{"id":"2c3da4d3-a900-45b1-a21c-b2840b37ceca","arxiv_id":"2411.15226","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":5,"one_line_summary":"Pervaporation-driven flow in PDMS artificial leaves generates electrokinetic power through a colloidal plug, with output growing with leaf area until cavitation at zero water pressure stops it.","lead":"The authors show that water pervaporating through PDMS \"artificial leaves\" can pull water through a packed bead plug and generate a small electric current. This is a proof of concept for a new evaporation-driven energy harvesting configuration, but the power is tiny and bubbles at zero pressure set a hard ceiling.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The cavitation threshold claim rests entirely on Eq. 5's indirect leaf pressure; an unvalidated Rh under pervaporation conditions could shift Pleaf and make 'systematic cavitation at 0 bar' a modeling artifact.","rationale":"The reader's weakest-assumption analysis correctly identifies the indirect leaf-pressure estimate, Eq. 5, as the load-bearing element of the cavitation claim. My stress-test agrees and sharpens it: the claim is not merely 'pressure could be different' but that the entire quantitative threshold P_cav ~ 0 bar is unvalidated against a direct measurement, with no error budget on the product RhQ. The paper's own observation of pressure-dependent PDMS deformation weakens the assumption that the system state is fully captured by the pre-measured Rh and the leaf-only Q values. This does not overturn the more robust achievements: the electrical characterization is internally consistent, the collapse in Fig. 7(a) supports electrokinetic conversion, and the linear scaling of Q with leaf area is convincing. The concern is specifically about the second headline claim, the cavitation limit, where the evidence is indirect. Because the reader already set the verdict to CONDITIONAL, my read does not change that verdict: the central claim is plausible but the cavitation threshold should be verified by direct pressure measurement before accepting P_cav ~ 0 as a quantitative result.","tokens_in":14114,"tokens_out":14817,"duration_ms":153084,"concrete_test":"Install a calibrated absolute pressure sensor directly at the leaf inlet (or on the downstream side of the colloid plug) during the Fig. 9 protocol, and compare the measured Pleaf at the moment of bubble appearance with the value computed from Eq. 5 using the pre-measured Rh and flowmeter Q. Repeat with at least one different plug resistance (e.g., Lp ~ 3 cm, Rh ~ 16 bar min/uL) and with thoroughly degassed water. If bubbles appear at a directly measured Pleaf significantly different from 0, or only in non-degassed water, the systematic-cavitation-at-zero claim should be revised.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central 'cavitation limit' result (Sec. III.D, Fig. 9) depends on Eq. 5, Pleaf = Pi - RhQ, which is not a directly measured pressure. Rh = 33 bar min/uL was characterized once at Pi = 8 bar with the plug outlet at atmospheric pressure, then assumed unchanged when the outlet is a dead-end PDMS leaf at Pleaf near 0 bar. The paper itself reports that the PDMS leaf deforms mechanically with Pleaf (Fig. 9 inset), and a small coupled change in the plug or an unaccounted resistance in the flowmeter/connectors would shift the inferred Pleaf at the moment bubbles appear. The claim that cavitation occurs 'systematically as soon as Pleaf ~ 0 bar' therefore rests on the accuracy of the product RhQ to within the claimed repeatability. No error bars are given for Q, Rh, or Pleaf; at Q ~ 4.3 uL/h, a 15% uncertainty in Rh or Q changes Pleaf by roughly 0.35 bar, enough to move the threshold from zero to a clearly positive or negative value. If Pleaf were actually positive, the observed bubbles could be dissolved air coming out of solution rather than vapor cavitation, and the 'intrinsic limit at 0 bar' would be an artifact of the pressure model. The introduction's statement that cavitation occurs at DeltaP ~ 1 bar is also not directly consistent with Fig. 9, where cavitation occurs at DeltaP ~ 2.4 bar (with Pi = 2.4 bar), although this may reflect the design rule for Pi = 1 bar rather than a measurement.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper reports an experimental validation of a series ('upstream conversion') configuration for evaporation-driven electrokinetic energy harvesting. A PDMS chip with 79 parallel dead-end channels serves as an artificial leaf driving a pervaporation flow; the flow passes through a packed polystyrene colloidal plug whose streaming potential is collected across a load resistance. The authors characterize the leaf (flow rate linear in the number of leaves, consistent with the Dollet/Noblin pervaporation model, with an inferred PDMS water transport coefficient q̃ ≈ 0.4 µm²/s matching literature), the plug's hydraulic and electrokinetic parameters (Rh ≈ 1 bar min/µL, Sstr ≈ 25 nA/bar, RC ≈ 25 MΩ, efficiency ε ≈ 0.14%), and the coupled system: pervaporation-driven streaming data collapse onto the same Eq. (4) parameters as mechanically driven flows, and the harvested power reaches Pe ≈ 0.18 nW for five leaves, scaling as Pe = εPh. In a second configuration with a high-resistance plug (Rh ≈ 33 bar min/µL) and imposed inlet pressure Pi, the authors infer the leaf pressure from Pleaf = Pi − RhQ and observe bubble formation when Pleaf ≈ 0 bar, which they attribute to cavitation and present as an intrinsic limit (Qmax = Pcav/Rh) for this configuration.","tokens_in":14481,"tokens_out":18901,"duration_ms":177866,"significance":"The central demonstration is genuinely quantitative and largely free of circularity: the pervaporation-driven data in Fig. 7(a) collapse onto the streaming-potential curve of Eq. (4) using parameters (Sstr, RC) determined from independent mechanically imposed flow experiments, and the design rule Pe = εRhQ² is validated. The inferred q̃ ≈ 0.4 µm²/s provides an external consistency check against the pervaporation literature, and the measurement of ε = Pe/Ph in a passively driven evaporation system requires measuring both Q and ΔP, which few evaporation-harvesting studies achieve. If the cavitation-threshold claim holds, the paper contributes a practically important design constraint. However, the threshold claim is the least supported element: it depends on an indirect pressure estimate with no propagated uncertainties, and the identification of vapor cavitation is not distinguished from dissolved-air exsolution. These issues are fixable with targeted controls; the electrokinetic harvesting result is not affected by them.","major_comments":[{"comment":"The central claim that cavitation occurs 'systematically as soon as Pleaf ≈ 0 bar' rests entirely on the indirect estimate Pleaf = Pi − RhQ (Eq. 5). Here Rh = 33 bar min/µL was measured once, under Pi = 8 bar with the plug outlet at atmospheric pressure, and is assumed unchanged when the outlet is a dead-end PDMS leaf and the flow rate is 5–15 times lower (Q ≈ 4–9 µL/h, i.e., 0.07–0.15 µL/min on a 0–1.5 µL/min flow sensor). The linearity check in Fig. 6(a) was performed on a different plug (Rt = 0.5 mm, Lp ≈ 2 cm), so the pressure independence of Rh for the long, thin plug of Fig. 9 is not directly established. No error bars are reported for Q, Rh, or Pleaf; at Q ≈ 4.3 µL/h, a 15% uncertainty in either Q or Rh shifts Pleaf by ≈0.35 bar, enough to move the threshold from 0 to a clearly positive or negative value. Because the 0-bar threshold is the paper's principal new insight, the authors should propagate measurement uncertainties, verify Rh for the specific plug used in Fig. 9 over the relevant range, and ideally measure Pleaf directly (a control with Pleaf imposed by a pressure controller is mentioned in Sec. III.D, suggesting this is feasible). The stated repetition with 'several different leaves' should also be documented with data.","section":"Sec. III.D, Eq. (5), Fig. 9"},{"comment":"The bubbles observed at Pleaf ≈ 0 bar are attributed to vapor cavitation, but the experiments do not exclude exsolution of dissolved air, and the identification matters for the 'intrinsic limit' conclusion in Sec. IV. In the Fig. 9 protocol the water is held at Pi = 8 bar for several hours in contact with the gas phase of the pressure controller before the ramp; water equilibrated at 8 bar contains up to about eight times its 1-atm dissolved-gas concentration, so after decompression the leaf, at Pleaf ≈ 0 (1 atm absolute), is strongly supersaturated and air exsolution there is expected. This would produce bubbles at approximately Pleaf ≈ 0 gauge without any cavitation, consistent with the observation that the threshold is 'systematically' zero across repeated runs, whereas true vapor cavitation at 1 atm absolute would require heterogeneous nucleation under essentially zero tension. A concrete discriminating test is to repeat the pressure ramp with thoroughly degassed water, and/or to probe the gas content of the bubbles (for example, their growth/dissolution response to small pressure changes). If the bubbles are dissolved gas, the limit is not intrinsic and the concluding claim that 'whatever the cavitation threshold... cavitation always limits' the configuration would need to be substantially qualified.","section":"Sec. III.D and Sec. IV (Conclusion)"}],"minor_comments":[{"comment":"The paper states that cavitation occurs when the pressure drop reaches ΔP ≈ 1 bar, but the experiment in Fig. 9 shows bubble formation at Pi = 2.4 bar with Pleaf ≈ 0, i.e., ΔP ≈ 2.4 bar; the '1 bar' statement is only the design point for Pi = 1 bar and should be rephrased accordingly.","section":"Sec. I and Sec. IV"},{"comment":"Typo: 'more informations' should be 'more information'; also, 'Carman-Koseny' (Sec. III.B, Eq. 3) is conventionally spelled 'Kozeny-Carman'.","section":"Sec. III.A"},{"comment":"No accuracy specifications are given for the flow meters, and the Fig. 9 flow rates (0.07–0.15 µL/min) lie near the bottom of the XS sensor's 0–1.5 µL/min range; the manufacturer's accuracy or an in-house calibration should be reported, since Q enters the central pressure estimate of Eq. (5).","section":"Sec. II.B and Fig. 9"},{"comment":"The uncertainty in RC (25 ± 5 MΩ) and the fact that all load resistances satisfy RL < RC are acknowledged, but the corresponding uncertainty in ε and thus in the predicted power line of Fig. 7(b) is not propagated; a short propagation of errors would strengthen the quantitative claim.","section":"Sec. III.B and Fig. 7(b)"},{"comment":"Rather than only visually comparing the pervaporation-driven data with the Eq. (4) curve fitted to mechanically driven data, fitting the pervaporation data alone (or reporting residuals) would quantify the claimed collapse and the 'same parameters' statement.","section":"Fig. 7(a)"}],"recommendation":"major_revision","confidential_remarks":"The stress-test concern about Eq. (5) does land: the cavitation threshold is the least supported element, and the dissolved-gas channel I emphasize (8-bar equilibration followed by decompression) strengthens the case for a required control. That said, the core electrokinetic demonstration is sound, independent, and within the journal's scope; the required fixes (uncertainty propagation, a direct Pleaf measurement or a degassed-water control, and rephrasing of the 'intrinsic limit' claim) are proportionate to a major revision and do not require redoing the central experiment. I also checked the citation pattern: the self-citations are to the authors' own methodological prior work (pervaporation modeling, microfluidic evaporation review) and are appropriate."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nThe paper is a genuine proof of concept: it is the first quantitative demonstration that the separated driver/converter configuration Yaroshchuk proposed actually works. The key evidence is Fig. 7: the streaming potentials measured under pervaporation collapse onto the same Eq. 4 curve as the mechanically driven calibration, and the inferred PDMS permeability q~ matches literature values. The scaling of flow rate and power with the number of leaves is clean. That part is solid.\n\nThe cavitation limit is a nice observation but softer. It rests entirely on Eq. 5, Pleaf = Pi - RhQ, with Rh measured once at 8 bar and assumed unchanged. The plug is rigid, so that assumption is reasonable, but there are no error bars on Q or Rh, and the leaf itself visibly deforms. A 15% uncertainty in either would shift Pleaf by ~0.35 bar, which is enough to make 'systematically at 0 bar' less sharp. The paper also says the limit is at DeltaP ~ 1 bar in the intro/conclusion, but the actual cavitation event in Fig. 9 occurs at DeltaP ~ 2.4 bar; that's a wording inconsistency, not a fatal one, since the physical claim is Pleaf ~ 0.\n\nThe dissolved-air alternative is not fully excluded. The high-pressure pre-fill should have removed some gas, but the authors don't report a control with degassed water or a direct leaf pressure measurement. So I read the cavitation threshold as plausible, not proven.\n\nThe omitted salt-concentration screening is disclosed, which is honest. I would have liked the raw data with error bars, but for a proof of concept the current level of detail is acceptable.\n\nWho is this for? Researchers working on evaporation-driven energy harvesting, microfluidic passive pumping, and anyone using PDMS pervaporation as a pump. The paper deserves a serious referee: the central demonstration is reproducible and the cavitation claim can be tested with relatively simple additional experiments. I would send it to review, and ask the authors to address the pressure estimate and the DeltaP wording in revision.","headline":"First clean proof-of-concept of the series config; the cavitation limit is plausible but needs a direct pressure measurement.","tokens_in":14991,"tokens_out":3539,"would_cite":true,"duration_ms":32994,"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 reports the first experimental evidence that pervaporation-driven flow through a colloidal plug generates electrical power, with output rising with evaporation area until cavitation in the PDMS leaf stops it.","keywords":["electrokinetic energy harvesting","pervaporation","PDMS microfluidics","streaming potential","cavitation","artificial leaves","colloidal plug","evaporation-driven flow"],"falsifier":"Embed a miniature pressure sensor or pressure tap directly inside the PDMS leaf while recording flow rate and cavitation onset; if bubbles first appear when the directly measured leaf pressure is well below 0 bar, or if the pressure estimated from $P_i - R_h Q$ disagrees with the sensor at onset, the claim that cavitation is triggered exactly at $P_\\text{leaf}\\simeq 0$ fails. Alternatively, vary $R_h$ systematically and check whether cavitation always occurs at the same estimated $P_\\text{leaf}$ rather than at a fixed $Q$ or $\\Delta P$.","tokens_in":13929,"feed_emoji":"💧","tokens_out":6296,"duration_ms":60010,"temperature":0.7,"pith_summary":"The paper sets out to prove that a passive evaporation-driven flow can be converted into electrical power when the conversion element sits upstream of the evaporating surface, and that the harvested power grows with the evaporation area until the water inside the evaporating element cavitates. The authors build PDMS artificial leaves whose pervaporation pulls water through a packed colloidal plug, generating a streaming potential. They report the first experimental evidence of electrokinetic conversion from pervaporation-induced flows, with the output following $P_e = \\varepsilon R_h Q^2$ as leaves are added in parallel. The key limit is intrinsic: cavitation in the PDMS leaf occurs whenever the leaf pressure estimated by $P_\\text{leaf} = P_i - R_h Q$ approaches $0$ bar, stopping the flow and capping the harvestable power.","feed_headline":"Evaporating PDMS leaves generate electricity until water cavitates","feed_subtitle":"First proof that pervaporation-driven flow can yield electricity, limited by cavitation in the leaf.","key_machinery":"The two building blocks are the pervaporation-driven leaf and the colloidal-plug converter. The leaf is a PDMS chip with $N=79$ parallel dead-end microchannels ($h=30\\,\\mu$m, $w=50\\,\\mu$m, $L=4$ cm) in a $200\\,\\mu$m-thick membrane, whose pervaporation rate is described by a 1D screening limit $Q_\\text{lim} \\simeq (LW/\\delta)\\,\\tilde{q}(1-R_H)$ with a geometric prefactor $\\alpha \\simeq 0.5$ verified by numerical resolution. The converter is a close-packed plug of charged polystyrene colloids of hydraulic resistance $R_h$ that produces a streaming current $S_\\text{str}\\Delta P$ and is described by the equivalent-circuit relation $V = R_C R_L S_\\text{str}\\Delta P/(R_C+R_L)$. The design identity that carries the argument is $P_e = \\varepsilon R_h Q^2$: the power scales with the square of the passively driven flow rate. Cavitation is the counter-term: when $P_\\text{leaf} = P_i - R_h Q$ crosses zero, bubbles nucleate in the PDMS leaves and terminate the flow.","core_discovery":"On its own terms, the paper claims that the series configuration, with the evaporator upstream and the converter downstream, works and can be scaled by area. Connecting one, two, and five PDMS pervaporation leaves in parallel to a colloid-plug converter produced streaming potentials whose normalized response $V/\\Delta P$ versus load resistance collapses onto the same curve as mechanically imposed flows, with conversion efficiency $\\varepsilon \\simeq 0.14\\%$ and an output power of about $0.18$ nW for five leaves. Because the measured flow rate scales linearly with the pervaporation area and the hydraulic resistance of the plug is fixed, the output follows $P_e = \\varepsilon R_h Q^2$, so adding leaves increases harvested power. The increase stops when the estimated water pressure inside the leaf reaches $P_\\text{leaf} \\simeq 0$ bar: bubbles form, pervaporation stops, and the flow rate drops to zero. This is presented as the first direct demonstration of electrokinetic energy harvesting from pervaporation-driven flows, together with the design rule that the maximum sustainable flow is set by the cavitation threshold of the driving element, $Q_\\text{max} = P_\\text{cav}/R_h$.","pith_inferences":["Editorial extension: because the load resistance used ($R_L \\le 10$ M$\\Omega$) stayed below the converter's internal resistance ($R_C \\simeq 25$ M$\\Omega$), the reported $\\varepsilon \\simeq 0.14\\%$ understates the element's efficiency; matching the load would raise the efficiency but would not alter the cavitation limit.","Editorial extension: the paper's pressure model could be checked directly by embedding a pressure sensor in the leaf; if the leaf deforms under tension, $R_h$ changes and the statement that cavitation occurs exactly at $P_\\text{leaf}\\simeq 0$ would need revision.","Editorial extension: the same series architecture suggests a general materials rule, pair a high-$R_h$ converter with a driving material whose cavitation pressure is as negative as possible, and predicts that per-area output saturates once $Q = P_\\text{cav}/R_h$ is reached."],"forward_implications":["Enlarging the pervaporation surface or raising the converter's hydraulic resistance increases harvested power quadratically in $Q$, so the series architecture is a viable route to scale up evaporation-driven harvesting.","In any passive series system, the maximum sustainable flow is $Q = P_\\text{cav}/R_h$; cavitation in the flow-driving element is an intrinsic limit, not a defect of this particular chip.","Hydrogel-based leaves, being hydrophilic and capable of stable negative pressures, could shift $P_\\text{cav}$ below 0 bar and thereby increase the harvestable power.","The collapse of the pervaporation-driven data onto the mechanically driven characterization curve confirms that the same electrokinetic conversion mechanism is at work in both cases."],"supporting_citations":[{"why":"Establishes the streaming-current and streaming-potential mechanism that the paper's conversion element relies on.","marker":"[1]"},{"why":"Critical review that identifies the series configuration and the $P_e = \\varepsilon R_h Q^2$ scaling, which this work tests experimentally.","marker":"[29]"},{"why":"Provides the parallel dead-end channel venation model and the pervaporation saturation limit used to design the PDMS leaves.","marker":"[35]"},{"why":"Supplies the exact conformal-mapping solution for pervaporation-driven flow in a single rectangular dead-end channel, giving the geometrical factor used in the leaf design.","marker":"[38]"},{"why":"Introduces the strategy of using a packed colloid plug as the electrokinetic conversion element.","marker":"[36]"},{"why":"Supports the nanofluidic electrokinetic behavior of nanoparticle crystals, the physical basis of the colloid-plug converter.","marker":"[37]"},{"why":"Reports a much more negative cavitation threshold for a defect-free cavity, providing the contrast that shows why the PDMS leaf cavitates near zero pressure.","marker":"[42]"},{"why":"Describes how surface defects can trap cavitation precursors, explaining the systematic cavitation observed in microfabricated PDMS leaves.","marker":"[43]"},{"why":"Models the kinetic competition of bulk, surface, and surface-defect cavitation nucleation, supporting the interpretation of the observed cavitation threshold.","marker":"[44]"}],"fun_headline_variants":["PDMS 'artificial leaves' harvest electricity from evaporation until bubbles stop it","First electrokinetic power from pervaporation, capped by cavitation","Evaporation-driven power in PDMS chips: area scales, cavitation caps","Pervaporation powers microchip electricity, then cavitation kills flow","Artificial leaves generate electricity, but bubbles limit the harvest"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The claim that cavitation always happens exactly at $P_\\text{leaf} \\simeq 0$ bar rests on the assumption that leaf pressure is accurately given by $P_\\text{leaf} = P_i - R_h Q$ with the separately measured $R_h$ unchanged during pervaporation and under negative pressure; if channel deformation changes $R_h$, the apparent threshold could be an artifact of the pressure estimate.","fun_headline_variants_meta":{"raw":{"variants":["PDMS 'artificial leaves' harvest electricity from evaporation until bubbles stop it","First electrokinetic power from pervaporation, capped by cavitation","Evaporation-driven power in PDMS chips: area scales, cavitation caps","Pervaporation powers microchip electricity, then cavitation kills flow","Artificial leaves generate electricity, but bubbles limit the harvest"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000271,"raw_usage":{"total_tokens":1674,"prompt_tokens":1035,"completion_tokens":639,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":651,"completion_tokens_details":{"reasoning_tokens":547}},"tokens_in":651,"tokens_out":639,"duration_ms":6782,"temperature":1.0,"reasoning_tokens":547,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T15:40:09.675995+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Embed a miniature pressure sensor or pressure tap directly inside the PDMS leaf while recording flow rate and cavitation onset; if bubbles first appear when the directly measured leaf pressure is well below 0 bar, or if the pressure estimated from $P_i - R_h Q$ disagrees with the sensor at onset, the claim that cavitation is triggered exactly at $P_\\text{leaf}\\simeq 0$ fails. Alternatively, vary $R_h$ systematically and check whether cavitation always occurs at the same estimated $P_\\text{leaf}$ rather than at a fixed $Q$ or $\\Delta P$.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Establishes the streaming-current and streaming-potential mechanism that the paper's conversion element relies on."},{"cited_title":"Yaroshchuk ,\\ title title Evaporation-driven electrokinetic energy conversion: Critical review, parametric analysis and perspectives , \\ @noop journal journal Adv","cited_arxiv_id":null,"evidence_quote":"Critical review that identifies the series configuration and the $P_e = \\varepsilon R_h Q^2$ scaling, which this work tests experimentally."},{"cited_title":"Noblin , author L","cited_arxiv_id":null,"evidence_quote":"Provides the parallel dead-end channel venation model and the pervaporation saturation limit used to design the PDMS leaves."},{"cited_title":"Dollet , author J.-F","cited_arxiv_id":null,"evidence_quote":"Supplies the exact conformal-mapping solution for pervaporation-driven flow in a single rectangular dead-end channel, giving the geometrical factor used in the leaf design."},{"cited_title":"Saha , author J","cited_arxiv_id":null,"evidence_quote":"Introduces the strategy of using a packed colloid plug as the electrokinetic conversion element."},{"cited_title":"Chen , author Y","cited_arxiv_id":null,"evidence_quote":"Supports the nanofluidic electrokinetic behavior of nanoparticle crystals, the physical basis of the colloid-plug converter."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Reports a much more negative cavitation threshold for a defect-free cavity, providing the contrast that shows why the PDMS leaf cavitates near zero pressure."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Describes how surface defects can trap cavitation precursors, explaining the systematic cavitation observed in microfabricated PDMS leaves."},{"cited_title":"Water cavitation results from the kinetic competition of bulk, surface and surface-defect nucleation events","cited_arxiv_id":"2410.17626","evidence_quote":"Models the kinetic competition of bulk, surface, and surface-defect cavitation nucleation, supporting the interpretation of the observed cavitation threshold."}],"review_version":1}