Three compounds with potassium, rubidium or cesium between WTe2 layers showed a different crystal symmetry and similar direct optical gaps. They were nominally K0.5WTe2, Rb0.5WTe2 and Cs0.5WTe2. The work’s electrical result came from Rb0.5WTe2 crystals, which the paper describes as showing semiconducting behavior.
Here, intercalation means placing atoms in the spaces between sheets of WTe2. The researchers combined powder X-ray diffraction, which reveals structure from the pattern made by a powdered sample, with diffuse-reflectance infrared Fourier-transform spectroscopy, temperature-dependent conductivity measurements and density functional theory, or DFT, calculations of electronic states. The K, Rb and Cs structure solutions were refined with the Rietveld method.
A different symmetry, wider spaces between layers
Compared with pristine WTe2’s orthorhombic Pmn21 arrangement, the intercalated K, Rb and Cs phases were assigned monoclinic P21/m symmetry. The individual WTe2 layers remained largely unchanged, although they showed slight distortions.
The change is also visible in the distance between layers. The relevant PXRD spacing was 7.009 angstroms for WTe2, rising to 8.6297 angstroms for K0.5WTe2, 8.9179 angstroms for Rb0.5WTe2 and 9.2606 angstroms for Cs0.5WTe2. The spacing increased across the series as the cation became larger.
Elemental checks supported the intended formula for the potassium and rubidium products. ICP-OES and RFA measurements gave an average A:W:Te ratio of 0.50(3):1:1.98(3), consistent with the nominal A0.5WTe2 composition.
The main powder synthesis used a 0.2 M alkali naphthalenide solution in anhydrous THF with a 1.5 molar excess, held at -50 °C for 3.5 hours. Single crystals were held at -80 °C for five weeks.
The clearest electrical signal came from rubidium
The reported electrical measurements centered on Rb0.5WTe2 crystals. The paper describes ohmic current-voltage behavior from 75 to 300 K, as well as semiconducting current-voltage behavior. Conductivity reached a maximum reported value of 24.3 mS/m at 300 K.
An activation-style fit over 150 to 300 K yielded a 0.476 eV activation barrier. Results below 125 K were limited by instrumental resolution.
That makes the transport finding phase-specific. Comparisons among K, Rb and Cs instead combine structural, optical and computational evidence, so the reported electrical result should not be read as a transport result for the full series.
Optical data and calculations point to different kinds of gaps
The study’s absorption-edge analysis estimated direct optical gaps of 0.431 eV for K0.5WTe2, 0.448 eV for Rb0.5WTe2 and 0.440 eV for Cs0.5WTe2. A direct gap here is an optical estimate of the energy separation associated with light absorption. The absorption edge was broad, and attempts to extract an indirect gap were unreliable because the result depended on the model used.
DFT offered a related prediction. It classified all three compounds as narrow-gap indirect semiconductors, with gaps from 0.35 to 0.4 eV and minimum direct gaps of about 0.5 eV. These are computational predictions rather than direct measurements, and the paper notes known limitations in DFT band-gap values.
The potassium compound prompted a separate computational explanation. In cation-free P21/m WTe2 calculations, the neutral model was a semimetal with a negative band gap, while a model carrying a -4 charge per unit cell was a zero-gap semimetal. The authors suggested charged vacancies as one possible explanation for the very small K0.5WTe2 gap, but the vacancies were not directly characterized experimentally.
The calculations also predicted strong sensitivity to interlayer strain in the K phase. At 2% compressive strain, the model classified the material as metallic; at 2% tensile strain, it classified the material as a zero-band-gap semimetal. This prediction was not experimentally tested.
Moisture complicates the picture
In a time-resolved experiment on Rb0.5WTe2, water co-intercalation came before hydrolysis and declining long-range order. The interlayer gallery expanded by about 1.91 angstroms, or roughly 21%, during the process.
The structural evidence carries an important limitation: single-crystal diffraction was not feasible for these moisture-sensitive compounds. The reported K, Rb and Cs structures therefore came from PXRD structure solutions and Rietveld refinements, supporting the assigned symmetry without providing a definitive single-crystal structure.
A useful materials result, with key questions unresolved
The clearest finding is structural. Across the K, Rb and Cs compounds, the study reports a lower-symmetry lattice, progressively wider spacing between layers and similar direct optical-gap estimates. The electronic picture is more limited because the reported transport result is for the Rb phase, while the wider comparison relies on optical and computational data.
Further work would need to test whether charged vacancies account for the very small K-phase gap, whether the predicted strain response can be reproduced experimentally and how the compounds can be stabilized against moisture.
Paper data and sources
Original title: Intercalation of Alkali Metal into WTe2, the Crystal Structure of A0.5WTe2 and Observation of a Metal-to-Semiconductor Transition
Authors: Patrick Schmidt, Fabian Strauß, Marcus Scheele et al.
Journal/Repository: Dalton Trans. (2026) 55 (8), 3296-3309
Status: Peer-reviewed
First online: 2026-08-25
DOI: 10.1039/d5dt02775f
Original paper · Full text