Monolayer 1T' WTe2: Structure, Topology & Tuning
- Monolayer 1T'-WTe2 is a two-dimensional material defined by a symmetry-lowered, distorted octahedral lattice with inequivalent W–Te bonds and a tunable narrow-gap electronic structure.
- Its synthesis via molecular-beam epitaxy and graphene encapsulation is characterized by precise phase identification through ARPES, STM/STS, and Raman spectroscopy.
- The material exhibits quantum spin Hall edge modes with anisotropic conduction, while external tuning via strain, electric field, and temperature modulates its topological behavior.
Monolayer “1T WTe” in contemporary electronic-structure and device literature almost always denotes the distorted octahedral monolayer $1T'$-WTe rather than an ideal undistorted $1T$ sheet. In that usage, the relevant single layer is the isolated distorted layer that underlies bulk -WTe, and its defining features are a symmetry-lowered lattice, inequivalent W–Te bonds, narrow-gap or positive-gap low-energy electronic structure, and quantum-spin-Hall-related edge phenomenology (Ando et al., 29 Jan 2025, Tang et al., 2017). The distinction is not semantic: the experimentally realized monolayer phase discussed across ARPES, STM/STS, transport, and first-principles studies is the $1T'$ polymorph, not an ideal high-symmetry $1T$ monolayer (Ando et al., 29 Jan 2025).
1. Structural identity and crystallographic conventions
The parent $1T$ structure is octahedrally coordinated and higher in symmetry, whereas monolayer $1T'$-WTe$1T'$0 is obtained by a symmetry-lowering distortion of that octahedral lattice. In the structural description used in the literature, the distortion produces an orthorhombic or rectangular layer with inequivalent local environments, zigzag or chain-like W motifs, and two kinds of W–Te bond lengths rather than the single bond length characteristic of $1T'$1-WTe$1T'$2 (Ando et al., 29 Jan 2025, Tang et al., 2017).
A widely used microscopic picture is that the distortion doubles the periodicity along one in-plane direction and shifts W atoms away from ideal octahedral positions into zigzag chains. One ARPES/STM study reports characteristic in-plane scales of about 3.48 Å and about 6.27 Å for the distorted monolayer cell, while RHEED and STM/Fourier analysis on epitaxial films gave about $1T'$3 Å, $1T'$4 Å, and $1T'$5 Å, together with an angular distortion of about $1T'$6 (Tang et al., 2017). In bulk language, $1T'$7-WTe$1T'$8 can be described as stacked distorted $1T'$9 layers related by 180° rotation, whereas the monolayer is the isolated 0 layer (Ando et al., 29 Jan 2025).
The symmetry assignment is not entirely uniform across the supplied literature. Several topological and ARPES studies describe monolayer 1-WTe2 as inversion symmetric (Tang et al., 2017, Cucchi et al., 2018, Qiao et al., 18 Apr 2026), whereas a later spin-Hall calculation assigns a polar orthorhombic 3 (No. 31) structure with mirror symmetry 4 to the monolayer (Wu et al., 20 Jun 2026). This suggests that structural parametrization and symmetry bookkeeping remain model-dependent in parts of the literature, even though the phase identity as distorted 5-WTe6 is consistent.
2. Epitaxy, sample preparation, and structural diagnostics
A central experimental route is molecular-beam epitaxy on graphene-terminated SiC. In one selective-growth study, bilayer graphene was first formed on an n-type 4H-SiC(0001) wafer by resistive heating at 1100 °C for 15 min under vacuum better than 7 Torr, after which WTe8 was grown by evaporating W in a Te-rich atmosphere (Ando et al., 29 Jan 2025). The decisive control parameter was the substrate temperature 9: 280 °C optimized monolayer 1H-WTe$1T$0, whereas 310 °C optimized monolayer 1T'-WTe$1T$1; the films were then annealed for 30 min at the same $1T$2 (Ando et al., 29 Jan 2025).
Phase identification is spectroscopic rather than purely diffraction-based. For the $1T$3 phase, ARPES is combined with Te $1T$4 core-level spectroscopy and first-principles calculations. In monolayer $1T$5-WTe$1T$6, each Te spin-orbit component shows a two-peaked structure, attributed to the two inequivalent bond lengths generated by the distortion; in $1T$7-WTe$1T$8, the same components appear as single peaks (Ando et al., 29 Jan 2025). A separate epitaxial QSH study further reported two Te $1T$9 doublets and a clean W 0 doublet characteristic of a pure 1 phase (Tang et al., 2017).
The substrate symmetry strongly affects domain structure. Because monolayer 2-WTe3 has twofold symmetry while bilayer graphene has threefold or sixfold symmetry, epitaxy produces three rotational domains related by 120° (Ando et al., 29 Jan 2025, Tang et al., 2017). That domain multiplicity limits conventional diffraction as a decisive phase fingerprint and motivates the suggestion that nano-ARPES is required to resolve single-domain band structure (Ando et al., 29 Jan 2025). In STM-based work on epitaxial islands, typical monolayer islands were about 20 nm 4 20 nm, with monolayer height above bilayer graphene around 5 Å (Tang et al., 2017).
Exfoliated monolayers can also be probed directly when protected by graphene encapsulation. Laser micro-ARPES on graphene-capped flakes achieved an effective spatial resolution of about 6m FWHM, with 6.01 eV photons, typical energy resolution 2 meV, and momentum resolution 0.003 Å7 (Cucchi et al., 2018). That work emphasized that encapsulation is essential because non-encapsulated ultrathin WTe8 shows heavily degraded, largely featureless spectra, whereas encapsulated flakes retain well-defined dispersions and can substantially recover after brief UHV annealing following air exposure (Cucchi et al., 2018).
3. Electronic structure and the band-gap problem
The low-energy electronic structure is the most intensively debated aspect of monolayer 9-WTe0. In ARPES on selectively grown monolayers, the 1 phase shows four hole bands near the relevant zone center or topological point at approximately 0.1, 0.6, 0.8, and 1.3 eV below 2, and the study describes the phase as a semiconducting monolayer with a nearly-zero energy gap rather than an unambiguous metal (Ando et al., 29 Jan 2025). In that work, plain GGA broadly reproduced the valence structure, but HSE06 was required to reproduce the narrow band gap around 3 (Ando et al., 29 Jan 2025).
Other spectroscopies report a clearly positive gap. In epitaxial ARPES with surface K dosing, the extracted bulk gap was 4 meV for the intrinsic monolayer and 5 meV after K dosing, while bulk-interior STS gave 6 meV from a histogram of 115 spectra (Tang et al., 2017). Laser micro-ARPES on encapsulated exfoliated monolayers found a CBM = 7 meV, VBM = 8 meV, and an inferred gap of 9 meV (Cucchi et al., 2018). Thickness-dependent MBE ARPES later reported an indirect monolayer gap of about 55 meV, but with the conduction band lying about $1T'$0 meV below $1T'$1 because the measured monolayer was $1T'$2-type self-doped (Qiao et al., 18 Apr 2026).
A temperature-dependent ARPES study widened the positive-gap picture. In epitaxial monolayer $1T'$3-WTe$1T'$4 on graphene/SiC(0001), FD-divided spectra directly resolved an electron-like conduction band above $1T'$5 together with four hole bands below $1T'$6, leading to the conclusion that the monolayer has an indirect band gap larger than about 0.1 eV; fitted values were about 120 meV at 400 K and 150 meV at 200 K (Ando et al., 29 Jan 2025).
Theoretical treatments differ strongly by functional. One hybrid-functional study reported that HSE06+SOC yields a positive indirect gap of 141 meV for the relaxed monolayer, with all tested relaxed geometries remaining positive in the range 100–141 meV, whereas without SOC the monolayer remained gapless semimetallic even within HSE06 (Zheng et al., 2016). A later functional benchmark found PBE gaps of $1T'$7 meV without SOC and $1T'$8 meV with SOC, HSE06+SOC = 69 meV, and MVS+SOC = 81 meV, with the additional claim that MVS yields a positive fundamental gap even without exact exchange, Hubbard $1T'$9, or SOC correction (Yin et al., 2024). The overall pattern is therefore not a single universally agreed number but a robust observation that monolayer $1T$0-WTe$1T$1 lies near a small-gap/topological regime whose precise gap is exceptionally sensitive to experimental conditions and exchange-correlation treatment.
| Study | Probe or method | Reported low-energy result |
|---|---|---|
| (Tang et al., 2017) | ARPES / STS | $1T$2 meV, $1T$3 meV, $1T$4 meV |
| (Cucchi et al., 2018) | Micro-ARPES | $1T$5 meV |
| (Ando et al., 29 Jan 2025) | Temperature-dependent ARPES | 120 meV at 400 K, 150 meV at 200 K |
| (Zheng et al., 2016) | HSE06+SOC | 141 meV indirect gap |
| (Yin et al., 2024) | HSE06+SOC / MVS+SOC | 69 meV / 81 meV |
4. Topological interpretation and edge-state phenomenology
Monolayer $1T$6-WTe$1T$7 is commonly interpreted as a two-dimensional topological insulator or QSH insulator. In the standard mechanism summarized in the literature, the $1T$8 lattice distortion induces a parity-changing band inversion, and SOC then gaps the inverted crossing. In the inversion-symmetric formulation this is expressed through the Fu–Kane criterion
$1T$9
with the distortion changing the topological index from $1T$0 to $1T$1 because the inverted bands have different parity at the relevant time-reversal-invariant point (Tang et al., 2017).
Experimentally, the strongest early evidence combined ARPES, STM, and STS. Polarization-dependent ARPES identified inverted orbital ordering, K-dosed ARPES resolved the SOC-opened bulk gap, and STS on monolayer-island interiors showed a clear bulk gap while spectra at the edges became V-shaped with finite in-gap conductance (Tang et al., 2017). The edge spectral weight was localized near the edge and decayed into the bulk over distances up to about 4 nm, and the edge signal persisted even for irregular or rough edges, which the authors interpreted as consistent with topologically nontrivial edge modes rather than trivial termination-specific resonances (Tang et al., 2017).
Subsequent local probes reinforced that picture. Back-gated STM on monolayer films grown on graphene showed that the edge states span the entire bulk gap: they are maximally localized near midgap, while their localization length increases as their energy approaches the VBM or CBM, where they merge into the bulk continuum (Maximenko et al., 2020). Microwave impedance microscopy on contacted exfoliated monolayers directly imaged conduction along the physical perimeter, along internal tears, and along boundaries between protected and oxidized regions, while the edge conduction remained present as the bulk was gated from $1T$2-type through insulating to $1T$3-type; the edge signal was visible at 77 K and remained observable at 100 K, and it was suppressed by magnetic field in the expected direction (Shi et al., 2018).
The topological classification also evolves with thickness. ARPES plus first-principles calculations on MBE-grown films found $1T$4 for the monolayer, $1T$5 for the bilayer, and $1T$6 again for the trilayer, with the monolayer gap suppressed by increasing thickness and the system evolving toward bulk Weyl-semimetal behavior (Qiao et al., 18 Apr 2026). In that study, the topological diagnosis was made from hybrid Wannier charge-center evolution rather than from a parity product alone (Qiao et al., 18 Apr 2026).
A further transport-level refinement concerns edge orientation. A symmetry-resolved magnetotransport calculation on nanoribbons reported that $1T$7-edge ribbons show clear field-induced spin splitting and strong angular conductance modulation, whereas $1T$8-edge ribbons remain nearly insensitive because nonsymmorphic operations protect degeneracy along $1T$9-$1T'$0 even when time-reversal symmetry is broken (Tapar et al., 24 Jul 2025). This suggests that the helical-edge phenomenology of monolayer $1T'$1-WTe$1T'$2 is intrinsically anisotropic, not merely edge-localized.
5. Electric field, strain, temperature, excitons, superconductivity, and magnetic response
Electrostatic gating does not simply shift the chemical potential through a rigid band structure. In back-gated STM on monolayer $1T'$3-WTe$1T'$4, the maximum gate voltage 80 V across 300 nm SiO$1T'$5 corresponds to an estimated perpendicular field of order $1T'$6, and the measured top-surface gap changes at a rate of $1T'$7, in excellent agreement with the calculated $1T'$8 (Maximenko et al., 2020). The microscopic interpretation is that the field both dopes the film and breaks inversion symmetry, producing Rashba-like spin splitting on the tens-of-meV scale. In the Wannierized model this is written as
$1T'$9
with the true indirect bulk gap reduced for either sign of the field, even though top-surface STM can make the apparent gap evolve asymmetrically because it preferentially probes the top Te layer (Maximenko et al., 2020).
Strain provides another strong tuning axis. STM/STS plus first-principles analysis showed that either compressive strain along $1T'$00 or tensile strain along $1T'$01 can drive monolayer $1T'$02-WTe$1T'$03 from a semimetallic regime into a fully gapped topological insulator while preserving metallic edges (Zhao et al., 2020). Experimentally, the gap closes when
$1T'$04
or when
$1T'$05
and strained insulating islands showed full gaps of about 50 mV or 55 meV, with edge penetration lengths around 3.0–3.6 nm (Zhao et al., 2020). A separate twisted-bilayer theory indicates that even 1% biaxial strain in the constituent monolayers can induce a striped moiré electrostatic landscape with peak-to-trough magnitude $1T'$06 meV, underscoring how sensitive monolayer-derived WTe$1T'$07 bands are to small lattice changes (Magorrian et al., 2024).
Temperature dependence is unusually large and non-rigid. ARPES between 40 K and 400 K showed that on cooling, the H1 band top shifts downward by about 35 meV, H2 by about 110 meV, while the conduction band E1 shifts upward, so the indirect gap increases on cooling (Ando et al., 29 Jan 2025). Comparison with DFT under structural perturbations identified the best descriptor as the W-atom internal distortion parameter $1T'$08, varied in the calculations from 0.11 to 0.26 Å in GGA and 0.12 to 0.28 Å in HSE06 (Ando et al., 29 Jan 2025). This suggests that local tungsten distortion must be included in interpreting the thermal evolution of transport and spectroscopy.
The bulk state near neutrality has also been argued to host strong many-body electron-hole physics. In transport and thermodynamic measurements on exfoliated monolayers, the bulk conductivity developed a sharp V shape on cooling, while the chemical potential showed a step of about 43 meV at neutrality; the authors argued that these features cannot be reproduced by an independent-electron DOS and instead require equilibrium excitons (Sun et al., 2021). In their phenomenology,
$1T'$09
where $1T'$10 is the exciton density, and first-principles calculations gave exciton binding energy larger than 100 meV and an exciton radius as small as 4 nm (Sun et al., 2021). Below about 100 K, more strongly insulating behavior appears and is interpreted as consistent with exciton condensation, although the absence of a conventional CDW is explained by the symmetry of the exciton wave function rather than by a trivial band-insulating picture (Sun et al., 2021).
At high electron doping, monolayer $1T'$11-WTe$1T'$12 also becomes superconducting. A first-principles study of charge-doped monolayers found that superconductivity emerges only on the electron-doped side because specific acoustic phonons soften at momenta $1T'$13, generating strong electron-phonon attraction,
$1T'$14
whereas hole doping slightly stiffens the phonons and does not generate superconductivity (Yang et al., 2020). The predicted dome peaks at
$1T'$15
for
$1T'$16
and the soft $1T'$17-mode becomes unstable above
$1T'$18
suggesting nearby CDW competition (Yang et al., 2020).
6. Vibrational fingerprints, thickness evolution, and derivative structures
Monolayer WTe$1T'$19 is vibrationally distinct from hexagonal $1T'$20 dichalcogenides because its distorted orthorhombic lattice contains tungsten-chain motifs that make it “structurally one-dimensional” in the terminology of the Raman work (Kim et al., 2015). On SiO$1T'$21/Si under 532 nm excitation, a fresh monolayer is identified by only two strong Raman peaks above 100 cm$1T'$22: $1T'$23 near 164–165 cm$1T'$24 and $1T'$25 near 217 cm$1T'$26, while the lower-frequency $1T'$27 and $1T'$28 peaks that are present in thicker samples are strongly suppressed (Kim et al., 2015). The unusual point is that the in-plane $1T'$29 mode changes by only about 1 cm$1T'$30 from bulk to monolayer, whereas most other modes blueshift as the thickness is reduced; the paper attributes the weak shift of $1T'$31 to vibration along the tungsten-chain direction (Kim et al., 2015). The same work also emphasized rapid air degradation of mono- and bilayers, with bilayer Raman intensities dropping to about one third after one hour and vanishing after two hours in ambient conditions (Kim et al., 2015).
Purely first-principles strain-Raman calculations reach similar conclusions about strong anisotropy. For monolayer $1T'$32-WTe$1T'$33, the predicted Raman-active frequencies include $1T'$34, $1T'$35, and $1T'$36, all of which red-shift under tensile strain, with the response strongest under biaxial loading (Yang et al., 2019). The same study predicts phonon instabilities at critical tensile strains of 11.55% along $1T'$37, 7.0% along $1T'$38, and 8.44% under biaxial strain, and identifies an anomalous two-regime strain dependence of the $1T'$39 mode, whose shift rate changes near 5% strain from $1T'$40 to $1T'$41 under biaxial strain (Yang et al., 2019).
Dimensional crossover is equally pronounced in the electronic sector. In thickness-resolved ARPES, the monolayer is gapped, the bilayer gap shrinks to about 30 meV, and by the trilayer the valence band crosses $1T'$42, yielding metallic or semimetallic behavior before the system evolves into the bulk type-II Weyl semimetal (Qiao et al., 18 Apr 2026). That study interpreted the topology as oscillating with thickness, with monolayer $1T'$43, bilayer $1T'$44, and trilayer $1T'$45, driven by interlayer-coupling-induced changes in band crossings (Qiao et al., 18 Apr 2026).
Finally, the monolayer building block can be modified into nonstoichiometric derivatives. STM-tip pulses applied to graphene-covered multilayer $1T'$46-WTe$1T'$47 can remove the topmost Te plane and generate a reconstructed “$1T'$48-layer WTe$1T'$49” with pit depth about 278 pm, a new interior lattice characterized by $1T'$50 Å and $1T'$51 Å, and an energy-dependent unidirectional stripe or CDW state with period about 1.2 nm (Zhao et al., 2024). This derivative is not pristine monolayer $1T'$52-WTe$1T'$53, but it shows that removing a single atomic subplane from the monolayer structural unit can drive qualitatively new reconstruction and charge-order phenomena (Zhao et al., 2024).
