Flatland
Reading as

Tantalum nickel selenide

Ta₂NiSe₅; sulfide analogue Ta₂NiS₅
van der Waals crystal semiconductor

The leading candidate for a state proposed in the 1960s: the excitonic insulator, in which electrons and holes in a semiconductor with an almost vanishing gap bind into pairs of their own accord and condense, opening a gap themselves. Ta2NiSe5 shows the expected signatures – a strikingly flat valence-band top and a gap that opens at 326 K, just above room temperature – but its crystal distorts at the same moment, and that distortion alone can open a gap. Deciding which comes first has made it the test case for telling electronic order from lattice order.

Crystal structure

  • Se
  • Ni
  • Ta
Cell
Rectangular, a = 3.50 Å (along chains), b = 15.68 Å (across chains)
Atoms per cell
16
Ta–Se bonds
2.56–2.68 Å
Ni–Se bonds
2.35 and 2.39 Å
Height
4.36 Å between the outer atom centres
Double chains of tantalum in selenium octahedra flank single chains of nickel in selenium tetrahedra, and all of them run along one direction. The valence band comes mostly from the nickel chains and the conduction band from the tantalum chains, and the two sit close together in energy – the setting in which electrons and holes might pair up, which is why Ta2NiSe5 is studied as a candidate excitonic insulator. The model shows the orthorhombic structure above the transition near 326 K; below it, the cell tilts very slightly into a monoclinic form. One layer of Ta2NiSe5 in its orthorhombic phase at 400 K (Nakano and colleagues, IUCrJ 5, 158, 2018; COD 1556756): a = 3.50 Å along the chains, b = 15.68 Å across them.

Key properties

  • A transition at ~326 K from an orthorhombic, nearly zero-gap phase to a monoclinic insulating phase
  • Photoemission: an extremely flat valence-band top below the transition, the fingerprint expected of an excitonic insulator
  • An optical gap of ~0.16 eV below the transition, comparable to the estimated exciton binding energy, and a transition entropy that looks mostly electronic
  • Raman scattering: excitonic fluctuations growing towards ~241 K, overtaken by a structural transition at 325 K driven by an acoustic phonon of the same symmetry
  • Femtosecond light pulses narrow the gap below a threshold fluence of ~0.2 mJ cm−2 and widen it above; quench experiments point to a mostly structural order parameter
  • Under pressure the rippled layers flatten at 3 GPa, and superconductivity up to 1.2 K appears near 8 GPa

How it is made

  • Bulk crystals by chemical vapour transport
  • Mechanical exfoliation down to a few layers: five-layer flakes keep the crystal and electronic structure, with a transition ~9% lower than in bulk
  • Chemical substitution and hydrostatic pressure tune the one-electron gap across a dome-shaped phase diagram

Uses, and how close they are

  • Model system for excitonic condensation and correlated band gapslab
  • Band gaps switched by light on femtosecond timescaleslab

Readiness runs lab → prototype → pilot → deployed.

Open problems

  1. Is the 326 K transition driven by electron–hole pairing, by a lattice instability, or by both together – and how much of the gap does each contribute?
  2. Which experiment can separate an excitonic and a structural order parameter that share the same symmetry?
  3. Does any part of the ordered phase behave like an exciton condensate able to carry energy with little dissipation?
  4. The ordered phase is already confined to single layers in bulk – what happens to it in an isolated monolayer?

Going deeper

Short notes for specialists. Choose a lens in the header and yours comes first.

For theoreticians · your lens

The textbook excitonic-insulator setting – a Ni 3d–Se 4p valence band and a Ta 5d conduction band almost touching at the zone centre – has become a symmetry problem. Mirror symmetries of the orthorhombic phase forbid the two bands from hybridising, so either an excitonic order parameter or a shear distortion of the same symmetry can break them and open the gap, and the two couple linearly: whichever drives the transition drags the other along. Chain models with electron–phonon coupling reproduce the flat band from exciton condensation, and symmetry analysis identifies a purely electronic order parameter. First-principles total-energy calculations, by contrast, find symmetry-breaking phonon instabilities in both Ta2NiSe5 and Ta2NiS5 and no need for excitons. Raman data place the bare excitonic instability near 241 K, below the observed transition.

For experimentalists · your lens

Because the electronic and structural order parameters share a symmetry, a static measurement of the gap or the monoclinic angle cannot tell them apart. Use probes that separate them by timescale or susceptibility: time-resolved photoemission and diffraction after an ultrafast quench, and Raman scattering in the B2g channel to follow critical fluctuations. Compare with Ta2NiS5, which shows no excitonic signatures. Report each crystal’s transition temperature and, for flakes, the thickness – the transition shifts in ultrathin samples.

For engineers · your lens

Not an engineering material. Its device interest lies in speed: light pulses shrink or enlarge its band gap within femtoseconds, and its ordered phase survives at room temperature. The hope that an exciton condensate might transport energy almost without loss is weakened by evidence that the gap is mostly structural.

In the research tracks

Theory frontiers

Growth methods

Recent news

The newest items tagged Ta2NiSe5, from the news feed updated 5 Oct 2026.

Preprintnot yet peer reviewed arXiv

Substrate tuning of the structural and electronic transition in thin flakes of the excitonic insulator candidate Ta2NiSe5

Ta2NiSe5 continues to draw interest for its Tc = 326 K phase transition, whose dual electronic and structural nature reflects a complex interplay of electron-hole (excitonic) and electron-lattice interactions. The majority of studies that have attempted to decipher the relative importance of these interactions…

ExperimentTa₂NiSe₅
Preprintnot yet peer reviewed arXiv

Defect-Mediated Nucleation and Dynamics across the Phase Transition in the Excitonic Insulator Candidate Ta2NiSe5

Ta2NiSe5 is a quasi-one-dimensional material that exhibits a structural and electronic phase transition from a low-temperature monoclinic (semiconductor) to a high-temperature orthorhombic (semimetal) phase at approximately TC = 326 K. Here, we used variable-temperature scanning tunneling microscopy and spectroscopy to…

ExperimentTa₂NiSe₅
Preprintnot yet peer reviewed arXiv

Defect-Mediated Nucleation and Dynamics across the Phase Transition in the Excitonic Insulator Candidate Ta2NiSe5

Ta2NiSe5 is a quasi-one-dimensional material that exhibits a structural and electronic phase transition from a low-temperature monoclinic (semiconductor) to a high-temperature orthorhombic (semimetal) phase at approximately TC = 326 K. Here, we used variable-temperature scanning tunneling microscopy and spectroscopy to…

ExperimentTa₂NiSe₅

All 8 items tagged Ta₂NiSe₅ in the news feed  ·  RSS feed for Ta₂NiSe₅

Key references

  1. Structure and physical properties of the new layered ternary chalcogenides tantalum nickel sulfide (Ta2NiS5) and tantalum nickel selenide (Ta2NiSe5)Sunshine & Ibers · Inorganic Chemistry 24, 3611 (1985)cited by 168doi:10.1021/ic00216a027
  2. Excitonic insulator state in Ta2NiSe5 probed by photoemission spectroscopyWakisaka et al. · Physical Review Letters 103, 026402 (2009)cited by 314doi:10.1103/PhysRevLett.103.026402
  3. Layer-confined excitonic insulating phase in ultrathin Ta2NiSe5 crystalsKim et al. · ACS Nano 10, 8888 (2016)cited by 71doi:10.1021/acsnano.6b04796
  4. Zero-gap semiconductor to excitonic insulator transition in Ta2NiSe5Lu et al. · Nature Communications 8, 14408 (2017)cited by 290doi:10.1038/ncomms14408
  5. Ultrafast electronic band gap control in an excitonic insulatorMor et al. · Physical Review Letters 119, 086401 (2017)cited by 187doi:10.1103/PhysRevLett.119.086401
  6. Nature of symmetry breaking at the excitonic insulator transition: Ta2NiSe5Mazza et al. · Physical Review Letters 124, 197601 (2020)cited by 131doi:10.1103/PhysRevLett.124.197601
  7. Direct observation of excitonic instability in Ta2NiSe5Kim et al. · Nature Communications 12, 1969 (2021)cited by 94doi:10.1038/s41467-021-22133-z
  8. Common microscopic origin of the phase transitions in Ta2NiS5 and the excitonic insulator candidate Ta2NiSe5Windgätter et al. · npj Computational Materials 7, 210 (2021)cited by 48doi:10.1038/s41524-021-00675-6
  9. Hybridization-gap formation and superconductivity in the pressure-induced semimetallic phase of the excitonic insulator Ta2NiSe5Matsubayashi et al. · Journal of the Physical Society of Japan 90, 074706 (2021)cited by 28doi:10.7566/JPSJ.90.074706
  10. The spontaneous symmetry breaking in Ta2NiSe5 is structural in natureBaldini et al. · PNAS 120, e2221688120 (2023)cited by 67doi:10.1073/pnas.2221688120
Transition metal dichalcogenides TiSe₂, ZrSe₂, HfSe₂ Titanium, zirconium and hafnium dichalcogenides van der Waals crystal
type
depends on form
thickness
~0.57–0.66 nm per layer
band gap
0.5–1.2 eV
most mature
prototype
Lithium and sodium battery electrodes (TiS2) Transition metal dichalcogenides WTe₂ Tungsten ditelluride van der Waals crystal
type
depends on form
thickness
~0.7 nm per layer
most mature
lab
Topological edge-state and superconducting research devices Other layered chalcogenides TiS₃, ZrTe₃, NbSe₃, TaSe₃ Quasi-one-dimensional trichalcogenides van der Waals crystal
type
depends on form
thickness
~0.9 nm per TiS₃ layer
band gap
1–2 eV
most mature
lab
Polarisation-sensitive photodetectors (TiS3, ZrS3)