{"id":"66156f92-c736-4b09-8cbb-ea853b78b83d","arxiv_id":"1909.13625","paper_version":1,"verdict":"REJECT","confidence":"LOW","novelty_score":4.0,"correctness_risk":"unknown","formal_verification":"none","parameter_count":3,"one_line_summary":"A graphene metasurface sandwiched between two Fourier lenses is designed to differentiate and integrate optical fields, with reconfiguration via chemical-potential tuning.","lead":"This paper proposes a thin layer of graphene squares, placed between two lenses, that can perform calculus operations on light beams, such as taking a derivative or an integral. The design is meant to be reconfigurable by applying voltage to change graphene's conductivity, which could lead to programmable optical computers.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Even if the missing geometry-to-phase mapping were supplied, the paper's own phase-only, high-transmission assumption cannot realize T = i k_x or T = 1/(i k_x), whose amplitudes must vary with |k_x|; no amplitude-control mechanism is described.","rationale":"The reader's weakest assumption was that the geometry-to-phase mapping for the graphene squares is unverified and deferred to missing Supplementary Tables. That is a real reproducibility gap, but it treats the mapping as the main missing link. My review identifies a more fundamental, internally visible inconsistency: the paper's own design premise is a phase-only, high-transmission metasurface, yet the transfer functions it claims to realize require strongly varying amplitude with spatial frequency. A phase-only device cannot multiply the Fourier spectrum by i k_x or 1/(i k_x); the output of such a device would be the Fourier transform of a phase-modulated spectrum, not a derivative or integral. This concern is independent of the missing tables and the ideal-lens idealization. If the intended design actually does use amplitude variation through the graphene squares, then the paper must say so explicitly, document the resulting |t(k_x)| profile, and reconcile it with the 'high transmission' statement. Since the current text provides no such mechanism, the central claim is unsupported and the rejection stands. I therefore recommend no change to the reader's verdict, while noting that my specific objection differs in emphasis from the reader's weakest-assumption identification.","tokens_in":4909,"tokens_out":4420,"duration_ms":52371,"concrete_test":"Obtain the complex transmission t(k_x) from the dimensions promised in Supplementary Table I, or from a full-wave simulation of a single graphene square as a function of size, and plot its magnitude and phase across the operating k_x band. If |t| is within 10% of unity across the band, compute E_out(x) = ∫ t(k_x) E_in(k_x) e^{i k_x x} dk_x for the Gaussian E_in of Fig. 2 and compare with its ideal derivative. If the normalized correlation is not close to one, the phase-only high-transmission premise cannot support the claimed transfer functions. If instead |t| ∝ |k_x|, then the paper's 'high transmission' claim must be revised and the amplitude mechanism explicitly specified.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Section 3 justifies the metasurface design by stating that varying the length and width of the graphene squares can provide almost 2π phase shift while maintaining a high level of transmission. Section 4 then claims that this same structure realizes the transfer functions T1 = i k_x (differentiator) and T2 = 1/(i k_x) (integrator). These transfer functions are not phase-only: their magnitudes are proportional to |k_x| and to 1/|k_x|, respectively. A high-transmission phase screen implements t(k_x) ≈ exp(iφ(k_x)) with |t| ≈ 1 over the band, whose inverse Fourier transform is not proportional to the derivative kernel δ'(x) or to the integral kernel. For the Gaussian input of Fig. 2, the output spectrum must be multiplied by |k_x| to produce the derivative; with constant |t|, the output cannot match the shown derivative profile. Neither Section 3 nor Section 4 explains how the required amplitude profile |k_x| or 1/|k_x| is obtained, and the paper's 'high transmission' wording implies the amplitude is not being used as a design degree of freedom. Additionally, T2 diverges at k_x = 0, so a bounded approximation over a specified spatial-frequency band is required, but no band limits or regularization are provided. The central demonstration therefore depends on an amplitude-control capability that is absent from the stated design, even before considering the ideal-lens approximation and the missing Supplementary Tables.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proposes a reconfigurable optical computing system in which a metasurface made of subwavelength graphene squares on a silicon dioxide substrate is placed between two Fourier lenses. The authors claim that by spatially varying the size of the graphene squares, the structure can realize the transfer functions T = i k_x (first-order differentiator) and T = 1/(i k_x) (integrator), so that the output field is proportional to the derivative or integral of the incident field. They further claim that the same geometry can be dynamically reconfigured to a different frequency range by tuning the chemical potential of the graphene flakes via an external bias voltage. The paper presents qualitative comparisons between simulated output fields and ideal derivative/integral profiles, without any quantitative error metrics, and defers the essential design parameters to Supplementary Tables that are not included in the manuscript.","tokens_in":5170,"tokens_out":4767,"duration_ms":52242,"significance":"If the central claims were fully supported, the paper would contribute a useful reconfigurable implementation of computational metamaterials, exploiting graphene's tunable conductivity. The conceptual direction is timely and builds on an established line of work (e.g., Silva et al.). However, as presented, the evidence is far from sufficient: the transfer functions are stated without derivation, the geometry-to-transmission mapping is missing, the Kubo formula is garbled, and the validation is purely qualitative. The paper also contains a potentially load-bearing inconsistency between the described phase-only, high-transmission design and the amplitude-varying transfer functions. The work would be significant if these gaps were closed, but in its current form it does not substantiate its claims.","major_comments":[{"comment":"The design described in Section 3 provides 'almost 2π phase shift' while maintaining 'a high level of transmission', which implies a phase-only screen with |t(k_x)| ≈ 1. In contrast, the transfer functions T1 = i k_x and T2 = 1/(i k_x) have magnitudes proportional to |k_x| and 1/|k_x|, respectively. A phase-only screen cannot implement these amplitude profiles. The authors must explain how the required amplitude modulation is achieved, and provide the complex transmission profile t(k_x) of the metasurface (e.g., from full-wave simulation) with its magnitude and phase plotted against the ideal transfer function over the operating spatial-frequency band.","section":"Sections 3 and 4"},{"comment":"The printed Kubo formula is garbled to the point of being unusable: symbols are concatenated (e.g., '2 ( ) 2 ( )[ 2ln( 1)] ln[ ] ln[ ] 4 2 ( ) 4 2 ( )'), brackets are mismatched, and the variables (relaxation time, Fermi energy, frequency, temperature) are not clearly assigned. The authors should provide a clean, correctly typeset formula and state the numerical values of all parameters used in the simulations, including substrate permittivity and flake dimensions.","section":"Equation (1)"},{"comment":"The central claim that the graphene squares realize the transfer functions T1 and T2 rests entirely on Supplementary Tables I and II, which are missing from the manuscript. The reconfigurability demonstration similarly depends on the chemical potentials in Table III. Without these tables, and without a description of how the square dimensions map to the complex transmission coefficient, the design is not reproducible and the validation cannot be checked. These materials are load-bearing and must be included.","section":"Section 4 and Supplementary Tables I–III"},{"comment":"The validation of the differentiator and integrator is qualitative: the authors state that comparing Figs. 2b and 2c 'one induces that the proposed differentiator is functioning perfect', but no quantitative error metric is provided. The authors should report a normalized mean-square error or a correlation coefficient between the simulated output and the ideal derivative/integral over the full spatial domain, and they should specify the spatial-frequency band over which the realized transfer function matches the ideal, including a regularized treatment of the integrator's divergence at k_x = 0.","section":"Section 4 (validation)"},{"comment":"The analysis assumes that the two graded-index Fourier lenses perform ideal Fourier transforms with no loss, finite aperture, or aberrations. This idealization is unquantified. Since the output fidelity depends on the accessible spatial-frequency band and on the lens transfer function, the authors should specify the lens parameters and justify that the ideal-lens approximation holds over the band of interest.","section":"Sections 3 and 4 (Fourier lens assumption)"}],"minor_comments":[{"comment":"The section numbering in the Introduction is inconsistent: the text refers to 'Section II, Section III, Section III, Section IIII' while the actual sections are numbered 2 through 6.","section":"Introduction"},{"comment":"The caption of Figure 2 is duplicated after the discussion of the integrator, and Figure 3 appears to be referenced but its caption is missing. The captions and figure numbers should be corrected.","section":"Figures 2 and 3"},{"comment":"The text contains numerous typographical errors and garbled phrases, such as 't he' in the abstract, 'Plank constant' instead of 'Planck constant', and 'Reconfigurabilty' in the Section 5 heading. A thorough language edit is needed.","section":"Throughout"},{"comment":"Reference [19] is incomplete ('Silva, Alexandre, et al. Science 342 (2014): 160-163.' lacks the article title), and the reference list contains duplicate entries (refs. 20 and 36, 21 and 30, 32 and 41).","section":"References"},{"comment":"The transfer functions are written in garbled notation: 'T_rx = i k_x' and 'T_rx = 1/(i k_x)' appear as '1rxT ik=' and '2 1/rxT ik='. The equations should be typeset properly, with a clear distinction between the operator T and the gradient index of the Fourier lens.","section":"Section 4"}],"recommendation":"major_revision","confidential_remarks":"The manuscript appears to be an incomplete draft: it contains placeholder text ('Insert PSN Here'), duplicated figure captions, garbled equations, and missing supplementary tables. The main scientific concern is not just presentation: the phase-only design statement in Section 3 appears to conflict with the amplitude-varying transfer functions claimed in Section 4. If the authors can supply the missing tables, correct the Kubo formula, and provide quantitative validation including the complex transmission profile, the manuscript could become a viable contribution. In its current state, it is far from publishable, and I would suggest the editor require a substantially expanded and carefully revised version before resubmission."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Plainly: this is an incomplete draft, and its central design assumption doesn't survive contact with the transfer functions it claims to implement.\n\nThe idea is a reasonable extension of Silva et al.'s computational metamaterials: put a metasurface between two Fourier lenses, and make the metasurface reconfigurable by gating graphene flakes. Graphene's voltage-tunable conductivity is a sensible handle, and the general architecture is standard. If the device worked, it would add dynamic programming to optical analog processing. That's worth something.\n\nBut the paper as submitted is not ready for review. There are placeholders, a garbled Kubo formula, duplicated figure captions, section numbering errors, and no actual dimensions for the flakes—they're deferred to missing Supplementary Tables. The validation is purely qualitative visual comparison, with no error metrics or simulation parameters. That alone would justify a desk reject.\n\nThe deeper problem is the one your stress test flags. Section 3 says the graphene squares can provide almost 2π phase shift while keeping high transmission. That describes a phase-only screen, |t| ≈ 1. The differentiator and integrator transfer functions, T = i k_x and T = 1/(i k_x), require amplitude to scale as |k_x| and 1/|k_x|. A phase-only screen cannot do that; its impulse response isn't a derivative or integral kernel. The paper never explains how amplitude is controlled—and the integrator diverges at k_x = 0 anyway, with no band limits given. This isn't just missing detail; it's an internal contradiction between the stated design mechanism and the claimed function.\n\nThe reconfigurability section is also weak: it raises the operation frequency by 10% and then tunes the chemical potentials to recover the original output. That's a tuning exercise, not a demonstration of a new function.\n\nCredit where it's due: graphene-based reconfigurable computational metasurfaces are a plausible direction, and the authors correctly identify the relevant prior art in computational metamaterials and graphene devices. The concept is worth working on. But this manuscript doesn't support the central claim.\n\nFor peer review: desk reject. If the authors can supply the missing tables and—more importantly—show how a high-transmission graphene metasurface yields the amplitude profiles in T1 and T2, it might deserve a serious look. As it stands, no.","headline":"An incomplete draft whose phase-only design assumption can't realize the amplitude-varying transfer functions it claims, on top of missing tables and garbled equations.","tokens_in":5724,"tokens_out":2827,"would_cite":false,"duration_ms":28719,"reading_group":"no","serious_thinker":"no","would_accept_peer_review":false},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"A graphene square metasurface flanked by two Fourier lenses can compute the derivative or integral of an incident optical field, retunable through the chemical potential.","keywords":["optical computing","graphene metasurface","spatial differentiation","optical integration","reconfigurable optics","Kubo conductivity","Fourier transform","subwavelength metamaterial"],"falsifier":"Measure the complex transmission coefficient of single graphene squares on silicon dioxide at the design frequency as a function of side length and width; if the achievable values do not span nearly $2\\pi$ of phase with high transmission, the transfer functions cannot be synthesized. At the system level, send a known Gaussian beam through the two-lens setup and compare the output profile to the analytic first derivative: a mismatch in the side lobes or a residual background would falsify the claim.","tokens_in":4641,"feed_emoji":"🧮","tokens_out":8239,"duration_ms":82272,"temperature":0.7,"pith_summary":"This paper aims to show that a thin planar structure—a metasurface made of subwavelength graphene squares on silicon dioxide—can perform two classic optical-computing operations on an incoming light field. Placed between two Fourier lenses, the metasurface multiplies the field's spatial-frequency spectrum by $T = i k_x$ or $T = 1/(i k_x)$, so the output field is the first derivative or the integral of the input. Because the surface conductivity of graphene depends on its Fermi level, the same fabricated geometry can be retuned to a new operating frequency by changing the chemical potential of the flakes, i.e., by applying a bias voltage. This matters because a single reconfigurable component could carry out real-time edge detection, image processing, or equation solving in the optical domain, without electronics.","feed_headline":"Graphene metasurface does calculus on light","feed_subtitle":"Two Fourier lenses and electrically tuned graphene squares turn an input beam into its derivative or integral.","key_machinery":"The central object is a computational metasurface: a planar array of subwavelength graphene squares resting on a silicon dioxide substrate, located in the Fourier plane between two Fourier lenses. Each square is sized so that its local complex transmission coefficient implements the required value of the transfer function at the corresponding spatial frequency $k_x$. The first lens Fourier-transforms the incident field, the metasurface multiplies the spectrum by $T(k_x)$, and the second lens transforms back to real space. The design uses the Kubo formula for graphene's surface conductivity, whose dependence on Fermi level is the handle that makes the response tunable by bias voltage.","core_discovery":"The core claim is that a single metasurface of subwavelength graphene squares can be engineered as a spatial-frequency filter with transfer function $T=i k_x$ (first-order differentiator) or $T=1/(i k_x)$ (integrator), with two graded-index Fourier lenses performing the forward and inverse transforms. For a Gaussian beam input, the paper's numerical output matches the ideal first derivative; for a Gaussian-derivative input, the output matches the ideal integral. The paper further claims that increasing the operating frequency by ten percent does not disrupt these operations provided the chemical potentials of the flakes are adjusted, so the device is reconfigurable without any change in geometry.","pith_inferences":["Beyond the paper, the same architecture could be pushed to higher-order operators such as $T=(i k_x)^2$ or $T=k_x^2+k_y^2$, turning the device into a general programmable spatial filter.","The chemical-potential knob that the paper uses for frequency retuning could in principle also switch the device between differentiation and integration, since a different bias changes which transfer function the metasurface approximates.","An immediate experimental test would be to fabricate a small array and compare its complex transmission against the design values; the missing supplementary dimension tables make this the first obstacle for any reproductions."],"forward_implications":["A thin optical component could differentiate or integrate an incoming field in a single pass, replacing an electronic or bulk-optical processing step.","The same fabricated metasurface can be shifted to a different frequency band by changing the graphene chemical potential, making one device serve multiple spectral ranges.","Because spatial differentiation in the Fourier domain is equivalent to edge enhancement, the configuration offers a direct path to optical edge detection and image sharpening.","The success of the square-flake design implies that other linear operators that are functions of spatial frequency can be imprinted into a metasurface by the same local-sizing recipe."],"supporting_citations":[{"why":"Introduces the computational-metamaterial concept that this design uses to realize mathematical operations through a structured slab.","marker":"[19]"},{"why":"Demonstrates plasmonic spatial differentiation, the same operation the graphene metasurface aims to make reconfigurable.","marker":"[11]"},{"why":"Shows an all-optical differential-equation solver, establishing the analog-computing application context for the proposed device.","marker":"[9]"},{"why":"Documents graphene-based tunable devices whose conductivity-tuning mechanism the authors invoke for reconfigurability.","marker":"[20-43]"}],"fun_headline_variants":["Graphene metasurface calculates light's derivative and integral","Tunable graphene does calculus on optical beams","Reconfigurable graphene chip computes with light","Graphene squares turn light into its math functions","Light gets its derivative or integral via graphene"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The design assumes that varying the length and width of subwavelength graphene squares can provide nearly $2\\pi$ of phase shift while keeping transmission high; the exact dimensions are not given in the manuscript, so if the flakes cannot realize the required complex transmission values, the output will not match a derivative or an integral.","fun_headline_variants_meta":{"raw":{"variants":["Graphene metasurface calculates light's derivative and integral","Tunable graphene does calculus on optical beams","Reconfigurable graphene chip computes with light","Graphene squares turn light into its math functions","Light gets its derivative or integral via graphene"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000275,"raw_usage":{"total_tokens":1551,"prompt_tokens":759,"completion_tokens":792,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":375,"completion_tokens_details":{"reasoning_tokens":720}},"tokens_in":375,"tokens_out":792,"duration_ms":8765,"temperature":1.0,"reasoning_tokens":720,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T12:32:48.682018+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the complex transmission coefficient of single graphene squares on silicon dioxide at the design frequency as a function of side length and width; if the achievable values do not span nearly $2\\pi$ of phase with high transmission, the transfer functions cannot be synthesized. At the system level, send a known Gaussian beam through the two-lens setup and compare the output profile to the analytic first derivative: a mismatch in the side lobes or a residual background would falsify the claim.","supporting_citations":[{"cited_title":"Science 342 (2014): 160-163","cited_arxiv_id":null,"evidence_quote":"Introduces the computational-metamaterial concept that this design uses to realize mathematical operations through a structured slab."},{"cited_title":"Plasmonic computing of spatial differentiation","cited_arxiv_id":null,"evidence_quote":"Demonstrates plasmonic spatial differentiation, the same operation the graphene metasurface aims to make reconfigurable."},{"cited_title":"All -optical differential equation solver with constant -coefficient tunable based on a single microring resonator","cited_arxiv_id":null,"evidence_quote":"Shows an all-optical differential-equation solver, establishing the analog-computing application context for the proposed device."}],"review_version":1}