Surface passivation for narrowing optical linewidth of silicon T centers in nanophotonic devices

Chang-Min Lee; Edo Waks; Fariba Islam; Kyu-Young Kim; Purbita Purkayastha; Sorah Fischer

REVIEW 3 major objections 6 minor 34 references

Surface passivation for narrowing optical linewidth of silicon T centers in nanophotonic devices

T0 review · 3 major / 6 minor · reviewed 2026-08-11 · deepseek-v4-flash

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

Pith's one-line read Al2O3 surface passivation reduces silicon T-center optical linewidth by up to 57%, with 11 nm as the optimal coating thickness.

desk verdict A solid, plausible demonstration that ALD Al2O3 shrinks T-center spectral diffusion, with a real caveat about the missing sham-ALD control; worth refereeing seriously. read the letter →

arxiv 2608.09904 v1 pith:WWSMZRYX submitted 2026-08-10 quant-ph physics.optics

Fariba Islam , Chang-Min Lee , Kyu-Young Kim , Sorah Fischer , Purbita Purkayastha , Edo Waks This is my paper · ORCID

classification quant-phphysics.optics

keywords siliconTcentersspectraldiffusionsurfacepassivationAl2O3atomiclayerdepositionopticallinewidthnanophotonicdevicesholeburningquantumphotonics

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

The paper aims to show that coating the exposed silicon surfaces of T-center nanophotonic devices with a thin Al2O3 layer, deposited by atomic layer deposition, suppresses a substantial part of the spectral diffusion that broadens the emitters' optical lines. In same-emitter before/after measurements, the linewidth narrowed by 32% on average and up to 57%, and thickness studies identify roughly 11 nm as the optimal coating. Complementary measurements show the remaining broadening is still dominated by slow spectral diffusion from charge traps, with the homogeneous linewidth bounded near 75 MHz. If correct, this offers a CMOS-compatible route toward generating indistinguishable photons from T centers, a key step for scalable silicon spin-photon interfaces.

What carries the argument

The central mechanism is the Al2O3 passivation layer formed by atomic layer deposition (ALD) at 150 °C using trimethylaluminum and water. The layer passivates the silicon surface by terminating dangling bonds and establishing a fixed charge at the interface, which stabilizes the fluctuating surface charge states that cause spectral diffusion. The argument is carried by before/after photoluminescence-excitation (PLE) linewidth measurements on the same emitters, by thickness-dependent statistical measurements across ten emitters per thickness, and by two complementary probes: above-bandgap (980 nm) illumination that fills nearby charge traps and narrows the line, and spectral hole burning with two O-band lasers that separates slow spectral diffusion from homogeneous broadening.

What would settle it

Run an unpassivated silicon T-center nanobeam through the identical cryostat-removal and 150°C ALD thermal cycle (with the aluminum precursor omitted); if its linewidth narrows by a comparable ~30%, the claim that Al2O3 causes the narrowing would be refuted.

Watch

Extended reading notes

Core claim

The central discovery is that atomic-layer-deposited Al2O3 acts as an effective surface passivation layer for silicon T centers in nanophotonic waveguides, reducing the optical linewidth of the zero-phonon line by up to 57% (32% on average in same-emitter comparisons). The passivation terminates dangling bonds and provides a fixed charge layer that stabilizes fluctuating surface charges, which are a major source of spectral diffusion since T centers sit 100–200 nm from the surface. Systematic thickness variation shows linewidth narrowing saturates near 11 nm, while thicker films strain the silicon and shorten the excited-state lifetime. Above-bandgap illumination further narrows the line to about 0.3–0.4 GHz, and spectral hole burning places an upper bound of roughly 75 MHz on the homogeneous linewidth, showing that slow spectral diffusion remains the dominant residual broadening.

Load-bearing premise

The causal attribution of the 32% average narrowing to Al2O3 assumes that removing the sample from the cryostat, depositing the coating, and returning to the same location does not itself change the linewidth through thermal cycling, contamination, or re-alignment, since no unpassivated control sample is reported for the same cycle.

Editorial extensions

If this is right

Al2O3 passivation gives a practical, foundry-compatible way to narrow T-center linewidths without relying on post-selection or cavity integration.
An 11-nm Al2O3 layer yields near-saturated linewidth improvement; thicker layers add strain and shorten excited-state lifetimes without further narrowing.
Residual spectral diffusion after passivation is dominated by charge traps not stabilized by the coating, so further gains will require interface engineering (such as HF:HCl pretermination or post-deposition annealing) or electrical charge stabilization.
The roughly 75 MHz upper bound on the homogeneous linewidth sets a target for cavity-enhanced Purcell acceleration to push emission toward the transform limit.

Reading between the lines

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

A natural next test would be a controlled thermal-cycle experiment on an unpassivated sample; if that control narrows comparably, the causal role of Al2O3 would need to be re-examined.
The strain-induced redshift with Al2O3 thickness suggests the coating could serve as a post-fabrication frequency-tuning knob, potentially aligning multiple T centers to a common resonance.
Because above-bandgap light still narrows the line by about a factor of two after passivation, a sizable share of the residual spectral diffusion likely comes from traps in the silicon bulk or near-surface region rather than the outer surface, shifting the focus to etch damage and diode stabilization.
Time-resolved spectral diffusion measurements (repeated PLE or photon-correlation-based linewidth tracking) would directly quantify how the passivation changes the amplitude and timescale of frequency jumps, not just the time-averaged linewidth.

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.

Desk Editor's Note

read the letter

This paper asks whether ALD Al2O3 passivation can narrow the optical linewidth of T centers in nanobeams. The answer it gives—yes, by up to 57%, 32% on average for same-emitter comparisons—is new and worth taking seriously. The central finding is credible.

What's actually new here is the combination: Al2O3 passivation applied to T centers. I don't know of prior work doing that, and the paper acknowledges concurrent related work (Ref 34) without comparing, which is honest but leaves novelty relative to that work partially open. The dataset itself stands on its own.

The paper does several things well. The same-emitter before/after design is the right way to remove emitter-to-emitter scatter, and all four emitters narrowed in the same direction. The thickness sweep (five values, ten emitters per value) shows saturation near 11 nm and a lifetime penalty at greater thickness, which the authors attribute to strain and support with a redshift of the emission wavelength. The 980 nm illumination and spectral hole burning measurements are complementary and give an upper bound of about 75 MHz on the homogeneous linewidth, honestly showing that spectral diffusion remains.

The main soft spot is the absence of a sham-ALD control. The before/after comparison involves taking the sample out of the cryostat, running an ALD cycle, and going back in. Without a sample that goes through the same thermal cycle without Al2O3, the 32% narrowing is not uniquely attributable to the passivation layer. That said, this is not a fatal flaw—the consistency across four emitters and the thickness trend make a pure thermal-cycling artifact unlikely—but it is a missing control that a referee should ask for. The claim of an 11 nm optimum rests on cross-emitter means with no significance test; the error bars are standard deviations, so the saturation is suggestive rather than statistically established. The power-dependence data in Fig. 4(a) come from a single emitter with no repeated measurements, so the factor-of-two narrowing there should be treated as preliminary. No raw data are provided.

Who is this for? People working on T centers, silicon photonics, or color-center spectral diffusion. It is a subfield-relevant result, not a breakthrough, but it gives a practical, CMOS-compatible lever that others will want to try.

I would send it to peer review. The experiments are coherent, the claims are calibrated, and the missing control is addressable. With a sham-ALD control and error bars on the key power-dependence data, this would be a useful contribution; even as is, it deserves referee time rather than a desk reject.

Referee Report

3 major / 6 minor

Summary. The manuscript reports an experimental study of atomic-layer-deposited Al2O3 passivation on silicon T centers integrated in nanobeam waveguides. The authors measure photoluminescence excitation (PLE) linewidths of individual T centers before and after Al2O3 deposition on the same four emitters, observing an average narrowing of 32% and a maximum of 57%. They also vary the Al2O3 thickness from 7 to 26 nm (ten emitters per thickness) and identify 11 nm as optimal, while reporting a thickness-dependent decrease of the excited-state lifetime. Additional experiments with 980 nm above-bandgap illumination show further linewidth narrowing, and spectral hole burning yields an upper bound of approximately 75 MHz on the homogeneous linewidth. The paper concludes that Al2O3 passivation suppresses a substantial component of spectral diffusion and provides a CMOS-compatible path toward indistinguishable photons from T centers.

Significance. If the central claim is established, this work would be a practical and scalable route to reducing spectral diffusion in silicon T-center nanophotonic devices, with the reported ~75 MHz homogeneous-linewidth upper bound being a useful benchmark for future cavity and Purcell-enhanced experiments. The measured narrowing trend is promising, and the authors are appropriately careful to note that residual charge-induced broadening remains. However, the causal attribution of the linewidth narrowing to the Al2O3 passivation layer is not yet rigorously established because the same-emitter comparison lacks a thermal-cycle control, and the optimal-thickness claim rests on cross-emitter statistics without significance testing. The study is therefore a valuable proof-of-principle, but its quantitative headline claims currently outpace the evidence.

major comments (3)

[Results, Figure 2(b)] The central claim that Al2O3 passivation reduces T-center linewidths rests on before/after measurements of four emitters where the sample was removed from the cryostat, heated to 150 °C during ALD, and reinserted. No sham-ALD or thermal-cycle control is reported, so the observed average narrowing of 32% (maximum 57%) cannot be uniquely attributed to the Al2O3 layer; warming and re-cooling alone can reconfigure surface charge traps and spectral diffusion. Because the headline improvement in the abstract and introduction is drawn from this same uncontrolled comparison, this is a load-bearing gap that should be addressed with an unpassivated control undergoing the same thermal cycle.
[Results, Figure 3(a)] The identification of 11 nm as the optimal thickness is based on mean linewidths at five thicknesses with ten emitters each, but no statistical significance test is reported, and the error bars (standard deviations) appear to overlap substantially between neighboring thicknesses such as 11 and 15 nm. Moreover, the thickness series does not include a 0-nm (unpassivated) condition, so the 'linewidth reduction saturates near 11 nm' statement cannot be distinguished from a weak dependence on thickness across the entire passivated range. A significance test (or explicit statement of overlapping confidence intervals) and, ideally, an unpassivated baseline are needed to support the optimality claim.
[Results, Figure 3(b) and Supplementary Section 3] The decrease in excited-state lifetime with increasing Al2O3 thickness is attributed to strain-related enhancement of non-radiative decay, supported by the observed emission redshift. However, no direct structural or strain measurement is provided, and the authors acknowledge that the microscopic mechanism is not identified. Alternative explanations, such as ALD-induced surface damage or additional interface traps, are not excluded. Since the trade-off that makes 11 nm 'optimal' depends on this lifetime reduction, the strain attribution needs either direct evidence or a more cautious wording that treats the lifetime decrease as an empirical trend without a claimed mechanism.

minor comments (6)

[Figure 2 and Figure 3 captions] Figure 2 states that linewidths are obtained from Voigt fits, while Figure 3 states that linewidths come from Lorentzian fits; the choice of fitting function and its possible impact on extracted linewidths should be clarified, especially since the two figures are used jointly to support the narrowing claim.
[Results, Figure 3(a)] The thickness series omits the unpassivated (0 nm) condition, which would make the 'linewidth reduction' visible in the same figure; the authors should either add this data point or explicitly state why it is not included.
[Supplementary Section 2] The autocorrelation confirmation of single emitters is performed on four randomly selected emitters before ALD; the authors should state whether these are the same four emitters used in the before/after comparison of Figure 2(b), as this affects the interpretation of the single-emitter claim.
[Discussion] The sentence comparing the observed improvement to 'linewidth reductions reported for surface-passivated quantum dots' would be more useful with a quantitative comparison rather than a citation-only reference.
[Methods] There is a typographical error in the sentence 'For evaluating the effectiveness the Al2O3 surface passivation' where the word 'of' is missing after 'effectiveness'.
[Supplementary Section 1] In the description of the Hanbury Brown-Twiss setup, the notation 'P = 0.35Pₐ' appears to contain a typo for P_sat; please correct the subscript.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity identified: the central linewidth-narrowing claim is a direct experimental measurement, and self-citations are methods references, not load-bearing inputs.

full rationale

The paper's central claim—that Al2O3 passivation narrows T-center optical linewidths—is established by direct PLE measurements before and after deposition, with Voigt fits to acquired spectra, and by a separate thickness series. The 57% maximum and 32% average reductions are extracted from measured linewidths, not from a fitted model that predicts the same quantity. The spectral-hole-burning bound (150.8 ± 22.0 MHz hole FWHM implying an upper bound of approximately 75 MHz on the homogeneous linewidth) uses the standard low-power relation between hole width and homogeneous linewidth and does not import the passivation conclusion. Self-citations (Refs 1, 8, 9) are used for fabrication details and an autocorrelation analysis procedure, not to justify the passivation result, so they are not load-bearing. The absence of a sham thermal-cycle control is a possible experimental confound for causal attribution, but that is a correctness or external-validity concern, not a circular derivation: no equation or fitted parameter in the paper reduces the headline claim to its own input.

Assumptions & free parameters 0 free parameters · 5 assumptions · 0 invented entities

The experimental claims rest on standard domain assumptions from the silicon photovoltaics and color-center literature; no ad hoc numerical parameters or invented entities are introduced. The most load-bearing assumption is that the measured linewidth changes reflect suppression of surface charge noise rather than strain or measurement artifacts.

assumptions (5)

domain assumption T centers in the measured nanobeams lie 100-200 nm from the silicon surface, so surface charge fluctuations are a significant source of spectral diffusion.
Used to motivate passivation (Fig. 1a); no per-emitter depth measurement is provided.
domain assumption Thin ALD Al2O3 terminates silicon dangling bonds and supplies a fixed charge layer that stabilizes surface charge states.
Borrowed from silicon photovoltaics literature (Refs 19-22); the interface state density is not directly measured here.
domain assumption A spectral hole in the low-power limit has FWHM equal to twice the homogeneous linewidth, and finite pump power only broadens it.
Standard hole-burning model; used to convert the 150.8 MHz hole width into a 75 MHz upper bound.
domain assumption PLE linewidth at 0.1P_sat is dominated by spectral diffusion rather than by power broadening or laser instability.
Needed for interpreting narrowing as suppression of spectral diffusion; no independent laser linewidth or power-broadening calibration is reported.
domain assumption The redshift of emission with thicker Al2O3 is caused by strain, not by charging or dielectric effects.
Authors state 'direct structural measurements are required to identify the microscopic mechanism' (Results), so this is an assumed interpretation.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Surface passivation for narrowing optical linewidth of silicon T centers in nanophotonic devices." pith.science (2026). https://pith.science/paper/WWSMZRYX

@misc{pith2026260809904,
  author       = {Pith},
  title        = {Pith review of: Surface passivation for narrowing optical linewidth of silicon T centers in nanophotonic devices},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/WWSMZRYX}},
  note         = {Machine review of arXiv:2608.09904}
}

read the original abstract

Silicon T centers are promising spin-photon interfaces in solid-state platforms for telecom-compatible, scalable quantum information technologies. A major challenge for T centers in nanophotonics is spectral diffusion, where fluctuations in the local electric-field environment from surface and bulk charge states broaden the optical transition and reduce photon indistinguishability. Strategies that directly suppress spectral diffusion are therefore critical for improving T-center-based quantum photonic devices. Here, we use atomic-layer-deposited Al2O3 to passivate the silicon surface and demonstrate a systematic narrowing of T center optical linewidths. Across our measurements, Al2O3 passivation reduces the T center emission linewidth by up to 57%. Complementary above-bandgap illumination and spectral hole burning measurements show that the remaining linewidth contains a significant spectral-diffusion component caused by adjacent charge traps, while placing an upper bound of approximately 75 MHz on the homogeneous linewidth. This work provides a CMOS-compatible path toward generating indistinguishable photons from silicon T centers for scalable quantum photonic applications.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

34 extracted references · 33 canonical work pages

[1]

Bergeron, L. et al. Silicon-Integrated Telecommunications Photon-Spin Interface. PRX Quantum 1 , 020301 (2020)

work page 2020
[2]

MacQuarrie, E. R. et al. Generating T centres in photonic silicon-on-insulator material by ion implantation. New J. Phys. 23 , 103008 (2021)

work page 2021
[3]

Song, H. et al. Entanglement of a nuclear spin qubit register in silicon photonics. Nat. Nanotechnol. 21 , 53–57 (2026)

work page 2026
[4]

DeAbreu, A. et al. Waveguide-integrated silicon T centres. Opt. Express 31 , 15045–15057 (2023)

work page 2023
[5]

& Chen, S

Johnston, A., Felix-Rendon, U., Wong, Y.-E. & Chen, S. Cavity-coupled telecom atomic source in silicon. Nat. Commun. 15 , 2350 (2024)

work page 2024
[6]

Bowness, C. et al. Laser-induced spectral diffusion and excited-state mixing of silicon T centers. PRX quantum 6 , (2025)

work page 2025
[7]

Zhang, X. et al. Laser-induced spectral diffusion of T centers in silicon nanophotonic devices. PRX quantum 6 , (2025)

work page 2025
[8]

Islam, F. et al. Cavity-enhanced emission from a silicon T center. Nano Lett. 24 , 319–325 (2023)

work page 2023

Show all 34 references

[10]

Higginbottom, D. B. et al. Optical observation of single spins in silicon. Nature 607 , 266–270 (2022)

2022
[11]

Clear, C. et al. Optical-transition parameters of the silicon T center. Phys. Rev. Appl. 22 , (2024)

2024
[12]

Chartrand, C. et al. Highly enriched Si 28 reveals remarkable optical linewidths and fine structure for well-known damage centers. Phys. Rev. B: Condens. Matter Mater. Phys. 98 , 195201 (2018)

2018
[13]

& Wang, H

Lekavicius, I., Oo, T. & Wang, H. Diamond Lamb wave spin-mechanical resonators with optically coherent nitrogen vacancy centers. J. Appl. Phys. 126 , 214301 (2019)

2019
[14]

& Reiserer, A

Früh, J., Salamon, F., Gritsch, A., Ulanowski, A. & Reiserer, A. Spectral stability of cavity-enhanced single-photon emitters in silicon. arXiv [quant-ph] (2026) doi:10.48550/arXiv.2601.13666

2026 doi arXiv
[15]

Wang, C. F. et al. Optical properties of single InAs quantum dots in close proximity to surfaces. Appl. Phys. Lett. 85 , 3423–3425 (2004)

2004
[16]

Yu, X. et al. Engineering diamond interfaces free of dark spins. Phys. Rev. Appl. 25 , (2026)

2026
[17]

Chu, Y. et al. Coherent optical transitions in implanted nitrogen vacancy centers. Nano Lett. 14 , 1982–1986 (2014)

2014
[18]

Zhao, J. et al. Deterministic resonance fluorescence improvement of single quantum dots by optimized surface passivation. Light Sci. Appl. 14 , 170 (2025)

2025
[19]

& Takagi, S

Kim, Y., Han, J., Takenaka, M. & Takagi, S. Low temperature Al2O3 surface passivation for carrier-injection SiGe optical modulator. Opt. Express 22 , 7458–7464 (2014)

2014
[20]

& Kessels, W

Dingemans, G. & Kessels, W. M. M. Status and prospects of Al2O3-based surface passivation schemes for silicon solar cells. J. Vac. Sci. Technol. A 30 , 040802 (2012)

2012
[21]

Hoex, B., Schmidt, J., Pohl, P., van de Sanden, M. C. M. & Kessels, W. M. M. Silicon surface passivation by atomic layer deposited Al2O3. J. Appl. Phys. 104 , 044903 (2008)

2008
[22]

Benick, J. et al. High efficiency n-type Si solar cells on Al2O3-passivated boron emitters. Appl. Phys. Lett. 92 , 253504 (2008)

2008
[23]

Hughes, E. T. et al. Dislocation ‐ induced structural and luminescence degradation in InAs quantum dot emitters on silicon. Phys. Status Solidi (A) 220 , (2023)

2023
[24]

Langer, T. et al. Strain-induced defects as nonradiative recombination centers in green-emitting GaInN/GaN quantum well structures. Appl. Phys. Lett. 103 , 022108 (2013)

2013
[25]

Defects in epitaxial multilayers I

Matthews, J. Defects in epitaxial multilayers I. Misfit dislocations. J. Cryst. Growth 27 , 118–125 (1974)

1974
[26]

J., Xu, J

Wu, F., Smart, T. J., Xu, J. & Ping, Y. Carrier recombination mechanism at defects in wide band gap two-dimensional materials from first principles. Phys. Rev. B. 100 , (2019)

2019
[27]

Alizadehherfati, A. et al. Electrical control of optically active single spin qubits in ZnSe. arXiv [quant-ph] (2025) doi:10.48550/arXiv.2512.21462

2025 doi
[28]

Majumdar, A., Kim, E. D. & Vučković, J. Effect of photogenerated carriers on the spectral diffusion of a quantum dot coupled to a photonic crystal cavity. Phys. Rev. B Condens. Matter Mater. Phys. 84 , (2011)

2011
[29]

Grant, N. E. et al. Activation of Al2O3 surface passivation of silicon: Separating bulk and surface effects. Appl. Surf. Sci. 645 , 158786 (2024)

2024
[30]

Dingemans, G., Seguin, R., Engelhart, P., Sanden, M. C. M. van de & Kessels, W. M. M. Silicon surface passivation by ultrathin Al 2 O 3 films synthesized by thermal and plasma atomic layer deposition. Phys. Status Solidi Rapid Res. Lett. 4 , 10–12 (2010)

2010
[31]

Kasperczyk, M. et al. Statistically modeling optical linewidths of nitrogen vacancy centers in microstructures. Phys. Rev. B. 102 , (2020)

2020
[32]

Pedersen, F. T. et al. Near transform-limited quantum dot linewidths in a broadband photonic crystal waveguide. ACS Photonics 7 , 2343–2349 (2020)

2020
[33]

Somaschi, N. et al. Near-optimal single-photon sources in the solid state. Nat. Photonics 10 , 340–345 (2016)

2016
[34]

& Chen, S

Johnston, A., Wong, Y.-E., Yan, S., Felix-Rendon, U. & Chen, S. Optical linewidth narrowing for device-coupled single T centers. arXiv [quant-ph] (2026) doi:10.48550/arXiv.2607.28485. Supplementary Information for Surface passivation for narrowing optical linewidth of silicon ...

2026 doi arXiv
[35]

Lee, C.-M. et al. High-efficiency single photon emission from a silicon T-center in a nanobeam. ACS Photonics 10 , 3844–3849 (2023)

2023

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.

[1] [1]

Bergeron, L. et al. Silicon-Integrated Telecommunications Photon-Spin Interface. PRX Quantum 1 , 020301 (2020)

work page 2020

[2] [2]

MacQuarrie, E. R. et al. Generating T centres in photonic silicon-on-insulator material by ion implantation. New J. Phys. 23 , 103008 (2021)

work page 2021

[3] [3]

Song, H. et al. Entanglement of a nuclear spin qubit register in silicon photonics. Nat. Nanotechnol. 21 , 53–57 (2026)

work page 2026

[4] [4]

DeAbreu, A. et al. Waveguide-integrated silicon T centres. Opt. Express 31 , 15045–15057 (2023)

work page 2023

[5] [5]

& Chen, S

Johnston, A., Felix-Rendon, U., Wong, Y.-E. & Chen, S. Cavity-coupled telecom atomic source in silicon. Nat. Commun. 15 , 2350 (2024)

work page 2024

[6] [6]

Bowness, C. et al. Laser-induced spectral diffusion and excited-state mixing of silicon T centers. PRX quantum 6 , (2025)

work page 2025

[7] [7]

Zhang, X. et al. Laser-induced spectral diffusion of T centers in silicon nanophotonic devices. PRX quantum 6 , (2025)

work page 2025

[8] [8]

Islam, F. et al. Cavity-enhanced emission from a silicon T center. Nano Lett. 24 , 319–325 (2023)

work page 2023