Pith. sign in

REVIEW 2 major objections 5 minor 4 references

Erbium-implanted WS2 flakes with room-temperature photon emission at telecom wavelengths

T0 review · 2 major / 5 minor · reviewed 2026-08-07 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read Erbium ions implanted into WS2 flakes and activated by a 400 °C anneal emit narrow, millisecond-lived, strongly polarized telecom-band photons at room temperature, which the paper attributes to substitutional ErW defects.

desk verdict The telecom PL from Er-implanted WS2 is real and worth publishing, but the paper's claim that Er penetrates 400 nm — and hence that the emitters are bulk substitutional ErW — rests on a fragile thickness-threshold argument that needs direct depth profiling. read the letter →

arxiv 2506.07746 v1 pith:KKHH3TC3 submitted 2025-06-09 cond-mat.mes-hall

Guadalupe García-Arellano , Gabriel I. López Morales , Zav Shotan , Raman Kumar , Ben Murdin , Cyrus E. Dreyer , Carlos A. Meriles This is my paper · ORCID
classification cond-mat.mes-hall
keywords Telecomphotonemissionspinqubitsrare-earthionstwo-dimensionalmaterialstungstendisulfideerbiumimplantationphotoluminescencequantumembedding
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper tries to establish that erbium ions can become telecom-band emitters inside a two-dimensional van der Waals crystal, specifically exfoliated WS2. After ion implantation and a mild 400 °C anneal, Er3+ ions in flakes thicker than roughly 200 nm emit narrow, long-lived photoluminescence lines around 1.5 µm at room temperature, with strong linear polarization and little change from 3.5 K to 300 K. The authors attribute the emission to substitutional ErW defects, erbium occupying a tungsten site, and support that assignment with quantum-embedding calculations that reproduce the activation, the spectral region, the millisecond lifetimes, and the polarization patterns. If correct, the result gives device builders a transfer-ready, two-dimensionally hosted telecom emitter that can be coupled to photonic structures designed separately for photon manipulation.

What carries the argument

The load-bearing object is the substitutional ErW defect: an Er3+ ion replacing a tungsten atom in the WS2 lattice, whose shielded 4f electrons weakly hybridize with the host. The argument is carried by the 4I13/2-to-4I15/2 transition of Er3+, pumped at 980 nm through the 4I11/2 manifold, and by a quantum-embedding effective Hamiltonian with hopping, screened Coulomb, and spin-orbit terms on the Er 4f manifold that yields many-body optical matrix elements. This machinery explains why the isolated Er3+ ion is nearly dark while the defect is bright, and it predicts the measured transition-dependent linear polarization.

What would settle it

Depth-profile an erbium-implanted WS2 flake with secondary-ion mass spectrometry or atom-probe tomography: if the erbium distribution peaks within tens of nanometers of the surface rather than around 400 nm, the bulk-ErW interpretation and the inferred stopping depth would be contradicted.

Watch

Extended reading notes

Core claim

The central claim is that the substitutional ErW defect in WS2 is a bright, narrow-line, room-temperature emitter in the telecom band once activated by annealing after ion implantation. Experimentally, the claim rests on photoluminescence lines from 1300 to 1600 nm with sub-nanometer inhomogeneous linewidths, sites across each flake giving nearly identical spectra, a fluorescence lifetime of about 4.5 ms at room temperature that lengthens to about 8 ms at 3.5 K, and a high degree of linear polarization whose dipole axis rotates by 90 degrees between the 1521 nm and 1540 nm transitions. Quantum-embedding calculations for the neutral ErW defect produce emission in the observed region, radiative lifetimes of 3–10 ms, and transition-dependent linear polarizations, and they show that hybridization with the WS2 lattice, rather than the crystal field of a free ion, activates the 4f–4f transitions.

Load-bearing premise

The conclusion that the emitters are bulk substitutional ErW defects assumes that 75 keV erbium ions penetrate roughly 400 nm into WS2, as inferred from the observation that only flakes thicker than about 200–250 nm emit; if the true stopping depth is much smaller, the fluorescence could come from a different population of erbium ions near the surface or the substrate interface.

Editorial extensions

If this is right

  • Er:WS2 flakes become transfer-ready telecom emitters that operate at room temperature, so photonic devices can be assembled by stacking flakes rather than by growing one material that must serve as both emitter and photon manipulator.
  • The strong, transition-dependent linear polarization means emitted photons can be routed into on-chip waveguides and cavities with predictable orientation, simplifying collection.
  • The weak temperature dependence of brightness and lifetime indicates non-radiative decay channels are inefficient, so the optical interface does not require cryogenic operation.
  • If the emitter is indeed ErW, the low natural abundance of nuclear-spin-active isotopes in WS2 makes the system a candidate for long-lived spin qubits and eventually a spin-photon interface.
  • The same implantation-and-anneal protocol may extend to other rare-earth ions and other two-dimensional hosts, widening the palette of telecom and visible emitters.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The flake-thickness threshold implies a stopping depth near 400 nm for 75 keV erbium, far beyond the 25 nm and 60 nm predictions of common ion-range codes; if a direct depth profile confirms this, stopping-power models for layered van der Waals crystals need revision.
  • A natural next measurement is photon-antibunching or high-resolution excitation spectroscopy on the narrowest lines; determining whether each line is a single Er3+ ion or a small ensemble would sharpen the single-photon and emitter-density picture.
  • Because the embedding calculation predicts that the crystal-field splitting depends on charge state, gating or doping should shift the emission lines; observing such shifts would test the neutral-ErW assignment.
  • The observation that excitation and emission dipoles stay parallel even for transitions with perpendicular emission polarization implies polarization-preserving relaxation within the 4f manifolds; resonant excitation into individual crystal-field levels could test that picture directly.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

2 major / 5 minor

Summary. The manuscript reports telecom-band photoluminescence from Er-implanted exfoliated WS2 flakes after 400 °C annealing in argon. The authors observe sub-nm, often spectrometer-limited emission lines near 1.5 µm, a room-temperature lifetime of 4.5 ± 0.3 ms, weak dependence of the spectra on temperature down to 3.5 K, and a strong, transition-dependent linear polarization. They attribute the emission to neutral substitutional ErW defects, supported by quantum-embedding calculations of the Er 4f manifold that yield telecom-range transitions with mostly linearly polarized dipoles. The paper includes dose, thickness, polarization, lifetime, and cryogenic measurements, with additional details in the Supporting Information.

Significance. If the interpretation holds, this is a valuable step toward 2D-hosted telecom-band emitters for quantum photonic applications: the ms-scale lifetime, narrow ensemble lines, weak temperature dependence, and polarization selection rules are all attractive. The authors provide a careful experimental characterization, including polarization calibration controls, dose series, and cryogenic data, and they present the quantum-embedding model with explicit parameters and documented limitations. The main weakness is that the depth profile of the implanted Er ions is not measured; the inferred ~400 nm penetration depth is load-bearing for the substitutional-ErW assignment, and it currently rests on a thickness threshold that has plausible alternative explanations. The paper is therefore timely and useful, but the site attribution needs stronger support before the central conclusion is accepted.

major comments (2)
  1. [Main text, p. 2–3; SI Section I.c] The claim that 75 keV Er ions penetrate ~400 nm into WS2 is inferred solely from the observation that only flakes thicker than ~200–250 nm luminesce. SRIM and IRADINA predict 25 nm and 60 nm, respectively, and the implantation was intentionally performed at 22° off-normal to avoid channeling, so the required 7–16× discrepancy cannot be attributed to the usual channeling-based underestimation. The cited precedents for SRIM underestimation (Refs. 33–36) are for different ion-target combinations and do not cover this factor. No depth-resolved measurement (RBS, SIMS, atom probe tomography, or cross-sectional analysis) is presented. Alternative explanations for the thickness threshold—such as optical collection volume, thermal coupling to the substrate, strain relaxation, or damage accumulation—are not controlled; the SI itself reports non-fluorescing sections that do not correlate with topography (SI Section I.d). Because the substitutional-ErW assignment and the comparison with a monolayer ErW supercell calculation assume a bulk-like, low-damage host, this missing depth profile is a load-bearing gap. I request a direct depth measurement or, at minimum, an explicit statement that the site assignment is tentative pending such data.
  2. [Main text, p. 4; SI Section II] The quantum-embedding calculations are presented as supporting the ErW assignment, but the comparison with experiment is qualitative: the authors state that a one-to-one correspondence between calculated and measured transitions is presently unwarranted, and the number of predicted lines depends on the assumed thermalization within the 4I13/2 manifold (Fig. 3a versus 3b). The Slater F^k integrals and the spin-orbit coupling are fitted to isolated Er3+ spectra, not to the WS2 data, so the calculation is a reasonable semi-empirical transfer rather than an independent ab-initio prediction of the transition energies. The observed narrow lines, 4.5 ms lifetime, and linear polarization are consistent with Er in a low-symmetry crystalline site, but they do not by themselves exclude near-surface Er ions, interface precipitates, or implantation-damage-related centers. I recommend adding a site-discriminating measurement (for example Zeeman spectroscopy, electron paramagnetic resonance, or atomically resolved elemental analysis) or explicitly presenting ErW as one plausible hypothesis rather than the assignment.
minor comments (5)
  1. [SI Section I.c] The text says the mean penetration depth 'must be comparable to 400 nm,' but the observed quantity is a thickness threshold, not a directly measured range; please distinguish 'mean range' from 'maximum penetration' and provide an uncertainty estimate for the inferred 400 nm value.
  2. [Main text, p. 3; SI Section I.f, Fig. S7] The main text states that there is 'little PL change in brightness, frequency, or lifetime' with temperature, but SI Fig. S7 shows the lifetime increasing from ~4 ms at 300 K to ~8 ms at 3.5 K; this factor of two should be reconciled with the statement or discussed as evidence for some non-radiative channel at room temperature.
  3. [SI Section II.a, Eq. (1)] Equation (1) is garbled in the submitted text, with symbols such as '%&!""' and '12∙45' appearing in place of the intended Hamiltonian and spin-orbit terms; please proofread the equation and define all operators and indices.
  4. [Abstract and Conclusion] The calculations are described as 'ab-initio,' but the Slater integrals and spin-orbit coupling are fitted to isolated Er3+ data; a more precise description would be 'semi-empirical quantum embedding built on DFT-derived crystal-field parameters.'
  5. [SI Section I.e] The statement that no cryogenic spectra could be recorded from the dimmer regions 'pointing to linewidths below the detection capability' is an overinterpretation; reduced signal-to-noise or sample drift could also explain the absence, and this should be acknowledged.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the core experimental observations are direct measurements, and the quantum-embedding parameters are fitted to isolated Er3+ spectra, not to the WS2 photoluminescence data.

full rationale

The central claims—narrow telecom-band PL lines, ~4.5 ms lifetime, strong linear polarization, and weak temperature dependence—are direct experimental measurements and do not derive from the theoretical model. The quantum-embedding calculation is not circular: the authors state that the hopping matrix elements come from DFT Wannierization, and that the Slater U and spin-orbit parameters are 'chosen such that for a calculation of an isolated Er3+, the spectrum best matches the experiment' (SI Section II.a). These parameters are fitted to isolated-Er3+ literature spectra, not to the WS2 emission data, so the predicted telecom-band activity is a genuine transfer rather than a fit to the target. The substitutional ErW site assumption is supported by prior DFT stability studies, including independent references (Refs. 24 and 29) in addition to the authors' own Ref. 31; even the self-cited calculation does not use any of the present PL data, so it is independent supporting evidence rather than a load-bearing self-citation. The paper also explicitly disclaims a one-to-one spectral assignment: 'establishing a one-to-one correspondence between the calculated and measured transitions seems presently unwarranted' (main text). Finally, the implanted-depth inference (≈400 nm from the flake-thickness threshold) is methodologically fragile and conflicts with SRIM/IRADINA, but it is an experimental inference about ion range, not a fitted parameter renamed as a prediction, and it does not make the derivation self-referential. No step in the paper reduces by construction to its own inputs.

Assumptions & free parameters 3 free parameters · 6 assumptions · 1 invented entities

The experiment itself is largely self-contained. The theory introduces semi-empirical free-ion parameters and a specific defect model; the most fragile added assumption is the large implantation depth inferred from flake thickness, which conflicts with standard ion-range simulations.

free parameters (3)
  • Hunds coupling parameter J (Slater F^k integrals) = not given in paper
    In SI Section II.a, J is chosen so the isolated Er3+ spectrum best matches experiment (refs 37-40). It is an input from free-ion spectroscopy, not fitted to the WS2 data.
  • Spin-orbit coupling strength zeta = not given in paper
    In SI Section II.a, only the quadratic part of the spin-orbit operator is retained, with zeta chosen to match isolated Er3+ experimental spectra. Again an external input, not a target fit.
  • Lorentzian broadening R = 0.10 meV
    Chosen in SI Eq. (5) to resolve closely spaced transitions in Figure 3; it is a visualization parameter and does not affect the central claim.
assumptions (6)
  • domain assumption PBE-DFT with PAW pseudopotentials, a 7x7x1 supercell, and a 20 Angstrom vacuum gap adequately describes the ErW defect structure.
    Used in SI Section II.b to obtain relaxed geometries and Wannier functions; no convergence study against larger cells or functionals is reported.
  • domain assumption The active space of 14 Er 4f states with 11 electrons and the Hamiltonian in Eq. (1) captures the relevant optical transitions.
    The model neglects other configurations and assumes 4f-to-4f transitions dominate, stated in SI Section II.
  • domain assumption Free-ion fitted J and zeta transfer unchanged to the ErW defect in WS2.
    The semi-empirical parameters are taken from isolated Er3+ spectroscopy and applied inside the WS2 environment; this transferability is assumed.
  • domain assumption The observed telecom PL originates from 4f-to-4f transitions of Er3+ ions.
    Assigned from spectral range and lifetime, but no direct excitation spectroscopy or isotope-shift test is performed.
  • domain assumption The emitter is the neutral substitutional ErW defect, based on prior DFT stability calculations.
    Main text and SI cite refs 23, 24, 31 for stability; no atomistic verification (STEM, single-emitter spectroscopy) is presented.
  • ad hoc to paper 75 keV Er ions penetrate about 400 nm into WS2 despite SRIM and IRADINA predicting 25 to 60 nm.
    Inferred in SI Section I.c from the flake thickness threshold for PL; needed to explain why only thick flakes emit, but no direct implantation profile is measured.
invented entities (1)
  • Substitutional ErW defect (neutral charge state)
    purpose: Assigned as the microscopic source of the narrow telecom PL lines
    No direct atomic-scale identification is provided; the assignment relies on prior DFT stability and qualitative spectral agreement. Predicted transitions are within the paper, not an external falsifiable handle.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Erbium-implanted WS2 flakes with room-temperature photon emission at telecom wavelengths." pith.science (2026). https://pith.science/paper/KKHH3TC3

@misc{pith2026250607746,
  author       = {Pith},
  title        = {Pith review of: Erbium-implanted WS2 flakes with room-temperature photon emission at telecom wavelengths},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/KKHH3TC3}},
  note         = {Machine review of arXiv:2506.07746}
}
read the original abstract

Optically addressable spin impurities in crystals along with device engineering provide an attractive route to realizing quantum technologies in the solid state, but reconciling disparate emitter and host material constraints for a given target application is often challenging. Rare-earth ions in two-dimensional (2D) materials could mitigate this problem given the atomic-like transitions of the emitters and the versatile nature of van der Waals systems. Here we combine ion implantation, confocal microscopy, and ab-initio calculations to examine the photon emission of Er-doped WS2 flakes. Optical spectroscopy reveals narrow, long-lived photo-luminescence lines in the telecom band, which we activate after low-temperature thermal annealing. Spectroscopic and polarization-selective measurements show a uniform response across the ensemble, while the fluorescence brightness remains mostly unchanged with temperature, suggesting non-radiative relaxation channels are inefficient. Our results create opportunities for novel solid state devices coupling 2D-hosted, telecom-band emitters to photonic heterostructures separately optimized for photon manipulation.

Figures

Figures reproduced from arXiv: 2506.07746 by the authors.

Figure 1
Figure 1. Infrared fluorescence confocal microscopy of exfoliated WS2. (a) Schematic of the experimental setup. Depending on the target experiment, we resort to cw or gated 980-nm lasers for excitation; both lasers are linearly polarized. We use three long-pass filters (one at 1200 nm and two at 1500 nm) as well as a short-pass filter (at 1600 nm) to minimize photon contributions away from the telecom bands. (b) Reflection-mo… view at source ↗
Figure 2
Figure 2. a), which hints at some residual lattice heterogeneity in sections of the crystal with the highest brightness. Flakes exposed to lower Er implantation doses show comparatively more uniform PL maps (as well as improved conversion efficiency upon annealing), possibly because lower ion concentrations prevent Er clustering37,38 (SI, Section I). Optical relaxation in rare-earth ions is typically slow given the shielding … view at source ↗
Figure 4
Figure 4. Polarization dependence of Er3+ photoluminescence. (a) PL images from Flake 3 (1014 ions/cm2 at 75 keV) for varying angle of polarization ) of the excitation beam (left row), or when changing the angle of a linear polarizer prior to PL detection (right row). (b) Excitation and emission polarization plots for four different regions of Flake 3 (colored circles in the PL images); dashed lines in each polar plot are fit… view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

4 extracted references · 4 canonical work pages

  1. [3]

    strength function

    show the room temperature emission linewidth (~12 GHz) is comparable to the spectrometer resolution; notably, no spectra from these regions could be recorded under cryogenic conditions, pointing to linewidths below the detection capability of our spectrometer. The dominant mechanism driving the observed change in inhomogeneous linewidth is presently unkno...

    work page 2004

  2. [4]

    Optically Resolved Exchange Splittings in the Doped Van der Waals Ferromagnet CrBr3:Yb3+

    Polarization dependence of Er3+ photoluminescence. (a) PL images from Flake 3 (1014 ions/cm2 at 75 keV) for varying angle of polarization ) of the excitation beam (left row), or when changing the angle of a linear polarizer prior to PL detection (right row). (b) Excitation and emission polarization plots for four different regions of Flake 3 (colored circ...

    work page Pith review arXiv 2018

  3. [2021]

    Ab initio calculations of erbium crystal field splittings in oxide hosts

    7 C. Borschel, C. Ronning, “Ion beam irradiation of nanostructures A 3D Monte Carlo simulation code”, Nucl. Instrum. Methods Phys. Res. B 269, 2133 (2011). 8 W. Möller, “TRI3DYN, “Collisional computer simulation of the dynamic evolution of 3-dimensional nanostructures under ion irradiation””, Nucl. Instr. Meth. Phys. Res. B 322, 23 (2014). 9 S.O. Kucheyev...

    work page Pith review arXiv 2011

  4. [2632]

    Range parameters study of medium-heavy ions implanted into light substrates

    2 P.L. Grande, F.C. Zawislak, D. Fink, M. Behar, “Range parameters study of medium-heavy ions implanted into light substrates”, Nucl. Instr. Meth, 61 (1991). 3 S. Moll, Y. Zhang, Z. Zhu, P.D. Edmondson, F. Namavar, W.J. Weber, “Comparison between simulated and experimental Au-ion profiles implanted in nanocrystalline ceria”, Nucl. Instr. Meth. Phys. Res. ...

    work page 1991

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.