{"id":"d9983d09-78a8-4c86-98bd-c3feb173cd03","arxiv_id":"1908.08516","paper_version":3,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"A single time-periodic LC resonator is experimentally shown to host an exceptional point whose resonance frequency responds to small capacitance changes with a square-root scaling, demonstrating a new route to EPD-based sensing.","lead":"This paper reports the first experimental realization of exceptional points of degeneracy (EPDs) in a single LC resonator, achieved by time-modulating its capacitor rather than by coupling two resonators with balanced gain and loss. It also demonstrates that a tiny capacitance perturbation shifts the resonance frequency by an amount proportional to the square root of the perturbation, which could enable a new class of sensitive sensors.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Sensitivity result rests on an unmeasured pump-voltage-to-capacitance calibration; a multiplier gain/offset error could invalidate the quantitative Δf∝√δ agreement.","rationale":"The reader's weakest assumption correctly identifies the calibration of δ as the load-bearing point. The EPD itself is supported by two independent observables: the measured dispersion diagram in Fig. 4(a) follows the theoretical branch coalescence, and the time-domain response at the EPD points (#1 and #3) shows the predicted linear-in-n secular growth, whereas the adjacent unstable regime shows exponential growth. Those signatures are hard to explain without a Floquet Jordan degeneracy, so I do not see a soundness problem with the EPD identification. The sensitivity measurement is where the argument is less secure. The theoretical Puiseux prediction is parameter-free in the sense that α1 is computed from the nominal L and C values; however, the experimental x-coordinate depends entirely on the ideal multiplier equation. Without a direct capacitance measurement, a calibration error could change the slope of Δf versus δ and, if nonlinear, the apparent exponent. Four points without error bars are insufficient to distinguish β=0.5 from nearby exponents if the calibration is uncertain. A direct capacitance measurement at the same pump levels is a simple, decisive check. If it confirms δ values, the CONDITIONAL verdict should be upgraded; if not, the quantitative sensitivity claim should be softened to a qualitative demonstration. The reader's CONDITIONAL verdict is therefore appropriate and unchanged.","tokens_in":17428,"tokens_out":9938,"duration_ms":105376,"concrete_test":"Measure the actual time-varying capacitance seen at Node A for the exact pump high levels used in Fig. 5 (e.g., apply a small sinusoidal probe voltage during the ON state and compute ic(t)/[dv/dt], or use an impedance analyzer with the pump running), then convert each pump setting to a measured δ. Re-plot the experimental frequency shifts versus measured δ, fit Δf = A δ^β, and check that β=0.5 is within the confidence interval and that the points collapse onto the exact/Puiseux curves. This would settle whether the apparent √δ agreement is an artifact of the assumed linear calibration.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's exceptional-sensitivity claim rests on Fig. 5, where the perturbation δ is not measured but inferred from the ideal multiplier relation C(t)=C0(1−vp(t)/V0), with V0=1.05 V taken from the AD835 datasheet (Supplement, Eqs. (7)–(10)). Each 5 mV change in the pump high level is converted to δ=0.01 via this linear relation. No direct measurement of the synthesized capacitance is reported. If the actual multiplier has a different normalization V0, a gain error, or an output offset, then the true δ values differ from the nominal ones. A constant multiplicative error preserves the square-root functional form but moves the data off the absolute theoretical curve; a nonlinearity (e.g., multiplier saturation or voltage-dependent capacitor response) could make four points appear to follow a power law even if the underlying response is not √δ. The paper shows no error bars and no independent calibration, and the 'excellent agreement' between green triangles and the Puiseux curve is asserted rather than quantified. Because the central sensing claim is precisely the quantitative √δ shift for small δ, this missing calibration is the most load-bearing weakness.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript reports an experimental demonstration of a second-order exceptional point of degeneracy (EPD) in a single LC resonator whose capacitance is modulated in time through a piecewise-constant pump voltage and an analog multiplier. The authors verify the EPD by two signatures: the measured dispersion diagram of the resonance frequencies versus modulation frequency, and the observed linear growth of the time-domain capacitor voltage at the predicted EPD modulation frequencies. They then perturb the high level of the pump voltage and report that the resonance frequency shifts follow the square-root Puiseux law Δf ∝ √δ, concluding that this provides a single-resonator, easily tunable EPD-based sensor that avoids the precise gain/loss balancing needed in PT-symmetric coupled-resonator schemes. The theoretical framework is a state-transition-matrix analysis carried over from the authors' prior paper [32]; the experimental realization and the sensitivity measurement are the new contributions.","tokens_in":17611,"tokens_out":7366,"duration_ms":77308,"significance":"If the sensitivity result withstands scrutiny, this would be a valuable experimental confirmation that EPDs in linear time-periodic systems can be realized in a single resonator and that the associated fractional frequency shift can be observed in a realistic noisy electronic circuit. The paper has clear strengths: the EPD existence is supported by two independent signatures (the dispersion diagram and the time-domain growth), the frequency-shift prediction in Fig. 5(a) is not a fit because α1 is calculated, and the authors candidly acknowledge the ongoing debate around EPD-enhanced sensing and address the role of noise. The main weakness is that the perturbation δ is not measured but inferred from the ideal multiplier relation, so the quantitative aspect of the exceptional-sensitivity claim is not yet established. The experimental demonstration is also somewhat incremental relative to the authors' earlier theory paper [32], but the hardware realization is a meaningful step toward practical EPD sensors.","major_comments":[{"comment":"The x-axis of Fig. 5(a) is set by assuming an exact linear relation between the pump high-level voltage and the capacitance perturbation δ. The authors infer δ from the ideal multiplier equation C(t)=C0(1−vp(t)/V0) with V0=1.05 V taken from the AD835 datasheet, but no direct measurement of the synthesized capacitance is reported (for example, by measuring ic(t)/[dv/dt] at Node A). A constant multiplier gain or offset error would scale all δ values and move the green triangles off the theoretical curve while preserving the √δ functional form, and a nonlinearity (e.g. multiplier saturation or switch nonidealities) could produce a spurious sub-linear dependence. Because the central sensing claim is the quantitative Δf∝√δ behavior, the δ calibration must be independently measured and the resulting curves must be shown with uncertainties before the 'excellent agreement' assertion in Section III.B is supported.","section":"Section III.B; Supplemental Eqs. (6)-(10)"},{"comment":"The sensitivity demonstration reports no quantitative uncertainty. The green triangles in Fig. 5(a) have no error bars, and the text does not state the FFT frequency resolution, the linewidths of the measured peaks, or the run-to-run repeatability across reset cycles. The claim that a δ=0.003 perturbation is 'distinguishable' requires a quantitative comparison between the peak separation and the peak width/noise floor. As written, the data support a qualitative √δ-like trend but do not fully substantiate the quantitative sensitivity enhancement claimed in the abstract and conclusion.","section":"Section III.B and Fig. 5"}],"minor_comments":[{"comment":"The rendered formula for α1 is garbled in the manuscript text ('/radical√p/radicalver√ex/...'); the production version should be checked carefully so that the Puiseux coefficient expression is readable.","section":"Section II, Eq. (9)"},{"comment":"The slight shift between the theoretical and experimental dispersion curves is attributed to parasitics and tolerances, but no quantitative estimate is given; a number (e.g. the frequency offset at the EPD) would make the 'good agreement' statement more concrete.","section":"Section III.A and Fig. 4(a)"},{"comment":"The sentence 'each 5 mV change in the positive level of the pump voltage will result in 1% change of the C1 capacitor value' relies on the ideal multiplier model and would be better accompanied by a statement of the measured C(t) or at least by a discussion of how the actual multiplier gain and offset were characterized.","section":"Section III.B"},{"comment":"The phrase 'Proof of exceptional sensitivity' is stronger than the evidence presented; 'Demonstration' or 'Measurement' would be more appropriate.","section":"Figure 5 caption"},{"comment":"The derivation leading to C(t)=C0(1−vp(t)/V0) ignores the switching transients in vp(t); it would be helpful to state explicitly that the relation is valid away from the switching instants and to comment on any charge-injection effects from the reset switches.","section":"Supplemental Material, Section I"}],"recommendation":"major_revision","confidential_remarks":"The paper is an experimental follow-up to the same group's theoretical work in Ref. [32]; the incremental novelty is acceptable for an applied physics journal if the sensitivity calibration and uncertainty analysis are strengthened. The central EPD existence result appears sound, but the quantitative sensitivity claim currently rests on an unmeasured conversion from pump voltage to capacitance perturbation. I recommend major revision to address this and the missing error analysis."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The interesting thing here is the experiment, not the theory. The EPD in a linear time-periodic resonator was predicted in the group's own prior paper (Ref. [32]); what this paper adds is the first actual circuit implementation and the first measurement of the square-root sensitivity. That is a real step forward for the EPD-sensing community, which has mostly worked with coupled PT-symmetric resonators that require careful gain/loss balancing. A single resonator tuned by modulation frequency is a simpler and arguably more practical route, and the paper shows it working.\n\nThe experimental evidence for the EPD itself is solid. The dispersion diagram in Fig. 4(a) matches the theoretical branches well, and the time-domain waveforms at the predicted modulation frequencies show the expected linear growth that distinguishes an EPD from an ordinary resonance. These two independent signatures make the central claim believable. The circuit design is carefully described, including the reset mechanism that prevents saturation, and the discussion of noise is honest rather than hand-waving.\n\nWhere I would push back is the quantitative sensitivity claim. The perturbation δ is not measured; it is inferred from the ideal multiplier equation, assuming each 5 mV change in the pump high level corresponds to δ = 0.01. The stress-test note is right that this is load-bearing. The four data points in Fig. 5(a) follow the square-root curve, but if the multiplier has a gain error or an output offset, the absolute x-axis moves, and a nonlinearity in the capacitance synthesis could make a non-√δ response look power-law. The paper offers no error bars, no independent calibration of δ, and no off-EPD baseline comparison. These are standard things to ask for in a sensing paper, and their absence does not invalidate the EPD demonstration, but it does mean the exceptional-sensitivity claim is not yet nailed down.\n\nFor a reader in this subfield, the paper is worth reading and worth citing as the first experimental version of this idea. It is also the kind of paper that deserves a serious referee: the physics is interesting, the implementation is genuinely new, and the weaknesses are fixable with a modest amount of additional measurement. I would send it to review with a request for a direct calibration of the pump-voltage-to-capacitance mapping and a proper uncertainty analysis.","headline":"A credible first experimental demonstration of EPDs in a single time-modulated resonator, with a sensitivity experiment that shows the expected square-root trend but rests on an unmeasured calibration that should be checked before the quantitative claim is taken at face value.","tokens_in":18184,"tokens_out":1590,"would_cite":true,"duration_ms":18312,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"A single time-modulated LC resonator hosts an exceptional point whose frequency shift scales as the square root of a small capacitance perturbation.","keywords":["exceptional point of degeneracy","linear time-periodic system","single-resonator sensor","Puiseux series","time-varying capacitance","square-root sensitivity","LC resonator","capacitive sensing"],"falsifier":"Measure the capacitance $C(t)$ actually presented by the multiplier circuit while the pump high level is swept over the range used in the sensitivity plot, using a direct impedance or bridge measurement; if the true $\\delta$ deviates from the assumed linear map, the claimed $\\sqrt{\\delta}$ scaling and sensitivity enhancement are not established by the current data. Alternatively, apply a known, independently calibrated capacitance change at fixed pump voltage and check that the FFT peaks shift by the predicted $\\sqrt{\\delta}$ amount.","tokens_in":17208,"feed_emoji":"⚡","tokens_out":7403,"duration_ms":72651,"temperature":0.7,"pith_summary":"This paper reports the first experimental demonstration that a single LC resonator, with one capacitor periodically switched between two values, can host an exceptional point of degeneracy (EPD): a parameter setting where two resonance modes coalesce into one. At the EPD, a small relative capacitance change $\\delta$ shifts the resonance frequency as $\\sqrt{\\delta}$ rather than $\\delta$, so a 1% perturbation yields roughly a 10% frequency shift. The authors built a circuit in which a multiplier synthesizes the time-varying capacitor, verified the predicted dispersion diagram and the linear-in-time voltage growth that marks the EPD, and showed that measured frequency shifts follow the square-root law down to at least $\\delta=0.003$. If correct, this gives a route to exceptionally sensitive capacitive sensing with a single resonator, tuned simply by adjusting a modulation frequency instead of matching gain and loss in two coupled resonators.","feed_headline":"Square-root sensitivity from one modulated resonator","feed_subtitle":"Switching one capacitor periodically creates an EPD, turning 1% capacitance changes into ~10% frequency shifts.","key_machinery":"The load-bearing object is the $2\\times2$ state transition matrix $\\Phi(T_m,0)$ that maps the resonator state, capacitor charge and inductor current, over one modulation period $T_m$. At the EPD, $\\Phi$ is non-diagonalizable with a degenerate eigenvalue $\\lambda_e=\\pm1$, so its $n$-th power grows linearly in the period index $n$; this algebraic growth produces the observed linear voltage rise. The sensitivity law comes from a first-order Puiseux expansion of the perturbed eigenvalues, with coefficient $\\alpha_1$ determined by the derivative of $\\det[\\Phi(\\delta)-\\lambda I]$. The physical implementation replaces the physical time-varying capacitor with a fixed capacitor $C_0$ and a four-quadrant multiplier that realizes $C(t)=C_0(1-v_p(t)/V_0)$, so the two capacitance levels $C_1$ and $C_2$ are set by the pump voltage levels and the EPD is tuned by the pump frequency $f_m$.","core_discovery":"The paper's central claim is that an exceptional point of degeneracy can be induced in a single linear time-periodic LC resonator by periodically modulating one capacitance value, and that operating at this EPD makes the resonance frequency exceptionally sensitive to perturbations. Theoretically, the two resonances of the unperturbed time-periodic system coalesce at the EPD, so the state transition matrix becomes a Jordan block; a small perturbation $\\delta$ to the capacitance $C_1$ splits the degeneracy with a Puiseux expansion $\\lambda_p(\\delta)=\\lambda_e+(-1)^p\\alpha_1\\sqrt{\\delta}$, giving a detected frequency shift $\\Delta f\\propto\\sqrt{\\delta}$. Experimentally, the authors observe both EPD types, at the center and at the edge of the fundamental Brillouin zone, in the dispersion diagram, measure the linear voltage growth characteristic of the degenerate state, and demonstrate that capacitance perturbations as small as 0.3% produce clearly resolvable frequency shifts. The authors argue that this single-resonator EPD is easier to realize and tune than PT-symmetric two-resonator EPDs and that its shifted resonances remain real-valued for a one-sided capacitance perturbation, a practical advantage for sensor readout.","pith_inferences":["If the pump-to-capacitance calibration is verified by direct measurement, the same circuit could be packaged as a compact capacitive sensor for pressure, humidity, or biochemical mass, where the quantity of interest changes an effective capacitance.","The same time-periodic EPD mechanism should extend to higher-order EPDs, for example by using more than two capacitance levels in one modulation cycle, with frequency shifts scaling as $\\delta^{1/m}$ and potentially larger sensitivity.","The experiments leave open a systematic noise study: quantifying how the reset duty cycle, measurement window, and multiplier noise floor affect the minimum resolvable $\\delta$ would test whether the $\\sqrt{\\delta}$ advantage survives at very low signal levels."],"forward_implications":["A sensor can be built from a single LC tank with an off-the-shelf multiplier, and the EPD is reached simply by tuning the pump frequency instead of matching gain and loss.","At the EPD, a relative capacitance perturbation $\\delta$ produces a frequency shift proportional to $\\sqrt{\\delta}$, so a 1% perturbation yields roughly a 10% shift, well above the noise floor observed in the measurements.","The linear voltage growth at the EPD gives a time-domain signature that can confirm EPD operation and could serve as a second readout channel.","Because only the modulation frequency needs adjusting, component tolerances in $L$ and $C$ do not prevent reaching the EPD; the sensor can be retuned in situ.","Unlike a PT-symmetric two-resonator sensor perturbed on one side, this single-resonator EPD keeps the split resonance frequencies real-valued for a capacitance perturbation, simplifying frequency readout."],"supporting_citations":[{"why":"Supplies the theoretical framework for EPDs in single-resonator time-periodic systems and the state-transition-matrix eigenvalue problem used to locate the EPD.","marker":"[32]"},{"why":"Provides the recursive formulas for the Puiseux fractional power expansion used to derive the $\\sqrt{\\delta}$ frequency splitting near the EPD.","marker":"[33]"},{"why":"Gives the closed-form expression for powers of the transition matrix at the EPD that yields the linear-in-time growth of the state vector.","marker":"[34]"},{"why":"Earlier demonstration of a PT-symmetric two-resonator circuit EPD, the baseline scheme this single-resonator approach is contrasted with.","marker":"[17]"},{"why":"PT-symmetric EPD sensor work showing enhanced sensitivity and the real-frequency readout obtained by rebalancing gain and loss; used as a comparison in the discussion.","marker":"[2]"},{"why":"PT-symmetric EPD sensor implementation where a varactor balances the sensing-side perturbation; the comparison case for complex versus real shifted resonance frequencies.","marker":"[19]"}],"fun_headline_variants":["One modulated resonator yields exceptional sensitivity","Single time-varying resonator achieves EPD boost","Exceptional sensitivity via one time-modulated capacitor","Time modulation alone creates an exceptional point for sensing"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The result depends on the calibration that a 5 mV increase in the pump voltage's high level corresponds exactly to a relative capacitance change of $\\delta=0.01$, which follows from the ideal multiplier equation and was not verified by directly measuring the capacitance seen by the circuit.","fun_headline_variants_meta":{"raw":{"variants":["One modulated resonator yields exceptional sensitivity","Single time-varying resonator achieves EPD boost","Exceptional sensitivity via one time-modulated capacitor","Time modulation alone creates an exceptional point for sensing"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000744,"raw_usage":{"total_tokens":3337,"prompt_tokens":983,"completion_tokens":2354,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":599,"completion_tokens_details":{"reasoning_tokens":2299}},"tokens_in":599,"tokens_out":2354,"duration_ms":15167,"temperature":1.0,"reasoning_tokens":2299,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T11:37:28.847612+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the capacitance $C(t)$ actually presented by the multiplier circuit while the pump high level is swept over the range used in the sensitivity plot, using a direct impedance or bridge measurement; if the true $\\delta$ deviates from the assumed linear map, the claimed $\\sqrt{\\delta}$ scaling and sensitivity enhancement are not established by the current data. Alternatively, apply a known, independently calibrated capacitance change at fixed pump voltage and check that the FFT peaks shift by the predicted $\\sqrt{\\delta}$ amount.","supporting_citations":[{"cited_title":"Kazemi, M","cited_arxiv_id":null,"evidence_quote":"Supplies the theoretical framework for EPDs in single-resonator time-periodic systems and the state-transition-matrix eigenvalue problem used to locate the EPD."},{"cited_title":"Welters, SIAM Journal on Matrix Analysis and Ap- plications 32, 1 (2011)","cited_arxiv_id":null,"evidence_quote":"Provides the recursive formulas for the Puiseux fractional power expansion used to derive the $\\sqrt{\\delta}$ frequency splitting near the EPD."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Gives the closed-form expression for powers of the transition matrix at the EPD that yields the linear-in-time growth of the state vector."},{"cited_title":"Schindler, A","cited_arxiv_id":null,"evidence_quote":"Earlier demonstration of a PT-symmetric two-resonator circuit EPD, the baseline scheme this single-resonator approach is contrasted with."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"PT-symmetric EPD sensor work showing enhanced sensitivity and the real-frequency readout obtained by rebalancing gain and loss; used as a comparison in the discussion."},{"cited_title":"Sakhdari, M","cited_arxiv_id":null,"evidence_quote":"PT-symmetric EPD sensor implementation where a varactor balances the sensing-side perturbation; the comparison case for complex versus real shifted resonance frequencies."}],"review_version":1}