{"id":"3912014a-dd95-4a6f-b76c-33ad9c681fd9","arxiv_id":"1908.05512","paper_version":4,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"high","formal_verification":"none","parameter_count":2,"one_line_summary":"An open-circuit, time-resolved conductive AFM method separates triboelectric and piezotronic currents and measures the direct axial piezoelectric coefficient of single GaAs nanowires (0.4-1 pm/V).","lead":"The authors show that the standard way of measuring electric current from squeezed nanowires mostly picks up surface-contact artifacts, not the material's piezoelectric response. They introduce an open-circuit atomic force microscopy approach that extracts the true piezoelectric coefficient, demonstrated on gallium arsenide nanowires.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The rate-independent intercept in Eq. 14 is not established as purely piezoelectric; the manuscript's own oxide control is 'lower, though comparable' and is not subtracted, so the quoted d33 may be dominated by flexoelectric or triboelectric background.","rationale":"My reading agrees with the reader's weakest_assumption: the load-bearing step is the claim that the rate-independent intercept in Eq. 14 is piezoelectric. The manuscript's own control paragraph concedes that native SiO2 gives a comparable electromechanical signal, attributed to flexoelectricity, and that signal is not removed from the NW numbers. Since d33 scales linearly with q_PE, the background directly endangers the headline coefficient. The absence of a release-current sign reversal is a second, explicitly acknowledged inconsistency, reinforcing the same concern. The qualitative message that standard short-circuit cAFM is dominated by non-piezoelectric mechanisms is independently supported by the Eq. 13 order-of-magnitude estimate, the bias-dependent maps, and the topography-current mismatch, so the paper retains real value. The CONDITIONAL verdict remains appropriate: the method concept is reasonable, but the quantitative extraction needs either a proper subtraction of the control background or another unambiguous isolation of piezoelectric charge. I do not move the verdict; hence UNCHANGED.","tokens_in":13914,"tokens_out":6040,"duration_ms":61242,"concrete_test":"Re-analyze the Fig. 7 data by applying the same Eq. 14 fit to the native-oxide control points, extract q_PE,oxide, and subtract it from each NW intercept before converting to d33. If the corrected d33 falls below roughly 0.2 pm/V or changes by more than 50%, the reported 0.4-1 pm/V range is not supported by the data.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The quantitative claim depends on the decomposition in Eq. 14, Q = a/rate + q_PE, where q_PE is interpreted as the rate-independent piezoelectric charge. No derivation of the 1/rate parasitic term is given, no fit statistics are reported, and nothing in the model excludes other rate-independent charge sources. The manuscript's control experiment on native SiO2 (Fig. 7 and the control paragraph in the Open-circuit configuration section) produced an electromechanical signal that was 'lower, though comparable' to the NW signal and was attributed to flexoelectricity; it is never subtracted from the NW intercepts. Because the conversion to d33 is directly proportional to q_PE (through q_PE/c_glass = V, E = V/L, T = F/A, and d = g*epsilon0*epsilon_r), a comparable flexoelectric or triboelectric background can substantially inflate or even fully account for the reported 0.4-1 pm/V. The paper itself also flags the absence of an opposite-polarity current on release as 'intriguing'; that is inconsistent with a purely conservative piezoelectric response unless the release transient is separately accounted for. The qualitative claim that short-circuit cAFM is unreliable is well supported, but the quantitative d33 values are not robust until the rate-independent background is isolated.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript proposes an open-circuit conductive atomic force microscopy (cAFM) methodology for extracting the axial piezoelectric coefficient of individual nanowires. The authors first argue, with an order-of-magnitude estimate (Eq. 13), that short-circuit cAFM cannot measure the direct piezoelectric current from a single GaAs nanowire because the expected charge is orders of magnitude below the current noise. They then introduce an open-circuit configuration in which a glass slide capacitor is placed in series with the sample, and the measured current reflects the time derivative of the voltage. By integrating the current during the loading ramp and plotting the integrated charge versus ramping rate, they fit the data to Q = a/rate + q_PE (Eq. 14) and interpret the rate-independent intercept q_PE as the piezoelectric charge. From q_PE they compute the voltage, stress, and finally d33 = 0.4–1.05 pC/N for GaAs nanowires, which they claim is in good agreement with literature values. They also report control experiments on native oxide, ITO, and undoped substrates to argue that the signal is not an artifact.","tokens_in":14129,"tokens_out":8388,"duration_ms":69031,"significance":"The short-circuit inadequacy argument is well founded and important: it provides a clear explanation for the long-standing controversy around single-NW current generation. The open-circuit configuration and time-resolved analysis are a promising and easy-to-implement alternative. If the quantitative extraction is validated, the method would allow direct piezoelectric coefficient measurements on a single nanowire, which is currently difficult. The paper also sheds light on the coexistence of piezoelectric, triboelectric, and piezotronic effects in cAFM. However, the quantitative claim is not yet fully supported because the background separation in Eq. (14) is not independently validated.","major_comments":[{"comment":"The decomposition Q = a/rate + q_PE is assumed without derivation, and the identification of the rate-independent intercept q_PE with the piezoelectric charge is not validated. The manuscript's own control on the native oxide (Fig. 7) produced an electromechanical signal 'lower, though comparable' to the NW signal, which the authors attribute to flexoelectricity; this signal is not subtracted from the NW intercepts. Since d33 is computed as d33 = g33 ε0 εr with g33 = (q_PE / c_glass / L) / (F/A), a comparable rate-independent background can substantially inflate or dominate the reported d33 values. To support the central quantitative claim, the authors should subtract the control intercept from the NW measurements or provide a quantitative estimate showing that the flexoelectric/triboelectric contribution is negligible for the NW geometry (e.g., using a non-piezoelectric nanowire of similar dimensions).","section":"Open-circuit configuration, Eq. (14) and Fig. 7"},{"comment":"The fits to Eq. (14) are not described with any statistics. The number of data points per NW, the scatter, and the confidence intervals on the fitted parameters a and q_PE are not reported, so the reader cannot assess whether the linear-in-1/rate model is appropriate or whether the intercept is well constrained. Please include the fit parameters, uncertainties, and R² or similar goodness-of-fit values for each nanowire and for the oxide control.","section":"Open-circuit configuration, Fig. 7 and Eq. (14)"},{"comment":"The paper states that the absence of an opposite-polarity current on release is 'intriguing' but does not resolve this issue. For a conservative piezoelectric response, unloading should produce a current of opposite sign. The integration in Eq. (14) only considers the loading (charging) portion of the ramp, so the release transient is not analyzed. This is a load-bearing point because the measured charge could include non-conservative (e.g., triboelectric or flexoelectric) contributions that are not reversed on release. The authors should either analyze the release portion of the data or provide a model explaining why no opposite current is observed.","section":"Time-resolved current measurements, p. 15"}],"minor_comments":[{"comment":"The numerical values E1300nN = 2.4 V/m and T1300nN = 258 N/m² are inconsistent with the stated geometry (0.505 V across 210 nm gives 2.4 MV/m; 1300 nN on a 40 nm radius gives 258 MPa). The ratio g33 is correct, but the intermediate values should be corrected to avoid confusion.","section":"Open-circuit configuration, Eq. (15)"},{"comment":"The claim of 'good agreement' with known GaAs values (1.5–2.5 pC/N, Ref. 38) is weakened by the fact that the measured range (0.4–1.05 pC/N) is several times lower; a more nuanced discussion (e.g., in terms of bending loss) is already present but could be expanded.","section":"Results and discussion, Open-circuit configuration"},{"comment":"The abstract uses 'pm/V' while the manuscript body uses 'pC/N'; these are equivalent for d33 but should be used consistently.","section":"Abstract and body"},{"comment":"The affiliation 'Cambrdige' (p. 1) should be corrected to 'Cambridge'.","section":"Title page"},{"comment":"In the caption of Fig. 5, 'the current axis for (e) and (d) are not identical' should read 'the current axes in (d) and (e) are not identical'.","section":"Fig. 5 caption"},{"comment":"The reference to the 'known' GaAs piezoelectric coefficients comes from the authors' own previous PFM work (Ref. 38). If possible, please include an independent experimental or computational reference to strengthen the comparison.","section":"References"}],"recommendation":"major_revision","confidential_remarks":"The quantitative claim depends on Eq. (14) and the background separation. The missing fit statistics and the lack of control subtraction are essential. In addition, the comparison value for GaAs d33 comes from the authors' own prior PFM paper (Ref. 38), so the 'agreement' is partly a self-comparison; the editor may want an independent verification. The paper is otherwise within the scope of the journal."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nRead this one if you care about nanoscale electromechanical measurements. The paper's strongest contribution is a quantitative argument that the standard short-circuit cAFM geometry cannot detect piezoelectric currents from single nanowires: for GaAs, d33 ~2 pm/V at 1300 nN gives Q ~3e-18 C, and even a fast ramp yields ~3e-17 A, well below the ~100 fA noise floor. That is a clean, correct order-of-magnitude result, and it makes sense of the contested nanogenerator literature.\n\nThe new measurement idea is also good. Putting a glass slide between sample and stage turns the circuit into a voltage differentiator, so integrating the measured current gives charge without needing a low-noise current amplifier. Time-resolved ramps separate the triboelectric hot-spots (instant current, no deformation correlation) from the top-of-NW signal that tracks force. The control experiments on ITO and undoped-substrate GaAs show the signal is not an artifact.\n\nThe problem is the extraction. The fitted form Q = a/rate + q_PE is assumed, not derived; no fit statistics or residuals are shown, so q_PE could be absorbing a curved background. More importantly, the native-oxide control produced an electromechanical signal 'lower, though comparable' to the NW signal. It is attributed to flexoelectricity and never subtracted. Since d33 is directly proportional to q_PE, a comparable flexoelectric background can easily account for the entire 0.4-1 pm/V. The authors acknowledge the missing opposite-polarity current on release but do not resolve it. And one of the three NW values is quietly left out ('the higher value for NW 2'), which does not inspire confidence.\n\nThe circularity is mild, not fatal. The comparison with ref 38, their own PFM study of similar GaAs NWs, is an overlapping-author comparison, but q_PE is not tuned to match, so it is a genuine measurement. Still, the agreement is not independent evidence.\n\nOverall: the qualitative claim—that short-circuit cAFM on NWs is dominated by triboelectric and piezotronic effects—is well supported and likely to stand. The quantitative d33 values are provisional at best. This paper should be sent to peer review; it deserves referee time and would trigger useful discussion. But a responsible referee should ask for a quantitative treatment of the oxide background, a derivation or validation of the 1/rate model, and a proper explanation of the discarded data point before the numbers are accepted.\n\nYours,","headline":"A genuinely useful negative result about short-circuit cAFM, wrapped around a clever open-circuit protocol whose quantitative d33 values are not yet trustworthy.","tokens_in":14764,"tokens_out":3355,"would_cite":true,"duration_ms":32447,"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":"The paper claims that standard short-circuit conductive AFM is inadequate for quantitative piezoelectric measurement of nanowires, and that an open-circuit time-resolved variant extracts a direct axial coefficient, giving $d_{33}\\approx…","keywords":["piezoelectric nanowires","conductive atomic force microscopy","open-circuit measurement","triboelectric effect","piezotronic effect","time-resolved current","gallium arsenide","flexoelectricity"],"falsifier":"Sweep the maximum ramp force $F$ on a single nanowire and check that $q_\\mathrm{PE}$ is proportional to $F$, as piezoelectric charge must be; if the intercept saturates or persists on a non-piezoelectric, non-flexoelectric control under identical tip and force conditions, the extracted $d_{33}$ is not purely piezoelectric. A direct cross-check would be to measure the same nanowire with converse PFM or a known piezoelectric standard and compare $d_{33}$.","tokens_in":13644,"feed_emoji":"⚡","tokens_out":7526,"duration_ms":65858,"temperature":0.7,"pith_summary":"This paper argues that the standard short-circuit conductive AFM measurement of piezoelectric nanowires is not quantitative: the true piezoelectric displacement charge is orders of magnitude below instrument noise, so the currents such experiments record are dominated by triboelectricity and the piezotronic effect. To get around this, the authors place a glass slide capacitor in series with the sample and record the current while ramping the tip into the nanowire at different speeds. Because the capacitor acts as a voltage differentiator, integrating the current gives the charge accumulated on a known capacitor, and fitting that charge as $Q=a/\\mathrm{rate}+q_\\mathrm{PE}$ isolates a rate-independent piezoelectric contribution $q_\\mathrm{PE}$. From it they extract a direct axial coefficient $d_{33}\\approx 0.4$–$1$ pm/V for GaAs nanowires, consistent with known values; if the method holds, a simple AFM modification yields quantitative direct piezoelectric coefficients for weak piezoelectric nanomaterials.","feed_headline":"Open-circuit AFM isolates true piezoelectric signal in nanowires","feed_subtitle":"Short-circuit cAFM currents are mostly triboelectric; a series glass capacitor isolates the true GaAs d33 near 1 pm/V.","key_machinery":"The load-bearing object is the rate-resolved charge fit $Q=a/\\mathrm{rate}+q_\\mathrm{PE}$ (Eq. 14 in the paper). The $1/\\mathrm{rate}$ term is proportional to contact duration and collects parasitic currents that scale with time, while the intercept $q_\\mathrm{PE}$ is a time-independent charge left on the isolating capacitor, identified as the piezoelectric displacement charge. The electrical circuit that makes this possible is the open-circuit voltage differentiator: a microscope glass slide between sample and AFM stage supplies $c_\\mathrm{glass}\\approx 5.3$ pF in series, so the recorded current is $c_\\mathrm{glass}\\,dv/dt$; the nanowire capacitance ($\\sim10^{-18}$ F) and resistance are negligible on the measurement timescale. This converts the impossible task of measuring a femtoamp piezoelectric current into the tractable task of integrating charge on a known macroscopic capacitor.","core_discovery":"The central discovery is that the measurement configuration, not the material, decides whether cAFM sees piezoelectricity. Under short-circuit conditions the piezoelectric element is a current source with $E_3=0$, and the expected current $Q/\\Delta t=d_{33}F/\\Delta t$ for a single GaAs nanowire under realistic force is about $10^{-17}$ A, far below the $\\sim$100 fA noise floor; the currents actually seen are triboelectric or piezotronic. In the open-circuit configuration, the nanowire is a voltage source and the series glass capacitor ($c_\\mathrm{glass}\\approx 5.3$ pF) makes the measured current proportional to the voltage derivative, $i=c_\\mathrm{glass}\\,dv/dt$. Integrating the current during a controlled ramp and separating the charge as $Q=a/\\mathrm{rate}+q_\\mathrm{PE}$ yields $q_\\mathrm{PE}$, from which the piezoelectric voltage coefficient $g_{33}$ and then $d_{33}=g_{33}\\epsilon_0\\epsilon_r$ follow. For GaAs nanowires the result is $d_{33}\\approx 0.4$–$1$ pm/V, somewhat below the $1.5$–$2.5$ pm/V expected from theory, which the authors attribute to partial bending and diameter uncertainty.","pith_inferences":["If the intercept is truly piezoelectric, then $q_\\mathrm{PE}$ should scale linearly with applied force; running the same ramp protocol at several peak forces on one nanowire would turn that implication into a direct validation test.","The oxide-control signal described as 'comparable' sets an upper bound on how much of the nanowire $q_\\mathrm{PE}$ could be flexoelectric; a quantitative subtraction using the control would tighten the $d_{33}$ estimate.","Because the open-circuit differentiator detects any rate-independent charge, the same setup could be used to measure flexoelectric coefficients of non-piezoelectric dielectrics by using tips of different radius to separate strain-gradient from uniform-strain responses.","The rate-dependent slope $a$ could be compared with contact-area and barrier-height changes extracted from I-V curves on the same nanowire, offering a direct probe of the piezotronic contribution."],"forward_implications":["If the claim is right, many short-circuit cAFM current maps reported for piezoelectric nanowires are largely triboelectric or piezotronic, not piezoelectric, and should be reinterpreted.","Single-nanowire direct piezoelectric coefficients can be measured with a standard AFM plus a glass slide, without an external current amplifier.","The rate-resolved fit separates piezoelectric from parasitic mechanisms, so the same dataset can also quantify triboelectric contributions.","The method works for a weakly piezoelectric semiconductor like GaAs, so it should extend to stronger piezoelectric materials and to triboelectric or flexoelectric characterization."],"supporting_citations":[{"why":"the foundational short-circuit cAFM nanogenerator experiment whose interpretation the paper disputes","marker":"(3)"},{"why":"provides the GaAs nanowire cAFM baseline and the laser-induced carrier-generation background","marker":"(17)"},{"why":"the main criticism that nanowire short-circuit current generation is not piezoelectric, which the paper corroborates","marker":"(20)"},{"why":"finite-element calculations showing expected ZnO nanowire generation is too small, reinforcing that short-circuit currents are not piezoelectric","marker":"(21)"},{"why":"the rebuttal defending the nanowire experiments, defining the controversy the paper addresses","marker":"(22)"},{"why":"the prior direct-PFM method that quantitatively extracted direct piezoelectric coefficients, which the paper adapts conceptually","marker":"(24)"},{"why":"the authors' own non-destructive PFM technique that limited direct piezoelectric measurement on non-planar nanowires","marker":"(26)"},{"why":"previous PFM work on III-V nanowires that supplies the known GaAs $d_{33}$ comparison values","marker":"(38)"},{"why":"demonstrates converse flexoelectricity in AFM, used to explain the comparable silicon-oxide control signal","marker":"(50)"}],"fun_headline_variants":["Time-resolved open-circuit cAFM separates piezoelectric from triboelectric","Open-circuit cAFM extracts true GaAs piezoelectric coefficient from nanowires","Short-circuit cAFM sees triboelectric; open-circuit sees true d33","Voltage-sensing cAFM eliminates triboelectric bias in nanowire d33","Open-circuit cAFM quantifies GaAs nanowire piezoelectricity directly"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The extraction assumes that the rate-independent intercept $q_\\mathrm{PE}$ in Eq. (14) is almost entirely piezoelectric charge, so the flexoelectric and triboelectric background seen on the silicon-oxide control—which the paper says was 'lower, though comparable' to the nanowire signal—does not contaminate the nanowire intercept.","fun_headline_variants_meta":{"raw":{"variants":["Time-resolved open-circuit cAFM separates piezoelectric from triboelectric","Open-circuit cAFM extracts true GaAs piezoelectric coefficient from nanowires","Short-circuit cAFM sees triboelectric; open-circuit sees true d33","Voltage-sensing cAFM eliminates triboelectric bias in nanowire d33","Open-circuit cAFM quantifies GaAs nanowire piezoelectricity directly"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000334,"raw_usage":{"total_tokens":1905,"prompt_tokens":1045,"completion_tokens":860,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":661,"completion_tokens_details":{"reasoning_tokens":762}},"tokens_in":661,"tokens_out":860,"duration_ms":8565,"temperature":1.0,"reasoning_tokens":762,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T13:12:20.895329+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Sweep the maximum ramp force $F$ on a single nanowire and check that $q_\\mathrm{PE}$ is proportional to $F$, as piezoelectric charge must be; if the intercept saturates or persists on a non-piezoelectric, non-flexoelectric control under identical tip and force conditions, the extracted $d_{33}$ is not purely piezoelectric. A direct cross-check would be to measure the same nanowire with converse PFM or a known piezoelectric standard and compare $d_{33}$.","supporting_citations":[],"review_version":1}