{"id":"e1b32bfb-a76e-4bcf-934b-d07be5cac323","arxiv_id":"2411.19118","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Direct flexoelectric charge generation is measured for the first time in a 50 nm HfO2 layer, with the extracted coefficient matching earlier inverse-effect values and a width-dependent effective coefficient consistent with tensor-component interplay.","lead":"This paper reports the first direct measurement of flexoelectric charges in a 50-nanometer-thick hafnium oxide layer, using vibrating cantilevers and lock-in detection. It matters because it confirms that flexoelectric transduction, a possible route for nano-electromechanical sensors and actuators, works at the thickness scales used in real devices.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The resonant signal is isolated only by a linear baseline subtraction with no control excluding a motion-modulated parasitic capacitance; a DC-bias sweep across the electrodes would settle whether the extracted peak is flexoelectric or parasitic.","rationale":"The reader's weakest_assumption already identifies the parasitic-crosstalk risk, and my stress-test converges on exactly that point: the linear baseline subtraction cannot reject a motion-synchronized capacitive current, and no control experiment rules it out. The DC-bias sweep is a clean discriminator because a parasitic dC/dt * V term scales with the bias across the time-varying capacitance, whereas the flexoelectric current does not. The width-dependence ratio matching 1/(1-nu) is real supporting evidence and should be credited, but it does not by itself prove the signal origin. Because the missing control is an experimental test that can still be performed, the appropriate disposition is to retain the reader's CONDITIONAL verdict rather than accept the result as definitive. I therefore recommend no change to the verdict.","tokens_in":9791,"tokens_out":10893,"duration_ms":121645,"concrete_test":"Apply a controlled DC bias V_DC across the top and bottom Pt electrodes (e.g., -5 V to +5 V) while keeping the cantilever displacement at resonance fixed by retuning the shaker drive, and record the extracted V_motional after the same linear baseline subtraction. If the resonant peak contains a motion-modulated parasitic capacitance term, its amplitude will grow linearly with V_DC and reverse sign when V_DC is reversed; a purely flexoelectric signal should be independent of V_DC. Repeating this on at least one narrow and one wide cantilever would decisively separate flexoelectric charge from parasitic capacitance modulation. A complementary control is to measure an identical-geometry metal cantilever with no HfO2 layer.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim rests on Eqs. 1.2 and 1.11, which convert the on-resonance voltage V_motional into mu_eff after a linear baseline subtraction of the parasitic background (Section 1.3, Fig. 1.5). That subtraction removes the non-resonant capacitive feedthrough between the piezoelectric shaker and the electrodes, but it cannot remove a background component that is itself modulated by the cantilever motion. If the electrode/substrate or electrode/electrode capacitance C(t) varies at the mechanical resonance and any low-frequency or DC potential exists across that capacitance (amplifier input offset, contact potentials, trapped charge), the resulting current contains a term dC/dt * V at the drive frequency. This term has exactly the same Lorentzian lineshape and the same linear dependence on actuation as the flexoelectric current (Eqs. 1.7-1.9), so a linear baseline subtraction cannot distinguish them. The paper reports no control cantilever without HfO2, no shorted-electrode device, and no DC-bias dependence test. The width-dependence ratio ~1/(1-nu) is a genuinely nontrivial internal check, but it is not a direct isolation of the signal source. The sequential displacement/charge acquisition (stated in Section 1.3) is a secondary concern, since the two measurements are not guaranteed to correspond to the same mechanical state. The result is plausible, but the 'first-ever direct flexoelectric charge measurement' claim is contingent on an untested assumption.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript reports the first claimed direct measurement of flexoelectric charges in nanoscale-thickness materials, using platinum-clamped 50 nm hafnium oxide cantilevers of 20 µm length and widths from 5 to 50 µm. The cantilevers are driven at resonance by a piezoelectric shaker under vacuum; the generated electrical signal is amplified and read with a lock-in amplifier, while the mechanical displacement is measured with a Laser Doppler Vibrometer. From the measured motional voltage, tip displacement, total capacitance, and the derivative of the first mode shape at the tip, the authors compute an effective flexoelectric coefficient using Eq. (1.2)/(1.11). They report that the extracted coefficient agrees with their earlier inverse-effect value of 105 ± 10 pC/m, and that the effective coefficient grows with cantilever width, with the wide-to-narrow asymptotic ratio equal to 1/(1−ν), i.e., about 1.42 for ν = 0.3, consistent with a claimed ~40–42% increase. A supplementary section derives the flexoelectric formula, provides an uncertainty budget, and sketches the tensor-component analysis for narrow and wide cantilevers.","tokens_in":10010,"tokens_out":3231,"duration_ms":32411,"significance":"If the measurement is genuinely flexoelectric, the result is an important step: it would demonstrate direct flexoelectric charge generation in nanometer-thickness dielectrics and strengthen the case for flexoelectric NEMS transduction. The derivations are transparent, the conversion from measured quantities to an effective coefficient is parameter-free in the sense that no target value is fitted, the mode-shape treatment is physically standard, and an uncertainty budget is provided. The width-dependence ratio 1/(1−ν) is a nontrivial internal consistency check that would be difficult to produce from an arbitrary parasitic background. However, the central claim rests on an untested assumption that the resonant electrical signal is flexoelectric after a linear baseline subtraction; the credibility of the 'first-ever' claim therefore depends on control experiments or quantitative bounds that are not yet reported.","major_comments":[{"comment":"The linear baseline subtraction used to isolate V_motional cannot exclude a motion-modulated parasitic capacitance contribution. If the electrode-to-substrate or electrode-to-electrode capacitance C(t) varies at the cantilever resonance and any low-frequency or DC potential exists across it (amplifier input offset, contact potentials, trapped charge), the resulting current contains a term dC/dt·V with the same Lorentzian lineshape and the same linear dependence on shaker actuation as the flexoelectric current of Eqs. (1.7)–(1.9). The manuscript reports no control cantilever without HfO2, no shorted-electrode device, and no DC-bias dependence test. Because the first-ever direct flexoelectric charge measurement is the central claim, this omission is load-bearing; a control experiment or an explicit quantitative bound on the parasitic-modulation pathway is needed.","section":"Section 1.3, Fig. 1.5, Eq. (1.11)"},{"comment":"The displacement and electrical measurements are not acquired simultaneously; the text states that the electrical charges and the vibrometer signal are monitored 'subsequently.' Since the tip displacement u enters μ_eff linearly, any drift, hysteresis, or slight frequency mismatch between the two acquisition steps is an unquantified systematic error. The manuscript should report the repeatability of the displacement and voltage measurements under nominally identical conditions, or a simultaneous acquisition scheme, to justify that u and V_motional correspond to the same mechanical state.","section":"Section 1.3, Eq. (1.11)"},{"comment":"The uncertainty budget (total 8.6–12.6%) accounts for readout, geometric, and capacitance uncertainties but does not include any contribution from the baseline-subtraction procedure or from the possible motion-modulated parasitic capacitance discussed above. The reported total is therefore a lower bound on the measurement uncertainty, not a complete budget. This should be stated explicitly, or the budget should be extended with a systematic term estimated from a control experiment.","section":"Section 1.6.2, uncertainty table"},{"comment":"The claimed 42% increase in μ_eff and its agreement with the predicted ratio 1/(1−ν) ≈ 1.42 are not supported by a quantitative fit or a statistical comparison. Figure 1.6 appears to show scattered data with asymptotic lines drawn by hand, and no error bars or confidence intervals are provided for the ratio. A quantitative fit or at least the number of devices and the standard error of the wide-to-narrow ratio should be reported to substantiate the width-dependence claim.","section":"Figure 1.6 and Table 1.1"}],"minor_comments":[{"comment":"The abstract and title describe 'measurements of flexoelectric-generated charges,' but the supplementary section 1.6.4 states that 'our methodology does not measure flexoelectric charges directly, we measure currents.' Please reconcile the wording so that the physical quantity actually measured (current, or charge inferred from current) is described consistently.","section":"Abstract and Section 1.6.4"},{"comment":"Several equations in the supplementary material are garbled by missing operators and OCR artifacts (e.g., Eq. (1.15) and the block around Eqs. (1.16)–(1.17)). Also, in the uncertainty table, the length L is listed as '5.9 nm − 19.9 nm,' which should be micrometers; please correct the units.","section":"Section 1.6.1 and 1.6.2"},{"comment":"Reference [21] is incomplete: no journal, volume, or DOI is given. If it is an unpublished manuscript or preprint, please indicate its status so readers can locate it.","section":"References"},{"comment":"Figure 1.5 lacks axis labels with units for the voltage and frequency axes, and Figure 1.6 lacks error bars and a statement of how many devices per width were measured. Adding these would improve interpretability.","section":"Figures 1.5 and 1.6"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is within the scope of an applied-physics journal and is concise and readable. The main risk is the absence of a control excluding motion-modulated parasitic capacitance; I would require that control or a quantitative bound before publication. The comparison to the authors' own prior inverse-effect value also reduces the independence of the validation, but this is not disqualifying if the forward measurement is convincingly isolated."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"First direct flexoelectric charge measurement on a 50 nm film, with a nice width-dependent ratio, but the signal isolation relies on a linear baseline subtraction that could be contaminated by motion-modulated parasitic capacitance.\n\nThe genuinely new thing here is the measurement itself: resonant cantilevers with a 50 nm HfO2 layer, voltage and displacement converted to an effective flexoelectric coefficient via a straightforward derivation. The width dependence is a real observation, and the ratio approaching 1/(1-nu) is a nontrivial internal check that matches the data. The agreement with the authors' own prior inverse-effect value of 105 +/- 10 pC/m gives some confidence that they are seeing the same physics.\n\nThe soft spot is the isolation of the flexoelectric signal. In Section 1.3, the resonant peak is extracted by subtracting a linear baseline from the in-phase and quadrature signals. That removes the non-resonant capacitive feedthrough from the shaker, but not a background that is itself modulated by the cantilever motion. If the electrode capacitance varies at resonance and there is any DC bias (amplifier offset, contact potentials, trapped charge), the resulting dC/dt * V term has the same Lorentzian lineshape and same linear dependence on actuation as the flexoelectric current. The paper reports no control cantilever without HfO2, no shorted-electrode device, and no DC-bias sweep. That is a load-bearing assumption, not a fatal flaw. A DC-bias sweep across the electrodes would settle it.\n\nSecondary issues: displacement and charge are measured sequentially rather than simultaneously; the 'first-ever' claim is not reconciled with Bhaskar et al. (ref [3]); the uncertainty table has unit inconsistencies (lengths listed as 5.9 nm-19.9 nm for 20 um cantilevers); and no raw data are provided. These are minor.\n\nOn balance, the paper is worth a serious referee. The derivation is clean, the width ratio is a good internal check, and the claimed first direct nanoscale measurement is important if it holds. I would send it to peer review with a request for the DC-bias sweep and a non-flexoelectric control. The NEMS and flexoelectric communities will want to see this result tested, and the paper provides enough detail to make that feasible.","headline":"First direct nanoscale flexoelectric charge measurement with a plausible width-dependent ratio, but signal isolation rests on an untested parasitic-crosstalk assumption.","tokens_in":10576,"tokens_out":2534,"would_cite":false,"duration_ms":21620,"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":"This paper reports the first direct measurements of flexoelectric-generated charges in nanoscale-thickness materials, using 50 nm hafnium oxide cantilevers, and shows the effective coefficient grows with width according to $1/(1-\\nu)$.","keywords":["flexoelectricity","direct flexoelectric effect","hafnium oxide","nanoscale cantilevers","NEMS","strain gradient","flexoelectric coefficient","MEMS transduction"],"falsifier":"A control experiment with a geometrically identical cantilever lacking the dielectric layer, or with a non-flexoelectric film, run through the same baseline subtraction would settle the matter: if a resonance-peaked electrical signal of comparable size survives, the measured voltage is parasitic rather than flexoelectric. Likewise, recording displacement and charge at the same instant and checking whether the extracted coefficient is unchanged would test the sequential-measurement assumption.","tokens_in":9540,"feed_emoji":"⚡","tokens_out":5319,"duration_ms":46953,"temperature":0.7,"pith_summary":"Flexoelectricity, the generation of polarization by a strain gradient, is expected to strengthen at small scales, but nobody had directly measured the charges it produces in nanometre-thick material films. This paper reports such measurements using microcantilevers with a 50 nm hafnium oxide layer between platinum electrodes, resonantly excited under vacuum, with the resulting motional charges read through a low-noise amplifier and the displacement tracked by laser vibrometry. The extracted effective flexoelectric coefficient agrees with values previously obtained from the inverse effect, and it rises with cantilever width by roughly 40 to 42 percent, matching the theoretical ratio $1/(1-\\nu)$ for the narrow-to-wide transition. The authors take this as evidence that flexoelectric transduction is real and usable at the scale of nano-electromechanical systems.","feed_headline":"First direct flexoelectric charge measured in 50 nm films","feed_subtitle":"Charge from bending a 50 nm hafnium oxide cantilever matches theory and the inverse effect.","key_machinery":"The load-bearing object is a resonant microcantilever made of a 50 nm hafnium oxide layer sandwiched between 20 nm platinum electrodes, driven at its first mode by a piezoelectric shaker. The electrical readout equation is $\\mu_{eff} = C_{meas}\\, V_{motional}/(W\\, \\phi_n'(L)\\, u)$, where $C_{meas}$ is the total capacitance of the measurement path, $V_{motional}$ is the flexoelectric voltage after subtracting a linear background, $W$ is the cantilever width, $\\phi_n'(L)$ is the derivative of the normalized mode shape at the tip, and $u$ is the tip displacement. This equation, together with the asymptotic narrow and wide tensor analysis, is what converts two separately measured quantities, voltage and displacement, into a material coefficient and its predicted geometry dependence.","core_discovery":"The central claim is that the direct flexoelectric effect, meaning charge generated by bending, can be measured in a 50 nm thick dielectric film, and that the effective coefficient extracted from those charges is governed by a width-dependent interplay of flexoelectric tensor components $\\mu_{11}$ and $\\mu_{12}$. For narrow cantilevers with $W\\ll L$, the effective coefficient is $\\mu_{eff,n} = -\\nu\\mu_{11} + \\mu_{12}(1-\\nu)$; for wide cantilevers with $L\\lesssim W$, it is $\\mu_{eff,w} = \\mu_{11}\\nu/(\\nu-1)+\\mu_{12}$. The ratio of the two asymptotic limits is exactly $1/(1-\\nu)$, about 1.42 for hafnium oxide with $\\nu=0.3$, matching the measured increase of roughly 40 percent. The paper also reports that its direct values align with inverse-effect measurements for hafnium oxide, closing the loop between the two manifestations of flexoelectricity at the nanoscale.","pith_inferences":["Because the ratio $1/(1-\\nu)$ is independent of $\\mu_{11}$ and $\\mu_{12}$, the same width-sweep experiment on any dielectric could double as a Poisson's-ratio probe, provided the tensor components stay constant across widths.","The sequential voltage and displacement acquisitions could be upgraded to simultaneous readout; if the extracted coefficient is unchanged, the reported values would be robust against drift between the two measurements.","The same cantilever method could be applied to other CMOS-compatible high-permittivity dielectrics, giving a route to survey nanoscale flexoelectricity in materials beyond hafnium oxide.","Since parasitic capacitance dominates the readout path, moving the amplifier closer or integrating it on chip should push the technique toward thinner or stiffer films with smaller generated charges."],"forward_implications":["Flexoelectric transduction of motion is achievable in NEMS-scale devices without adding a separate piezoelectric layer.","Cantilever geometry, specifically width relative to length, is a design knob for tuning the effective flexoelectric coefficient.","The measured charge levels, on the order of tens to hundreds of equivalent electrons per resonance cycle, give a concrete benchmark for the sensitivity required of readout electronics in flexoelectric NEMS.","The width-dependence ratio $1/(1-\\nu)$ serves as an internal consistency check that the extracted resonant signal is genuinely flexoelectric."],"supporting_citations":[{"why":"Reports the prior inverse-effect measurement of hafnium oxide's effective flexoelectric coefficient that the direct values are compared against.","marker":"[20]"},{"why":"Supplies the theoretical treatment of flexoelectric tensor components whose width-dependent interplay yields the asymptotic formulas and their ratio.","marker":"[4]"},{"why":"Gives the basic constitutive relation $P = \\mu \\, \\partial\\varepsilon/\\partial z$ on which the polarization-to-current derivation rests.","marker":"[5]"},{"why":"Provides the thermomechanical-noise-based method for converting laser vibrometer readings at an arbitrary point to tip displacement.","marker":"[23]"},{"why":"Is the standard cantilever beam model used to translate measured response into mode shapes and tip motion.","marker":"[24]"},{"why":"Specifies the low-noise preamplifier whose input capacitance enters $C_{meas}$ in the extraction formula.","marker":"[22]"},{"why":"Represents the established millimetre-scale direct flexoelectric measurements that this work extends to nanometre thicknesses.","marker":"[1]"},{"why":"Documents the absence of nanoscale direct measurements and motivates the need for them.","marker":"[2]"}],"fun_headline_variants":["First direct flexoelectric charge from 50 nm film","Direct flexoelectric effect measured in nanoscale films","Bending nanoscale hafnia yields direct flexoelectric charge","Nanoscale direct flexoelectricity measured in hafnia","Direct flexoelectric charge measured at nanoscale thickness"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing assumption is that, after a linear baseline subtraction, the resonant electrical signal is entirely flexoelectric charge and contains no motion-modulated parasitic contribution from changing capacitances or electrostatic coupling, and that the displacement and charge measurements, taken sequentially rather than simultaneously, represent the same mechanical state.","fun_headline_variants_meta":{"raw":{"variants":["First direct flexoelectric charge from 50 nm film","Direct flexoelectric effect measured in nanoscale films","Bending nanoscale hafnia yields direct flexoelectric charge","Nanoscale direct flexoelectricity measured in hafnia","Direct flexoelectric charge measured at nanoscale thickness"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000413,"raw_usage":{"total_tokens":2145,"prompt_tokens":965,"completion_tokens":1180,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":581,"completion_tokens_details":{"reasoning_tokens":1113}},"tokens_in":581,"tokens_out":1180,"duration_ms":8426,"temperature":1.0,"reasoning_tokens":1113,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T10:30:55.189783+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A control experiment with a geometrically identical cantilever lacking the dielectric layer, or with a non-flexoelectric film, run through the same baseline subtraction would settle the matter: if a resonance-peaked electrical signal of comparable size survives, the measured voltage is parasitic rather than flexoelectric. Likewise, recording displacement and charge at the same instant and checking whether the extracted coefficient is unchanged would test the sequential-measurement assumption.","supporting_citations":[{"cited_title":"Flexoelectricity in amorphous hafnium oxide (HfO2),","cited_arxiv_id":null,"evidence_quote":"Reports the prior inverse-effect measurement of hafnium oxide's effective flexoelectric coefficient that the direct values are compared against."},{"cited_title":"Piezoelectric Effect during Inhomogeneous Deformation and Acoustic Scattering of Carriers in Crystals,","cited_arxiv_id":null,"evidence_quote":"Gives the basic constitutive relation $P = \\mu \\, \\partial\\varepsilon/\\partial z$ on which the polarization-to-current derivation rests."},{"cited_title":"Nonlinearity in nanomechanical cantilevers,","cited_arxiv_id":null,"evidence_quote":"Provides the thermomechanical-noise-based method for converting laser vibrometer readings at an arbitrary point to tip displacement."},{"cited_title":"Ultra-sensitive Preamplifier,","cited_arxiv_id":null,"evidence_quote":"Specifies the low-noise preamplifier whose input capacitance enters $C_{meas}$ in the extraction formula."},{"cited_title":"Large flexoelectric polarization in ceramic lead magnesium niobate,","cited_arxiv_id":null,"evidence_quote":"Represents the established millimetre-scale direct flexoelectric measurements that this work extends to nanometre thicknesses."},{"cited_title":"Advancements of Flexoelectric Materials and eir Implementations in Flexoelectric Devices,","cited_arxiv_id":null,"evidence_quote":"Documents the absence of nanoscale direct measurements and motivates the need for them."}],"review_version":1}