{"id":"38d41b36-a632-4f91-b83f-904351565f17","arxiv_id":"2607.06135","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"In open-loop AM-He-KPFM, the heterodyne-driven second eigenmode back-acts on the first eigenmode via a dissipation-dominated, resonant inter-mode coupling.","lead":"Open-loop heterodyne Kelvin probe force microscopy is shown to produce an \"inverse heterodyne\" back-action: the electrostatically driven second cantilever mode feeds energy back into the first mode, mainly as dissipation. The paper supplies closed-form expressions and UHV experiments whose resonant line shapes match the predictions.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Unverified capacitance-gradient coefficients in the close-approach f2-sweep conditions could undermine the quantitative match of Eq. (23), though line-shape predictions survive.","rationale":"The reader's weakest assumption correctly identifies the companion's CG coefficients as the least-secure input. I agree that the quantitative strength of the inverse effect depends on the unverified Ki expansion. However, this concern is partially mitigated because the leading-order ZOTR coefficients K1≈K2≈C'' are standard Taylor-expansion results, and the FOTR expressions are stated in the main text. Moreover, the most distinctive experimental signatures (the |G2|^2 narrowing and the sign-changing Δf1 lobe) depend only on the transfer-function structure, not on the Ki prefactors; the experiments in Fig. 5 directly test those line shapes. Thus the existence claim is robust even if the companion were flawed. The quasi-quadrature condition Eq. (19) is less load-bearing than the reader implies, since Eq. (23) can be obtained by substituting the exact linear response of the second mode. I therefore keep the verdict at CONDITIONAL (UNCHANGED), but with the concrete test above as the decisive check on whether the quantitative agreement is genuine. The reader and I differ only in emphasis: I would require the companion's convergence proof to match the close-approach f2-sweep conditions, rather than treating the entire CG treatment as a black box.","tokens_in":26483,"tokens_out":30236,"duration_ms":261829,"concrete_test":"Obtain companion [25] and compute the effective coefficients K1, K2 from its non-truncated CG series for the exact Fig. 5 conditions (z1,0=12 nm, Δf1=−100 Hz, reported cantilever parameters, a realistic tip-sample capacitance model such as sphere-plane). Compare with the FOTR expressions of Eq. (9) and with the values implicitly used to generate Fig. 5(a,b). If the deviation exceeds ~20%, the quantitative match is unsupported. Alternatively, numerically evaluate C'(t) along the bimodal trajectory z(t)=zc+z1,0 cos(ω1t+Φ1)+z2,0 cos(ω2t+Φ2) with z2,0=0.18 nm and extract the Fourier amplitudes at ω1 and ω2; check whether they equal α2z2,0 and α1z1,0 respectively under the reported bias and zc. Also re-derive Eq. (23) from Eqs. (20)–(21) without invoking the small-ε condition, using the exact phase of the transfer function, to verify that the line-shape expressions hold for the full f2 sweep.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central quantitative claim is Eq. (23): the inverse-heterodyne contributions to Fd and Δω1 are proportional to α1α2, with αi set by the effective capacitance-gradient coefficients K1,K2 of Eq. (9). The non-truncated derivation, convergence proof, and numerical validation of these Ki are relegated to the concurrently submitted companion [25], which is not available in this manuscript. The paper does state FOTR expressions (Eq. 9b–9d), but these are truncations whose validity depends on the oscillation amplitudes and on zc through higher derivatives C^(3), C^(4). The key f2-sweep experiment (Fig. 5) was performed at Δf1 = −100 Hz (close approach), z1,0 = 12 nm, Umod = 1 V, VDC = +1 V — conditions under which ZOTR is not guaranteed; higher-order CG terms may be non-negligible. If the companion's convergence analysis is inaccurate for these parameters, the predicted magnitude of the inverse effect changes, and the claimed quantitative agreement in Sec. V B 1 could be coincidental. The qualitative signatures — the squared-transfer-function resonance peak and the sign-changing (1−u2^2) contribution to Δf1 — are independent of the prefactor, so the existence claim is not at risk; only the quantitative strength and distance dependence are. The quasi-quadrature condition Eq. (19) is a secondary premise: away from f2 resonance, ε is not small, yet Eq. (23) is compared with the full f2 sweep; an independent re-derivation from Eq. (21) using the exact transfer-function phase is needed to confirm no hidden truncation.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper develops an analytical bimodal virial/power-balance description of open-loop amplitude-modulated heterodyne KPFM and claims that the electrostatically driven second eigenmode feeds back onto the first eigenmode, producing an 'inverse heterodyne' contribution to the first-mode frequency shift and dissipation. The central results are Eq. (23): the first-mode drive acquires a term proportional to alpha1*alpha2*|G2|^2, and the frequency shift acquires a term proportional to alpha1*alpha2*|G2|^2*(1-u2^2). The theory predicts that the inverse effect appears mainly as a dissipation signal with a resonance peak in Vd(f2) that is about 1/sqrt(3) narrower than the second-mode transfer-function modulus, together with a sign-changing contribution to Delta-f1. The UHV experiments in Sec. V are presented as direct evidence, showing the predicted narrowing, the dispersive Delta-f1 lobe, and approximately quadratic VDC/Umod dependences. The manuscript also connects the effect to DHe-KPFM and He-PiFM and to broader multimode back-action phenomena.","tokens_in":26902,"tokens_out":15555,"duration_ms":140662,"significance":"If correct, the paper identifies a previously overlooked inter-mode back-action mechanism in a widely used KPFM variant and provides falsifiable predictions: the squared-transfer-function dissipation peak, the 1/sqrt(3) linewidth ratio, and the sign-changing (1-u2^2) conservative term. These are derived, not fit, and the f2-sweep data in Fig. 5 appear to match them. The direct-heterodyne scalings in Figs. 2-4 also provide useful independent cross-checks. However, the quantitative strength of the inverse effect is controlled by capacitance-gradient coefficients K1, K2 taken from an unavailable companion manuscript, and the central equations contain a dimensional inconsistency that must be resolved before the quantitative claims can be accepted.","major_comments":[{"comment":"The inverse-heterodyne terms in Eq. (23) have inconsistent dimensions. Since alpha_i has units N/m, ek2 has N/m, and |G2| has m/N, the term alpha1*alpha2/(2 ek2)*|G2|^2 in Eq. (23a) has units m/N inside a bracket that must have N/m; the same problem appears in Eq. (23b) with alpha1*alpha2/(ek2 Q2)*|G2|^2. Re-deriving from Eqs. (21)-(22) with z2,0 = z1,0*alpha1*|G2|, sin(epsilon)=cos(Phi_G2)=(1-u2^2)*ek2*|G2|, and cos(epsilon)=-sin(Phi_G2)=(u2/Q2)*ek2*|G2| gives bracket terms alpha1*alpha2*ek2/2*(1-u2^2)*|G2|^2 and alpha1*alpha2*u2*ek2/Q2*|G2|^2. This is not a minor typo: the magnitude difference is huge (order ek2^2) and directly affects the theoretical curves in Sec. V B 1. Please correct the equations or show an alternative derivation.","section":"Eq. (23a,b)"},{"comment":"The effective capacitance-gradient coefficients K0, K1, K2, and the FOTR expressions of Eq. (9), are taken from the companion manuscript [25], which is not available to the reader or reviewer. The key f2-sweep experiment (Fig. 5) was performed at Delta-f1 = -100 Hz, z1,0 = 12 nm, Umod = 1 V, VDC = +1 V, i.e., at close approach where the truncation may not be controlled. The quantitative agreement claimed in Sec. V B 1 depends on the companion's derivation and convergence analysis. Please include the necessary definitions, convergence tests, or validation in the main text or SI, or otherwise make the companion accessible. The qualitative line-shape predictions are independent of this issue, but the force magnitudes and distance dependence are not.","section":"Sec. II C and Sec. V B 1"},{"comment":"Eq. (23) is stated to be derived under the quasi-quadrature condition |epsilon| << 1, Eq. (19). However, the experimental comparison in Fig. 5 sweeps f2 over a range where epsilon is not small. If Eq. (23) is intended to hold over the full sweep, the derivation must be shown to follow from the exact transfer-function phase (Eq. 27), not from the small-epsilon approximation; otherwise the apparent match in Fig. 5 may be deceptive. Please either re-derive Eq. (23) without the small-epsilon assumption, or restrict the quantitative comparison to a narrow band around resonance and justify the validity outside it.","section":"Eq. (19) and Sec. V B 1"},{"comment":"The Vd(VDC) curves in Fig. 6(a) are explicitly stated to be dominated by a VDC-dependent dissipative contribution that is not described by the model and is formally absorbed into k_int,1^(d). The inverse-heterodyne contribution is isolated only by subtracting the Umod = 0 curve (Fig. 6b). This subtraction assumes the background dissipation is independent of Umod, but the same k_int,1^(d) term is later invoked to explain the Umod=0 offset in Fig. 7(a). The assumption is thus load-bearing for the claimed quadratic-in-Umod scaling. Please provide a test of this assumption, e.g., a Vd(VDC) measurement at Umod=0 as a baseline at the same zc, or a model for the VDC dependence of k_int,1^(d).","section":"Sec. V B 2 and Eq. (23b)"}],"minor_comments":[{"comment":"The signs of the (1-u2^2) term and the 'changes sign at resonance' discussion should specify that the sign change occurs at the interaction-shifted resonance ef2,0(zc), not at the free resonance f2,0.","section":"Sec. IV C"},{"comment":"The theoretical curves are 'scaled by prefactors' but the scaling is not defined; please specify the exact normalization so the reader can compare magnitudes.","section":"Fig. 5 caption"},{"comment":"When plotting square roots of differences of Vd curves, the quantity under the square root may be negative on one side of VCPD; please specify whether absolute values or only positive branches are used, and add error bars or uncertainty estimates.","section":"Sec. V B 2/3 and Eqs. (28)-(29)"},{"comment":"Table I lists VDC = +200 mV and Umod = 200 mV, but the f2-sweep experiment uses VDC = +1 V and Umod = 1 V. Please clarify which parameters are nominal and which are changed in specific measurements.","section":"Table I and Sec. V B 1"},{"comment":"The statement that z2,0 is estimated from the Fourier peak ratio z1,0/68 is acknowledged as calibration-uncertain. This is acceptable because the main predictions for Fd and Delta-f1 do not depend on z2,0, but the uncertainty should be propagated if z2,0 values are quoted quantitatively.","section":"Sec. V experimental setup"},{"comment":"The companion manuscript is listed as 'submitted to APS Open Sci'; if available as a preprint, a link should be provided, and key results should be summarized in the SI for the present paper to be self-contained.","section":"Reference [25]"}],"recommendation":"major_revision","confidential_remarks":"The dimensional inconsistency in Eq. (23) is the most serious concern: it affects the quantitative magnitude of the central prediction, even though the line shapes and sign-changing features are unaffected. The reliance on the companion for K0,K1,K2 and the unverified quasi-quadrature condition are additional risks. I recommend asking the authors to correct the equations, clarify the derivation, and provide the companion's key results or a self-contained derivation in the SI before acceptance."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Here's the honest take. This paper reports a real effect—inverse heterodyne back-action in open-loop AM-He-KPFM—and derives closed-form expressions for the first-mode frequency shift, drive force, second-mode amplitude, and phase. The two sharpest predictions are a dissipation peak proportional to |G2|^2 that is 1/sqrt(3) narrower than the second-mode transfer function, and a sign-changing (1-u2^2) contribution to Δf1. Both are derived, not fitted, and both match the UHV data. That is a genuine step forward, and the preceding empirical reports of intermode energy transfer [20,21] are properly acknowledged as lacking analytical support.\n\nThe main soft spot is exactly where the authors themselves put it (Sec. VII and ref. [25]): the effective capacitance-gradient coefficients K0, K1, K2, which set the force amplitudes α1, α2, are derived and validated in a concurrently submitted companion manuscript that is not available here. The FOTR expressions in Eq. (9b–d) are explicitly truncated; their convergence and validity for the close-approach f2 sweep (Δf1 = −100 Hz) are not demonstrated in this paper. If the companion treatment is inaccurate, the quantitative strength and distance dependence in Eq. (23) change. The qualitative line-shape predictions survive, since they follow from the transfer-function structure rather than the prefactor, so the existence claim is not at risk. The stress-test concern about ZOTR/FOTR at close approach is legitimate for the same reason.\n\nSecond, Eq. (19) imposes a quasi-quadrature condition |ε| << 1, yet Eq. (23) is compared with data across the full f2 sweep, where ε is not small away from resonance. I would like to see an independent derivation from Eq. (21) using the exact transfer-function phase, or a plot of ε over the sweep. This is a real but not fatal issue.\n\nThird, the experimental reporting is thinner than I would like: no error bars in the key figures, data available only on request, and the z2,0 calibration is admitted to be an order-of-magnitude estimate. The authors also note a dominant VDC-dependent dissipation left in k(d)_int,1 and unmodeled. They isolate the inverse effect by subtraction, which is reasonable, but it weakens the quantitative case.\n\nBottom line: this deserves a serious referee, not a desk reject. The algebra is internally consistent, the sharpest predictions are derived and match the data, and the limitations are stated rather than hidden. Acceptance should be conditional on access to the companion manuscript or a self-contained derivation of Ki, plus deposited data with error bars and a direct check of the quasi-quadrature assumption.","headline":"The inverse heterodyne back-action is real and the line-shape predictions are sharp, but the quantitative core leans on an unavailable companion paper.","tokens_in":842,"tokens_out":1013,"would_cite":true,"duration_ms":33320,"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":"Open-loop amplitude-modulated heterodyne KPFM sustains an inverse heterodyne effect: the second eigenmode feeds back onto the first, predominantly through dissipation.","keywords":["Kelvin probe force microscopy","heterodyne","bimodal AFM","inverse heterodyne effect","inter-mode coupling","dissipation","capacitance gradient","frequency shift"],"falsifier":"A decisive check would be to measure the first-mode drive force while sweeping the demodulation frequency across the second-mode resonance and compare the full width at half maximum of the dissipation peak to 1/sqrt(3) times the width of the directly measured second-mode amplitude resonance; if the width ratio differs significantly from 1/sqrt(3), the inverse heterodyne back-action is not the dominant mechanism.","tokens_in":26370,"feed_emoji":"🔬","tokens_out":7484,"duration_ms":69855,"temperature":0.7,"pith_summary":"Open-loop amplitude-modulated heterodyne Kelvin probe force microscopy (AM-He-KPFM) is shown to intrinsically support an inverse heterodyne effect: the electrostatically sustained second cantilever eigenmode feeds back onto the first eigenmode, altering both its dissipation and frequency shift. Combining a bimodal virial and power-balance framework with a non-truncated capacitance-gradient expansion, the paper derives closed-form expressions for the first-mode drive and frequency shift, including contributions proportional to the squared second-mode transfer function and to that same square times (1-u2^2). The theory predicts that the inverse effect appears mainly in the dissipation channel, with a resonance peak about 1/sqrt(3) narrower than the second-mode amplitude resonance, and a weaker, sign-changing contribution to the frequency shift. Ultrahigh-vacuum experiments sweeping the demodulation frequency, DC bias, and modulation amplitude validate these predictions and isolate the inverse heterodyne signature. The result matters because it identifies an intrinsic inter-mode energy-transfer process that must be considered when interpreting dissipation-based contrast and quantitative electrostatic measurements in heterodyne KPFM.","feed_headline":"Inverse heterodyne effect couples KPFM eigenmodes","feed_subtitle":"Heterodyne-driven second mode alters first-mode dissipation and frequency shift, with a resonance peak 1/√3 narrower.","key_machinery":"The central object is the interaction-shifted second-eigenmode transfer function G2(omega2) and the dimensionless frequency ratio u2 = omega2 / omega-tilde_2,0, which together shape the inverse heterodyne terms: |G2|^2 governs the dissipation back-action and |G2|^2*(1-u2^2) gives the sign-changing conservative back-action. The argument is carried by inserting the non-truncated capacitance-gradient expansion (effective coefficients K0, K1, K2) into the bimodal electrostatic force, then projecting the total force onto the virial and power-balance equations of each eigenmode.","core_discovery":"The central claim is that in open-loop AM-He-KPFM, heterodyne frequency conversion does more than excite the second eigenmode; it also produces a back-action force at the first eigenmode frequency. That back-action enters the first-mode power balance as a term proportional to alpha1*alpha2*|G2(omega2)|^2, making the required drive force increase with a resonance peak at the interaction-shifted second-mode frequency, and enters the first-mode virial as a term proportional to alpha1*alpha2*|G2(omega2)|^2*(1-u2^2), which changes sign across that resonance. Here alpha1 and alpha2 are electrostatic force amplitudes that vanish when the applied DC bias equals the contact potential difference or wh","pith_inferences":["The line-shape predictions (|G2|^2 and (1-u2^2)) are independent of the specific capacitance-gradient coefficients; if the companion's coefficients were found to be inaccurate, the strength and distance dependence would change but the resonance and sign-change signatures would survive, providing a way to separate the two parts of the theory.","The dissipation-channel readout may offer a sensitive way to locate the interaction-shifted second-mode resonance even when the second-mode amplitude signal is weak, since the inverse effect amplifies via the square of the transfer function.","The paper's experiments show an additional VDC-dependent dissipative contribution not captured by the model; identifying its microscopic origin (e.g., non-contact friction from surface charges) could refine quantitative dissipation-based KPFM.","The effect could be tested in other multimode systems: any two coupled oscillators with a nonlinear force and frequency conversion should show an analogous back-action peak in the drive of the lower mode, scaled by the square of the higher-mode transfer function."],"forward_implications":["When sweeping the demodulation frequency around the second-mode resonance in open-loop AM-He-KPFM, the first-mode drive force will show a narrow resonance peak with a full width at half maximum about 1/sqrt(3) times that of the second-mode amplitude resonance, directly revealing the inverse back-action.","The first-mode frequency shift will show a smaller, dispersive-like signature that changes sign at the interaction-shifted second-mode resonance, acting as an apparent stiffening or softening of the first mode.","The inverse effect makes both the drive force and frequency shift depend quadratically on (VDC minus V_cpd) and on U_mod, so voltage- and modulation-dependent measurements must account for it to recover the true capacitance-gradient signal.","In ideal closed-loop KPFM, the inverse effect vanishes because alpha1 and alpha2 go to zero, so the bimodal force expressions reduce to the usual monomodal ones; but any residual (VDC minus V_cpd) will reintroduce back-action.","The same back-action mechanism transfers to dual-heterodyne KPFM and heterodyne photo-induced force microscopy, where a similar dissipation resonance should appear."],"fun_headline_variants":["Inverse heterodyne back-action alters KPFM dissipation","Heterodyne coupling flips first-mode energy flow in KPFM","Second eigenmode feed-back reshapes KPFM frequency shift","Bimodal KPFM reveals inverse heterodyne energy exchange"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The load-bearing premise is that the effective capacitance-gradient coefficients (K0, K1, K2) from the companion manuscript are correct, and that the phases satisfy the quasi-quadrature condition |epsilon|<<1; if either fails, the quantitative predictions lose accuracy, although the sign-changing and squared-transfer-function line shapes may persist.","fun_headline_variants_meta":{"raw":{"variants":["Inverse heterodyne back-action alters KPFM dissipation","Heterodyne coupling flips first-mode energy flow in KPFM","Second eigenmode feed-back reshapes KPFM frequency shift","Bimodal KPFM reveals inverse heterodyne energy exchange"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000488,"raw_usage":{"total_tokens":2292,"prompt_tokens":846,"completion_tokens":1446,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":590,"completion_tokens_details":{"reasoning_tokens":1383}},"tokens_in":590,"tokens_out":1446,"duration_ms":10714,"temperature":1.0,"reasoning_tokens":1383,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-02T08:18:36.068155+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A decisive check would be to measure the first-mode drive force while sweeping the demodulation frequency across the second-mode resonance and compare the full width at half maximum of the dissipation peak to 1/sqrt(3) times the width of the directly measured second-mode amplitude resonance; if the width ratio differs significantly from 1/sqrt(3), the inverse heterodyne back-action is not the dominant mechanism.","supporting_citations":[],"review_version":2}