (ARPES) is the standard probe that maps experimental band dispersions and is the tool of choice for observing these Dirac crossings.

Problem Statement

Establish the nonsymmorphic two-dimensional Dirac semimetal of Young and Kane as a real and robust phase of matter, completing the physical picture beyond the clean, non-interacting tight-binding model in which it was defined. The common objective ties three linked questions about one and the same phase. (i) Experimental realization: identify a two-dimensional material, or an engineered heterostructure or superlattice, that has the required nonsymmorphic layer-group symmetries (glide mirror and/or screw axes), an electron filling of $4n+2$, and sizable spin-orbit coupling, such that the symmetry-protected Dirac points at the Brillouin-zone-edge momenta lie at the Fermi level — which entails the band-structure engineering needed to counter the particle-hole-breaking second-neighbor hopping $t_2$ and remove extraneous electron and hole pockets — and demonstrate, in such a system, the lattice distortions that gap the semimetal into both the topological and the trivial insulating phase. (ii) Interacting fate: determine what strong electron-electron interactions do to this semimetal — in particular, if the interactions gap it without breaking the protecting spatial symmetries, establish whether the gapped state must carry nontrivial topological order, and determine which topological orders (or whether a short-range-entangled symmetric mass-generated phase) can occur. (iii) Disorder: determine the disorder-strength threshold at which this strictly two-dimensional, symplectic-class semimetal localizes (weak disorder gives antilocalization) and the associated transport behavior, given that its trivial and topological insulating neighbors are not related by a symmetry. The three parts are jointly posed and refer to one and the same phase: (i) Realization: a concretely specified material (or engineered heterostructure or superlattice) meeting the nonsymmorphic-symmetry, $4n+2$-filling, sizable-spin-orbit-coupling, and Dirac-points-at-$E_F$ criteria and backed by an explicit band structure in which the $t_2$ offsets are countered — a design-level realization rather than a laboratory synthesis — together with a demonstration, from that band structure or an explicit model, that symmetry-preserving distortions gap the system into both the topological and the trivial insulating phase. (ii) Interactions: an explicit determination of whether a symmetry-preserving interaction-induced gap must carry nontrivial topological order, and of which topological orders (or short-range-entangled symmetric mass-generated phase) can occur. (iii) Disorder: an explicit disorder-strength threshold at which the strictly two-dimensional, symplectic-class semimetal localizes, together with the associated transport behavior.

References

1. S. M. Young and C. L. Kane, 'Dirac Semimetals in Two Dimensions', Phys. Rev. Lett. 115, 126803 (2015). DOI: 10.1103/PhysRevLett.115.126803; arXiv:1504.07977; https://doi.org/10.1103/PhysRevLett.115.126803
2. Yige Chen and Hae-Young Kee, 'Topological phases in iridium oxide superlattices: Quantized anomalous charge or valley Hall insulators', Phys. Rev. B 90, 195145 (2014). DOI: 10.1103/PhysRevB.90.195145; https://doi.org/10.1103/PhysRevB.90.195145
3. Siddharth A. Parameswaran, Ari M. Turner, Daniel P. Arovas, and Ashvin Vishwanath, 'Topological order and absence of band insulators at integer filling in non-symmorphic crystals', Nat. Phys. 9, 299 (2013). DOI: 10.1038/nphys2600; https://doi.org/10.1038/nphys2600
4. Jing Wang, 'Antiferromagnetic Dirac semimetals in two dimensions', Phys. Rev. B 95, 115138 (2017). DOI: 10.1103/PhysRevB.95.115138; arXiv:1609.03259; https://doi.org/10.1103/PhysRevB.95.115138
5. Steve M. Young and Benjamin J. Wieder, 'Filling-enforced magnetic Dirac semimetals in two dimensions', Phys. Rev. Lett. 118, 186401 (2017). DOI: 10.1103/PhysRevLett.118.186401; arXiv:1609.06738; https://doi.org/10.1103/PhysRevLett.118.186401
6. Y. J. Jin, B. B. Zheng, X. L. Xiao, Z. J. Chen, Y. Xu, and H. Xu, 'Two-Dimensional Dirac Semimetals without Inversion Symmetry', Phys. Rev. Lett. 125, 116402 (2020). DOI: 10.1103/PhysRevLett.125.116402; arXiv:2008.10175; https://doi.org/10.1103/PhysRevLett.125.116402
7. Pawel J. Kowalczyk, Simon A. Brown, Tobias Maerkl, Qiangsheng Lu, Ching-Kai Chiu, Ying Liu, Shengyuan A. Yang, Xiaoxiong Wang, Ilona Zasada, Francesca Genuzio, Tevfik Onur Menteş, Andrea Locatelli, Tai-Chang Chiang, and Guang Bian, 'Realization of Symmetry-Enforced Two-Dimensional Dirac Fermions in Nonsymmorphic α-Bismuthene', ACS Nano 14, 1888-1894 (2020). DOI: 10.1021/acsnano.9b08136; arXiv:1906.08456; https://doi.org/10.1021/acsnano.9b08136
8. Qiangsheng Lu, Kyle Y. Chen, Matthew Snyder, Jacob Cook, Duy Tung Nguyen, P. V. Sreenivasa Reddy, Tay-Rong Chang, Shengyuan A. Yang, and Guang Bian, 'Observation of symmetry-protected Dirac states in nonsymmorphic α-antimonene', Phys. Rev. B 104, L201105 (2021). DOI: 10.1103/PhysRevB.104.L201105; arXiv:2101.05793; https://doi.org/10.1103/PhysRevB.104.L201105
9. Shan Guan, Ying Liu, Zhi-Ming Yu, Shan-Shan Wang, Yugui Yao, and Shengyuan A. Yang, 'Two-dimensional Spin-Orbit Dirac Point in Monolayer HfGeTe', Phys. Rev. Materials 1, 054003 (2017). DOI: 10.1103/PhysRevMaterials.1.054003; arXiv:1706.08692; https://doi.org/10.1103/PhysRevMaterials.1.054003
10. Xiaotong Fan, Dashuai Ma, Botao Fu, Cheng-Cheng Liu, and Yugui Yao, 'Cat's-cradle-like Dirac semimetals in layer groups with multiple screw axes: Application to two-dimensional borophene and borophane', Phys. Rev. B 98, 195437 (2018). DOI: 10.1103/PhysRevB.98.195437; arXiv:1809.09431; https://doi.org/10.1103/PhysRevB.98.195437
11. Seoung-Hun Kang, Wei Luo, Sinchul Yeom, Yaling Zheng, and Mina Yoon, 'Two-dimensional Dirac semimetal based on the alkaline earth metal CaP$_3$', Phys. Rev. Materials 7, 124202 (2023). DOI: 10.1103/PhysRevMaterials.7.124202; arXiv:2303.01627; https://doi.org/10.1103/PhysRevMaterials.7.124202
12. V. Damljanović, 'Non-magnetic layers with a single symmetry-protected Dirac cone: Which additional dispersions must appear?', EPL 147, 56003 (2024). DOI: 10.1209/0295-5075/ad7317; arXiv:2402.11982; https://doi.org/10.1209/0295-5075/ad7317
13. Yi-Zhuang You, Yin-Chen He, Cenke Xu, and Ashvin Vishwanath, 'Symmetric Fermion Mass Generation as Deconfined Quantum Criticality', Phys. Rev. X 8, 011026 (2018). DOI: 10.1103/PhysRevX.8.011026; https://doi.org/10.1103/PhysRevX.8.011026
14. Yu-Te Hsu, Danil Prishchenko, Maarten Berben, Matija Čulo, Steffen Wiedmann, Emily C. Hunter, Paul Tinnemans, Tomohiro Takayama, Vladimir Mazurenko, Nigel E. Hussey, and Robin S. Perry, 'Evidence for strong electron correlations in a nonsymmorphic Dirac semimetal', npj Quantum Materials 6, 92 (2021). DOI: 10.1038/s41535-021-00396-5; arXiv:2105.03740; https://doi.org/10.1038/s41535-021-00396-5
15. Zhi-Ming Yang and Huan Li, 'Correlated Dirac semimetal states in nonsymmorphic MIrO3 (M = Sr, Ba and Ca)', J. Phys.: Condens. Matter 37, 325602 (2025). DOI: 10.1088/1361-648X/adf67c; arXiv:2505.12673; https://doi.org/10.1088/1361-648X/adf67c
16. Yu-Li Lee and Yu-Wen Lee, 'The phase diagram of a two-dimensional dirty tilted Dirac semimetal', Phys. Rev. B 100, 075156 (2019). DOI: 10.1103/PhysRevB.100.075156; arXiv:1904.00514; https://doi.org/10.1103/PhysRevB.100.075156
17. Ilias Amanatidis and Ioannis Kleftogiannis, 'Robust non-integer conductance in disordered 2D Dirac semimetals', J. Phys.: Condens. Matter 34, 275301 (2022). DOI: 10.1088/1361-648x/ac6786; arXiv:2110.09177; https://doi.org/10.1088/1361-648x/ac6786