The Role of Defect Geometry in Localized Emission from Monolayer Tungsten Dichalcogenides

Understanding the mechanism of single photon emission (SPE) in two-dimensional (2D) material is an unsolved problem important for quantum optical materials and the development of quantum information applications. In 2D transition metal dichalcogenides (TMDs) such as tungsten diselenide (WSe2), quantum emission has been broadly attributed to exciton localization from atomic point defects, yet the precise microscopic origins are not fully understood. This work introduces an empirically grounded computational framework that explains both the origins of facile SPE in WSe2 and its relative scarcity in related TMD, tungsten disulfide. High resolution microscopy identifies native defect geometries existing in monolayer WSe2 lattices providing the ingredients necessary to build a realistic model. The qualitative effects of chalcogen type, defect geometry, and mechanical strain on the electronic structure are then individually assessed using density functional theory, from which a specific divacancy configuration emerges as the candidate for localized single-electron transitions that match observed spectral energies. Spectroscopy and photon correlation measurements further validate this model, establishing a self-consistent link between defect geometry, electronic structure, and quantum emission. By isolating the distinct roles of chalcogen type, defect configuration, and mechanical strain, this work provides a thorough investigation of exciton localization and optical behavior, contributing to a clearer picture of the physical drivers of single photon emission in tungsten-based TMDs.

💡 Research Summary

**
The paper tackles the long‑standing question of why monolayer tungsten diselenide (WSe₂) so readily exhibits single‑photon emission (SPE) while related transition‑metal dichalcogenides (TMDs) such as WS₂, MoSe₂, and MoS₂ do not. The authors combine three complementary approaches—high‑resolution scanning transmission electron microscopy (HAADF‑STEM), density‑functional theory (DFT) with the hybrid HSE functional, and low‑temperature photoluminescence (PL) together with photon‑correlation (g²) measurements—to build a self‑consistent picture linking defect geometry, electronic structure, and optical response.

First, HAADF‑STEM imaging of exfoliated WSe₂ monolayers (electron beam ≤ 80 kV to avoid damage) reveals three dominant point‑defect motifs: a single selenium vacancy (V1), a vertical divacancy (V2) where a selenium atom is missing from both the top and bottom chalcogen layers at the same lattice site, and a lateral divacancy (V‑V) consisting of two neighboring vacancies in the same layer. The vertical divacancy appears with a density of ~0.17 nm⁻², consistent with prior reports, confirming that V2 is a native, statistically significant defect.

Second, the authors perform systematic DFT‑HSE calculations for each defect type in both WSe₂ and WS₂, with and without modest biaxial strain (±1 %). The pristine WSe₂ monolayer shows a direct band gap at the K point. The V1 defect introduces two degenerate mid‑gap states, while the lateral V‑V creates four mid‑gap levels. Both of these configurations leave the conduction band largely untouched, implying that exciton recombination would have to proceed via indirect relaxation pathways that are inefficient for SPE.

In stark contrast, the vertical divacancy V2 generates two true mid‑gap states and two additional states that hybridize strongly with the conduction band at K. This hybridization pulls a defect‑derived band down into the gap, creating a transition pathway directly from the valence‑band maximum (VBM) to a mixed conduction‑band/defect state. The calculated transition energy (~1.7 eV) aligns closely with the experimentally observed SPE range in WSe₂ (1.5–1.7 eV). Strain further tunes this gap: 1 % tensile strain widens the VBM‑defect separation, while compressive strain narrows it, offering a mechanism for the experimentally reported strain‑dependent red‑shifts of SPE lines.

Electron localization functions (ELF) corroborate the electronic picture: V2 exhibits the highest ELF peak (0.57) among the three defects, indicating stronger confinement of charge density around the defect core. This is consistent with a more localized exciton that can emit a single photon rather than a delocalized, multi‑photon recombination.

Formation‑energy calculations show that while the single vacancy V1 has the lowest formation energy (≈ 2.8 eV), the vertical divacancy V2 is only modestly higher (≈ 5.0 eV). Given the high overall defect density in exfoliated monolayers (10¹²–10¹³ cm⁻²), the probability of two neighboring vacancies coalescing into a V2 is appreciable, especially because WSe₂ exhibits the highest single‑vacancy binding energy among TMDs, favoring divacancy formation.

Third, the optical validation: low‑temperature PL spectra taken on regions identified by STEM as containing V2 defects display sharp, narrow lines at ~1.6 eV. Second‑order correlation measurements yield g²(0) < 0.5, confirming true single‑photon emission. Areas dominated by V1 or V‑V defects either show broader, multi‑photon features or no distinct SPE peaks, reinforcing the computational prediction that only V2 provides the necessary electronic structure for SPE.

Finally, the authors extend the DFT analysis to WS₂. In WS₂, both the formation energies and ELF values for V2 are higher, and the hybridization with the conduction band is weaker, explaining why SPE is rarely observed without deliberate strain or defect engineering. This comparative study isolates the role of the chalcogen atom (Se vs S) and demonstrates that the larger, more polarizable selenium orbitals facilitate stronger defect‑band mixing.

In summary, the paper delivers a comprehensive, experimentally anchored theoretical framework that identifies the vertical selenium divacancy (V2) as the primary microscopic origin of SPE in monolayer WSe₂. It clarifies how defect geometry, chalcogen chemistry, and mechanical strain each independently modulate the electronic structure and, consequently, the optical response. The work not only resolves a key puzzle in 2D quantum optics but also provides actionable guidance for defect‑engineered quantum emitters: targeting V2 formation through controlled growth or post‑processing, and applying calibrated strain to fine‑tune emission energies. This integrated methodology sets a benchmark for future studies of quantum light sources in layered materials.