Pith. sign in

REVIEW 4 major objections 6 minor 3 references

All-optical diode via nonreciprocal nonlinear absorption and interfacial charge transfer in two-dimensional van der Waals heterostructures

T0 review · 4 major / 6 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read A passive all-optical diode made of stacked NbC and GaS two-dimensional crystals shows a 1.73 dB nonreciprocity factor and 0.81 transmittance symmetry, with simulation and experiment agreeing only when interfacial charge transfer is…

desk verdict Solid new material system for a passive all-optical diode, but the interfacial charge-transfer mechanism is over-claimed and the supporting numbers are internally inconsistent. read the letter →

arxiv 2505.24678 v1 pith:LZ4PEH6M submitted 2025-05-30 physics.optics

Erkang Li , Jinhong Liu , Yanqing Ge , Mingjian Shi , Yijie Wang , Chunhui Lu , Yixuan Zhou , Xinlong Xu This is my paper
classification physics.optics PACS 42.65.-k42.79.-e
keywords all-opticaldiodenonreciprocalnonlinearabsorptionsaturablereverseinterfacialchargetransfervanderWaalsheterostructureNbCMXeneZ-scan
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

An all-optical diode is a component that lets light pass in one direction more easily than the reverse direction, with no moving parts and no external power. This paper claims to have built a passive one by stacking two two-dimensional crystals: NbC, a metallic MXene that absorbs less light as intensity rises (saturable absorption, SA), and GaS, a wide-gap chalcogenide that absorbs more (reverse saturable absorption, RSA). In the measured device, forward-incident light sees an SA response and reverse-incident light sees an RSA response, giving a nonreciprocity factor of 1.73 dB and a transmittance symmetry of 0.81. The authors further claim that a simulation of the two layers alone cannot match the experiment, and that adding a rate-equation model of interfacial charge transfer between the crystals is required to reproduce the measured numbers. If right, this is a simple, scalable, CMOS-compatible route to all-optical diodes for optical signal processing.

What carries the argument

The device works by axial asymmetry in nonlinear absorption: pairing a SA layer with a RSA layer makes the stack's transmittance rise with intensity in one propagation direction and fall in the other. The argument is carried by three coupled propagation equations — Eq. (1) for the SA layer (NbC), Eq. (2) for the RSA layer (GaS), and Eq. (3)/(S31) for the interface, the last derived from the four-level rate equations (S25–S30) of a Schottky heterojunction and containing the two-photon, excited-state, ground-state, and excited-state absorption cross sections $\sigma_{\rm TPA}$, $\sigma_{12}$, $\sigma_s$, and $\sigma_e$. A fourth-order Runge–Kutta scheme integrates these equations slice by slice, with 1 nm steps, in forward and then reverse order to produce the diode curves. Scanning the parameters shows that the linear absorption coefficients mostly set the nonreciprocity factor, the nonlinear coefficients $\beta_{\rm SA}$ and $\beta_{\rm RSA}$ set the transmittance symmetry, and among the interface terms $\sigma_e$ has the dominant effect; these are the knobs the paper proposes for designing future 2D all-optical diodes.

What would settle it

Insert an inert spacer (a few nanometres of hBN or oxide) between the NbC and GaS films and re-measure the forward and reverse Z-scan responses at 280 GW/cm²: if $F\approx1.73$ dB and $S\approx0.81$ survive with the interface broken, the interfacial charge-transfer term is not load-bearing. Alternatively, independently measure $\sigma_s$ and $\sigma_e$ by transient absorption on the isolated NbC film and check whether the fitted interface values are physically real rather than free parameters.

Watch

Extended reading notes

Core claim

The central claim is that a NbC/GaS van der Waals heterostructure is an all-optical diode whose directionality comes from opposite nonlinear absorption in the two layers: light entering from the NbC side first bleaches the saturable absorber and exits more intense than it otherwise would (forward bias), while light entering from the GaS side first encounters the reverse saturable absorber and exits less intense (reverse bias), mimicking the forward and reverse bias of an electric diode. Femtosecond Z-scan measurements at 800 nm under a 280 GW/cm² pump show exactly this asymmetry, with a nonreciprocity factor $F = 10\log_{10}(T_{\rm forward}/T_{\rm reverse})$ of 1.73 dB and a transmittance symmetry $S$ of 0.81 that the paper takes as near the ideal on–off balance. The authors state that solving the propagation equations for the two films alone reproduces the main diode shape but with poor symmetry, and that only after inserting an interface propagation equation — derived from a four-level rate-equation model of the NbC/GaS Schottky junction, with interfacial ground- and excited-state cross sections $\sigma_s = 1.2\times10^{-16}\ \mathrm{cm}^2$ and $\sigma_e = 1.6\times10^{-18}\ \mathrm{cm}^2$ — does the simulation match the experiment with the same F and S. Their conclusion is that nonreciprocal nonlinear absorption and interfacial charge transfer together drive the diode response.

Load-bearing premise

The whole mechanism story rests on the four-level rate-equation model of the NbC/GaS interface, whose two absorption cross sections $\sigma_s$ and $\sigma_e$ were chosen as 'best values' so the simulation reproduces the measured 1.73 dB and 0.81; if that interface model is not the right physics, the raw diode effect could still be real but the claimed essential role of interfacial charge transfer would be unsupported.

Editorial extensions

If this is right

  • Larger |$\beta_{\rm SA}$| and $\beta_{\rm RSA}$ nonlinear absorption coefficients directly improve the nonreciprocity factor $F$, so the design rule is to pair the strongest available SA and RSA 2D materials.
  • Transmittance symmetry $S$ is set by the nonlinear absorption coefficients, not by the linear absorption coefficients, which mainly control $F$.
  • Because $\sigma_e$ mostly sets the interfacial contribution, engineering the interface charge transfer of the heterostructure is a separate tuning knob for diode performance beyond the choice of SA and RSA films.
  • The demonstrated NbC/GaS diode is passive and compatible with planar CMOS fabrication, so it can be integrated into on-chip optical signal processing without external magnetic or acoustic fields.
  • The simulation roadmap in Figures 2–3 gives quantitative guidance — target values of $\beta_{\rm SA}$, $\beta_{\rm RSA}$, $\sigma_s$, and $\sigma_e$ — for optimizing future all-optical diodes from 2D van der Waals stacks.

Reading between the lines

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

  • The SA/RSA axial-asymmetry recipe is not specific to NbC and GaS: any pair of complementary 2D nonlinear absorbers with a charge-transfer interface should produce a diode, so the demonstrated F and S values are likely a lower bound for what this material family could reach with stronger absorbers.
  • The claimed interface sensitivity of $\sigma_e$ could be tested directly: deliberate control of the interface — stacking order, twist angle, or an inserted spacer layer — should move F and S in the direction predicted by Eq. (3).
  • Because F and S are pump-intensity dependent, the same stack may double as a tunable, even switchable, nonreciprocal element if driven below and above the intensity threshold near 8.4 GW/cm² that the authors identify as the analog of a knee voltage.
  • The diode's operating bandwidth may extend beyond 800 nm: NbC is a broadband absorber and GaS's wide bandgap keeps two-photon absorption available across much of the near infrared, although the paper only demonstrates 800 nm.
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

4 major / 6 minor

Summary. The paper reports an all-optical diode based on a two-dimensional NbC/GaS van der Waals heterostructure. Open-aperture Z-scan measurements show saturable absorption in NbC and reverse saturable absorption in GaS, with nonlinear coefficients of about -282 cm/GW and 812 cm/GW, respectively. For the heterostructure, forward and reverse incidence produce opposite nonlinear responses, yielding a nonreciprocity factor F = 1.73 dB and transmittance symmetry S = 0.81 at a pump intensity of 280 GW/cm^2. The authors model the device by sequentially solving propagation equations for the SA layer, the interface, and the RSA layer, and they argue that agreement with experiment requires including an interfacial charge-transfer term described by a four-level rate-equation model with fitted absorption cross sections sigma_s and sigma_e. The paper concludes that both nonreciprocal nonlinear absorption and interfacial charge transfer are responsible for the diode action.

Significance. If the central mechanistic claim were fully supported, the work would be a useful step toward passive, CMOS-compatible all-optical diodes based on 2D materials, and the reported nonlinear coefficients of NbC and GaS together with the comparative tables would be a valuable reference for the nonlinear optics community. The experimental diode response itself is plausible and the individual Z-scan fits are broadly consistent. However, the claim that interfacial charge transfer is essential to the diode response currently rests on a fitted interface model whose parameters are not independently constrained, and there are internal inconsistencies in the reported cross sections. The significance is therefore contingent on additional experimental or modeling evidence that directly probes the interface contribution.

major comments (4)
  1. [§6, Fig. 6(d), Eq. (3)/(S31)] The interfacial charge-transfer mechanism is not established because the interface parameters sigma_s and sigma_e are chosen as "best values" to reproduce the measured F = 1.73 dB and S = 0.81, and the same two scalar outputs are then cited as evidence of agreement. This is circular: matching two scalar outputs with an interface equation containing at least six parameters (sigma_TPA, sigma_12, sigma_s, sigma_e, and the lifetimes entering I1, I2, I3) cannot identify the microscopic mechanism. The authors should provide an independent constraint on the interface parameters, for example transient absorption measurements, thickness-dependent studies, or independently measured cross sections, before claiming that interfacial charge transfer is responsible for the improved symmetry.
  2. [§6 vs. Fig. 2 caption] The interface cross sections used in the main simulation are not stable within the manuscript. The Fig. 2 caption lists invariant parameters sigma_s = 1.0 x 10^-16 cm^2 and sigma_e = 1.4 x 10^-18 cm^2, whereas §6 reports "best values" sigma_s = 1.2 x 10^-16 cm^2 and sigma_e = 1.6 x 10^-18 cm^2 for the same interface model. This discrepancy indicates that the interface parameters are not uniquely determined by the data and undermines the quantitative match claimed for Fig. 6(d).
  3. [§5 vs. Table S2] The GaS excited-state absorption cross section is inconsistent across the manuscript. The main text states that fitting the 375 GW/cm^2 Z-scan data with Eq. (7) gives sigma_12 = 1.9 x 10^-18 cm^2, while Table S2 lists sigma_12 = 4.07 x 10^-17 cm^2 for the same on-focus intensity, a discrepancy of more than a factor of 20. Since sigma_12 enters the interface propagation equation Eq. (3)/(S31), the model inputs are not reliably grounded in the measured GaS response.
  4. [§6, Fig. 6(c) and 6(d)] The no-interface simulation already captures the main shape of the diode response, and only a "notable difference regarding the transmittance symmetry" is attributed to the interface. No quantitative goodness-of-fit metric, uncertainty estimate, or control experiment is provided to show that this residual is specifically caused by interfacial charge transfer rather than by sample thickness variations, intensity calibration, or other systematic effects. The claim that the interface has an "essential effect" on the diode response therefore goes beyond what the current comparison demonstrates.
minor comments (6)
  1. [Eq. (3)] Equation (3) is garbled in the typeset text, with symbols and fractions not rendered cleanly; it should be reset to match Eq. (S31) in the Supporting Information.
  2. [§4, Fig. 4(d)] The statement that the SA beta of NbC is "2-5 orders in magnitude larger" than other materials should specify whether this refers to one particular reference condition, since the comparisons in Table 1 span different pulse durations and excitation intensities.
  3. [§2, definition of S] The transmittance symmetry factor S is defined only through a formula in the text, and several reported values exceed 1 (e.g., S = 56.4 in Fig. 2(f) and S = 1.15 in Fig. 3(b)). The normalization and allowed range of S should be stated explicitly, and the definition should be typeset clearly.
  4. [Fig. 6(b) and 6(d)] The comparison between experiment and simulation is presented as a match with "the same performances of F and S," but no error bars or fitting residuals are shown. A quantitative comparison with uncertainty would strengthen the claim.
  5. [Table S2] The units of sigma_TPA are given as cm^4/GW, which is unusual for a two-photon absorption cross section; please verify the definition and units, and ensure consistency with the values reported in the main text.
  6. [References] Reference [7] is incomplete: it lacks the journal name and volume/page information. Please complete the reference.

Circularity Check

1 steps flagged · score 6.0 of 10

Interfacial charge-transfer mechanism is partially circular: σs and σe are fitted to the measured F/S, then the same F/S are quoted as the simulation's match.

  1. fitted input called prediction [Main text, Section 2 (Results and Discussion), Figure 6(c)-(d) and final paragraph; Eq. (3) with Supporting Eq. (S31).]
    "The σs of 1.2×10-16 cm2 and σe of 1.6×10-18 cm2 at the interface of van der Waals NbC/GaS heterostructures are the best values in the simulation in Figure 6(d). ... both the experiment (Figure 6(b)) and the simulation (Figure 6(d)) matches well with each other, which exhibits the all-optical diode behavior with the same performances of F (1.73 dB) and S (0.81) as that of the experimental result in Figure 6(b)."

    The interface simulation is accepted because σs and σe were chosen as 'best values' so that Eq. (3)/(S31) reproduces the measured F=1.73 dB and S=0.81, and the same F and S are then quoted as the match. The agreement is therefore a post-hoc fit, not an independent validation of interfacial charge transfer. Since the no-interface simulation already 'catches the main features' and only the transmittance-symmetry residual is assigned to the interface, and Eq. (S31) contains at least six interface parameters, matching two scalar outputs cannot identify the mechanism. The fitted inputs are also unstable within the paper: Fig.

full rationale

The measured all-optical diode data and the SA/RSA material parameters are independent experimental results, so the core nonreciprocal-absorption account is not circular. The self-contained simulation without an interface (Fig. 6c) already reproduces the main diode shape using measured β values, so that portion is genuinely tested. The circularity is confined to the interfacial charge-transfer claim: the interface model's σs and σe are selected as 'best values' to reproduce the measured F and S, and the reproduced F/S are then presented as the simulation-experiment agreement of Fig. 6(d). This is fitted-input-called-prediction: the target outputs set the model inputs, so the agreement cannot independently establish that interfacial charge transfer is essential. Internal parameter inconsistencies (σs/σe differ between Fig. 2 caption and final values; σ12 differs between text and Table S2) further indicate the interface parameters are not pinned by independent data. No load-bearing self-citation or imported uniqueness theorem was found. Overall score 6: the paper is partially circular because the central mechanistic claim about interfacial charge transfer reduces to a fit, while the raw diode measurement and material nonlinearities remain valid.

Assumptions & free parameters 9 free parameters · 6 assumptions · 0 invented entities

No new particles, forces, dimensions, or conserved quantities are introduced. The interface charge transfer is modeled using established concepts (SA, RSA, Schottky junction) and does not constitute an invented entity. The ledger instead shows many fitted parameters, the most important being the interface cross sections sigma_s and sigma_e that are chosen after the fact to reproduce the diode metrics.

free parameters (9)
  • beta_SA (NbC nonlinear absorption coefficient) = -350 to -300 cm/GW; -282 cm/GW used in diode simulation
    Extracted by fitting open-aperture Z-scan traces of the NbC film with Equation (5), then used directly in the diode simulations (Fig. 6c,d).
  • beta_RSA (GaS nonlinear absorption coefficient) = 752 to 831 cm/GW; 812 cm/GW used in diode simulation
    Extracted from Z-scan fits of the GaS film and used in the diode model.
  • sigma_s (ground-state absorption cross section of NbC at the interface) = 1.2e-16 cm2
    Selected as the 'best value' at the NbC/GaS interface so that the simulation reproduces the measured F = 1.73 dB and S = 0.81 in Figure 6(d).
  • sigma_e (excited-state absorption cross section of NbC at the interface) = 1.6e-18 cm2
    Also chosen by hand in the interface simulation to match the experimental diode response.
  • kappa (sigma_e/sigma_s for NbC film) = 0.28 to 0.72 depending on intensity
    Fitted from the three-level SA model to Z-scan data; used to compute the SA transmittance curve in Equation (4).
  • I_s (saturable intensity of NbC) = about 150 GW/cm2; 148-154 GW/cm2 across fits
    Fitted from the Z-scan transmittance versus intensity; needed to reproduce the SA curve.
  • sigma_TPA (two-photon absorption cross section of GaS) = 0.71e-27 to 7.65e-27 cm4/GW
    Fitted from GaS Z-scan data with Equation (7); used to model RSA and later as an input at the interface.
  • sigma_12 (excited-state absorption cross section of GaS) = 1.9e-18 cm2 in main text for 375 GW/cm2; Table S2 lists 4.07e-17 cm2
    Fitted from GaS Z-scan data; contributes to the RSA interpretation. The text/table discrepancy is itself a red flag.
  • fitting constants a, b, c = not given explicitly
    Introduced in Equations (S10) and (S22) 'considering the influence of experimental error' to describe the intensity dependence of beta_SA and beta_RSA; they are ad hoc fit parameters.
assumptions (6)
  • standard math The open-aperture Z-scan analysis of Equation (5) is valid for extracting nonlinear absorption coefficients from the NbC and GaS films.
    Equation (5) is taken from reference [27] and used without derivation; the thin-sample approximation is standard for Z-scan.
  • domain assumption The steady-state three-level model of Equations (S1)-(S6) describes saturable absorption in NbC.
    The model assumes one ground and one excited state with population relaxation time, sufficient for Pauli-blocking SA; used to derive Equation (4).
  • domain assumption The TPA plus excited-state absorption model of Equations (S11)-(S18) describes GaS response below its band gap at 800 nm.
    Assumes two-photon absorption from the ground state plus excited-state absorption; the slope of 1.22 in ln(1-T) vs ln(I) is interpreted as TPA plus ESA.
  • ad hoc to paper The NbC/GaS interface can be represented as a Schottky heterojunction described by the four-level rate equations (S25)-(S31).
    This model is specific to this paper; the interface cross sections are not independently measured and are adjusted to make the simulation match experiment.
  • ad hoc to paper The diode can be modeled as sequential propagation through the SA layer, an interface region, and the RSA layer, discretized in 1 nm slices with fourth-order Runge-Kutta integration.
    The decomposition and slicing scheme is introduced in Supporting Information section 8; the interface region thickness and parameters are not independently verified.
  • domain assumption The linear absorption coefficients and film thicknesses measured by Lambert-Beer law, UV-Vis, and AFM are accurate inputs for the nonlinear fits.
    Used to compute alpha0, effective thickness, and nonlinear coefficients; standard characterization assumptions.

how reviews work

0 comments
Cite this review

Pith. "Pith review of All-optical diode via nonreciprocal nonlinear absorption and interfacial charge transfer in two-dimensional van der Waals heterostructures." pith.science (2026). https://pith.science/paper/LZ4PEH6M

@misc{pith2026250524678,
  author       = {Pith},
  title        = {Pith review of: All-optical diode via nonreciprocal nonlinear absorption and interfacial charge transfer in two-dimensional van der Waals heterostructures},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/LZ4PEH6M}},
  note         = {Machine review of arXiv:2505.24678}
}
read the original abstract

Nonreciprocity is fundamental to photonic and optoelectronic devices such as all-optical diodes for ultrafast optical signal processing. However, previous nonreciprocity is mainly based on linear optical response instead of nonlinear optical response based on recently developed two-dimensional (2D) van der Waals heterostructures. Herein, an all-optical diode prototype based on nonreciprocal nonlinear absorption and interfacial charge transfer is proposed and designed by both simulation and experiment based on ready van der Waals heterostructures. The giant saturable absorption from 2D MXenes (NbC) and reverse saturable absorption from 2D chalcogenides (GaS) play a synergistic role in the designed all-optical diodes, which is characterized by a femtosecond laser based Z-scan system. The comprehensive physical mechanism of this all-optical diode based on 2D van der Waals NbC/GaS heterostructure designed by simulations, is consistent with experiments under the consideration of both nonreciprocal nonlinear absorption and interfacial effect. This all-optical diode based on the 2D van der Waals heterostructure features the simplicity, scalability, stability, integration, and compatibility with the complementary planar fabrication technology, which can further extend and miniaturize the nonlinear photonic and optoelectric devices.

Figures

Figures reproduced from arXiv: 2505.24678 by the authors.

Figure 1
Figure 1. (a) Schematic of all-optical diode with forward bias and reverse bias; (b) charge transfer process at the interface of van der Waals heterostructures [PITH_FULL_IMAGE:figures/full_fig_p006_1.png] view at source ↗
Figure 2
Figure 2. Simulation results of all-optical diode with different (a) SA 0 , (c) RSA  0 , (e)  SA , and (g)  RSA ; nonreciprocal factor (F, black) and transmittance symmetry (S, red) versus (b) SA 0 , (d) RSA  0 , (f)  SA , and (h)  RSA . (In addition to the variable parameters, the other invariants are SA 0 = 6.7  104 cm-1 , RSA  0 = 7.7  104 cm-1 ,  SA = -300 cm/GW,  TPA = 2.310-18 cm4 /GW, 12 = 1.710-18 cm2… view at source ↗
Figure 3
Figure 3. (g) with the e from 1.410-19 to 2.310-17 cm2 . As shown in [PITH_FULL_IMAGE:figures/full_fig_p009_3.png] view at source ↗
Figures from the paper (3 more)
Figure 4
Figure 4. Figure 4: (a) Open-aperture Z-scan traces of NbC film under different on-focus intensity excitation, (b) three energy level model, (c) normalized optical transmittance as a function of Iz, and (d) nonlinear absorptive coefficient of NbC film versus on-focus intensity. To accurat…
Figure 5
Figure 5. Figure 5: (a) Open-aperture Z-scan results of GaS film under different on-focus intensities, (b) plot of ln(1-T) vs. ln(Iz), (c) normalized optical transmittance (inset is TPA energy-level model), and (d) TPA coefficient of GaS film versus optical intensity. Because the optical …
Figure 6
Figure 6. Figure 6: (a) Z-scan curves of NbC/GaS heterostructures in forward (blue) and reverse (orange) biases, (b) all-optical diode response in forward (blue) and reverse (orange) biases with different pump intensity, (c) simulated all-optical diode response of NbC/GaS heterostructures…

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

3 extracted references · 3 canonical work pages

  1. [1]

    Reconfigurablelight-drivenopto-acousticisolatorsinphotoniccrystalfibre

    IntroductionNonreciprocity, breaking the time-reverse symmetry, is fundamental to the operation inall-opticaldiodedevices,whichplays asignificantandpowerfulroleinultrafastopticalsignalprocessing. However, previously reported diode responses were fabricated by means ofopto-acoustic effect[1], magneto-optic effect[2], and all-silicon microrings[3] fornonrec...

    work page 2011

  2. [4]

    Two-Dimensional Gallium Sulfide Nanoflakes for UV-SelectivePhotoelectrochemical-type Photodetectors

    Relationship between nonlinear absorption coefficient of NbC andI0For the saturable absorption response,SA( )Ihas been described in Equation (S6).Besides,SA( )Ihasthegeneralexpression: S A 0S A S A) +( I I   (S8)WhereSAisthenonlinearabsorptioncoefficientofNbCfilm.CombiningEquation(S6)with(S8),therelationshipbetweenSAandI0isexpressas: SA 0SA 0 ...

    work page 2021

  3. [28]

    Saturable absorption and self-defocusingresponse of 2D monoelemental germanium nanosheets in broadband spectra

    Zhou, L, Wang, Y,Wang,Y, Xiao, S and He, J. "Saturable absorption and self-defocusingresponse of 2D monoelemental germanium nanosheets in broadband spectra." Optics Express29(24),39115-39124(2021).29.Wang,Y,Wang,Y,Dong,Y,Zhou, L,Wei,H, Long, M, Xiao, Sand He, J. "The nonlinearopticaltransitionbleachingintellurene."Nanoscale13(37),15882-15890(2021).30.Wang...

    work page 2021

Pith tools

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