{"id":"08d816b4-54b3-422c-bb51-f37632a872f9","arxiv_id":"1908.00560","paper_version":1,"verdict":"REJECT","confidence":"MODERATE","novelty_score":4.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"Molecular dynamics simulations of supercooled water droplets on grooved graphite show penetration followed by retraction, with a claimed velocity threshold and a dominant temperature effect on retraction time.","lead":"Using molecular dynamics simulations, this paper studies supercooled water droplets striking graphite surfaces with nanoscale grooves, and reports that the droplets penetrate the grooves and then retract. It claims that below a threshold impact speed penetration stops, and that droplet temperature, not speed or groove size, controls how long retraction takes.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Abstract's no-penetration threshold is contradicted by the paper's own static reference; v* is only the onset of enhanced penetration, not the onset of penetration.","rationale":"In good faith, the paper's intended contribution is a threshold impact velocity v* below which a droplet does not penetrate nanogrooves, with retraction time governed mainly by temperature. For that claim to hold, the static reference droplet should show zero penetration into the grooves. The paper's own analysis, however, treats the non-impacted droplet as having a nonzero maximum penetration depth: Section 3.2 defines initial penetration as the maximum H from Fig. 5 and Figure 8 subtracts the \"non-impinged droplet\" MDP values, which is only meaningful if those values are positive. The text also states that velocity profiles at 0.5M and 1.0M show \"no significant different\" from the static case, implying that these impinged droplets penetrate to at least the static depth. Thus the abstract's absolute \"penetration is not observed\" is an overstatement; v* as operationalized is the onset of enhanced penetration beyond static wetting, not the onset of penetration itself. This is an internal inconsistency, not merely an underspecified parameter. The reader's weakest assumption, the Mach-number velocity initialization, is a serious reproducibility concern, but it is secondary: even with a fully specified conversion from Mach number to molecular velocities, the central threshold claim would still be contradicted by the paper's own static reference and subtraction procedure. Our concern is therefore more directly load-bearing for the central claim, and it supports the REJECT verdict without changing it. A single quantitative check on the static H values would settle whether the abstract's wording is misleading or the entire threshold picture requires revision.","tokens_in":8340,"tokens_out":6662,"duration_ms":60493,"concrete_test":"Read the static (non-impinged) wetting profiles in Fig. 5 (black triangle data) and compute the maximum H using Eq. (4) for d=2.0nm and 4.7nm at 250K and 265K. If H_static > 0, the abstract's \"below v* penetration is not observed\" is contradicted. Then compare H(t) for 0.5M and 1.0M runs against H_static(t); if they coincide, the threshold is only an onset of enhanced penetration, not of penetration. This check requires only the plotted data or a single rerun of the static NVT simulation described in Section 2, and settles whether the central claim is misstated.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim, stated in the abstract, is: \"There is a threshold value for the value of impinging velocity, v*, below which penetration is not observed which depends on size of the surface roughness.\" This is not what Section 3.1 actually establishes. The results define v* as the velocity at which a \"fundamental change in wetting behavior\" occurs, and state that wetting profiles for velocities equal to or less than 1.0M show \"no significant different\" from each other. Section 3.2 defines \"initial penetration\" as \"the maximum H value represented in figure 5\" and, in Figure 8, plots retracting time against initial penetration \"relative to the non-impinged droplet\", explicitly subtracting the MDP of the non-impacted droplet. Therefore the non-impacted (static) droplet itself has a nonzero maximum penetration H. Since low-velocity impinged droplets are said to resemble the static wetting profile, they too penetrate into the grooves. Consequently, \"penetration is not observed\" below v* is false on the paper's own terms; v* is at best the onset of a noticeable increase in maximum penetration depth relative to static wetting. The practical inference that grooves smaller than a certain size \"can stop water from penetration\" at v~0.5M is not supported, because at 0.5M the droplet penetrates to at least the static equilibrium depth.","agreement_with_reader":"disagree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This manuscript reports molecular dynamics simulations of supercooled water nanodroplets impinging on a grooved graphite substrate, using the TIP4P/Ice water model and OPLS carbon parameters. The authors vary the groove spacing (d = 2.0 nm and 4.7 nm), the droplet temperature (250 K and 265 K), and an impact velocity expressed as Mach numbers (0.5M–2.5M). They compute a density-weighted penetration depth H, obtain wetting profiles over time, and define a critical impact velocity v* = 1.5M where wetting behavior changes significantly. They report that the droplet retracts from the grooves after maximum penetration, that retraction time depends mainly on temperature, and that grooves smaller than a certain size can stop water penetration at low impact velocities.","tokens_in":8662,"tokens_out":6967,"duration_ms":64455,"significance":"If the threshold claim were fully supported, the work would provide a useful atomistic-scale design rule for ice-phobic nanostructured surfaces. The study has several strengths: it uses a water model appropriate for supercooled conditions, calibrates the water–carbon interaction to reproduce the experimental contact angle, and systematically compares static and impinged droplets. The observation of retraction from grooves and the temperature dependence of retraction time are plausible and potentially interesting. However, the headline conclusion about a penetration threshold is over-stated relative to the evidence presented, and the impact velocity initialization is not defined well enough to ensure reproducibility. These issues can likely be fixed by rewriting the claims and adding technical details.","major_comments":[{"comment":"The abstract's claim that below v* 'penetration is not observed' is inconsistent with the manuscript's own data and analysis. Section 3.2 and Figure 8 explicitly subtract the maximum penetration depth of the non-impinged droplet from the impinged values, and Section 3.1 states that wetting profiles for velocities equal to or less than 1.0M are similar to the static wetting profile. Since the static droplet has a nonzero H, low-velocity droplets do penetrate the grooves; v* is at best a threshold for enhanced penetration, not for the onset of penetration. This misstatement affects the abstract, the design recommendation, and the conclusion.","section":"Abstract and Section 3.1"},{"comment":"The impinging velocity is defined only as 'Much numbers (M)' added to thermal velocities, with no speed of sound, no conversion formula, no specification of the direction of the collective velocity, and no information on how the NVT thermostat interacts with this imposed velocity. Consequently, the magnitude of the impact velocities is not defined and the simulations are not reproducible. The authors should provide actual velocities in m/s for each M value and a detailed initialization protocol.","section":"Section 2"},{"comment":"Equation (3) as printed reads ε_c−o = sqrt(ε_c−c + ε_o−o), but the values in Table 3 (0.50840 kJ/mol for the mixing rule) correspond to the geometric mean sqrt(ε_c−c · ε_o−o). A reader implementing the printed formula would obtain an interaction strength roughly a factor of two too large, which would change the predicted contact angle. The equation should be corrected to use the product of the well depths.","section":"Equation (3) and Table 3"},{"comment":"The threshold v* = 1.5M is introduced as a 'fundamental change' without a quantitative criterion. To support a threshold claim, the authors should define an operational measure (e.g., a minimum increase in MDP over the static value) and apply it consistently to both groove sizes and temperatures.","section":"Section 3.1"}],"minor_comments":[{"comment":"The term 'Much numbers' should be 'Mach numbers' throughout the manuscript.","section":"Section 2"},{"comment":"The statement that 'there have so far been no MD simulation incorporated to investigate the dynamic wetting of impinged water droplet on solid surfaces' is incorrect; many prior MD studies of droplet impact on solid surfaces exist. This sentence should be revised.","section":"Section 1 (Introduction)"},{"comment":"Reference 12 (Park et al., J. Hydrol. 2011) is unrelated to wetting and should be replaced with a proper citation for the effect of interfacial energy on wettability.","section":"References"},{"comment":"There are several typographical errors, including 'imping velocity' instead of 'impinging velocity' and 'no significant different' instead of 'no significant difference.'","section":"Throughout"},{"comment":"The phrase 'retracting as a function of \"pure\" depth of penetration' is unclear; please define 'pure' as the relative MDP used in Figure 8.","section":"Section 3.2"},{"comment":"In Equation (5), please clarify that L_y and L_z are the groove dimensions, not the simulation box dimensions.","section":"Equation (5)"}],"recommendation":"major_revision","confidential_remarks":"The manuscript needs a substantial rewrite of the abstract and conclusions to align with the actual measurements. The reference list contains a clearly irrelevant citation (ref. 12), and the novelty claim in the Introduction is overstated. Since the central physical scenario is plausible and the technical issues appear fixable, I recommend major revision rather than rejection."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Here's what I'd tell you about this one. The abstract promises a sharp threshold below which water doesn't penetrate grooves in graphite. That claim is stronger than what the simulations show. Their own static-droplet results give a nonzero penetration depth, and they subtract that static reference in Figure 8. So at 0.5M the droplet penetrates about as much as a non-impacted droplet; 'penetration is not observed' is just wrong. What the data actually support is that 1.5M is the impact speed above which extra penetration becomes noticeable. That's a softer and less interesting result.\n\nWhat's genuinely in the paper: they model supercooled water with TIP4P/Ice, which is a reasonable choice, and they calibrate the water-carbon interaction to match the 90-95° contact angle instead of just taking mixing rules. The qualitative result that colder droplets take longer to retract from grooves is plausible and consistent with higher viscosity. That part is worth keeping.\n\nThe soft spots are real and in proportion: (1) the velocity initialization is underspecified — 'impinging velocity in terms of Mach numbers' with no sound speed or conversion; that makes the whole velocity series hard to reproduce. (2) No replicate runs, no error bars, just single trajectories. (3) There are citation problems — reference 12 is about rainfall, not wetting — and equation (3) has the mixing rule printed with a plus sign instead of multiplication. These are fixable but they add up.\n\nWho is this for? Someone in the anti-icing niche looking for design hints might read the temperature-dominance trend. But I wouldn't rely on any threshold number from this paper.\n\nI'd suggest the authors fix the abstract, define the impinging velocity properly, and re-run with at least three independent trajectories per condition. As it stands, I would not send this to peer review; a desk reject with an invitation to resubmit after major revision would be the honest path.","headline":"A plausible MD parameter study whose headline threshold claim is contradicted by its own static reference and whose impinging-velocity protocol is underspecified.","tokens_in":9108,"tokens_out":2935,"would_cite":false,"duration_ms":28673,"reading_group":"no","serious_thinker":"yes","would_accept_peer_review":false},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"On a graphite surface cut with nanoscale grooves, water droplets that hit slowly enough never enter the grooves, and once they do, temperature controls how quickly they pull back.","keywords":["molecular dynamics","supercooled water","nanodroplet impact","wetting","nanostructured graphite","penetration threshold","ice-phobic surfaces","TIP4P/Ice"],"falsifier":"Repeat the simulations with the droplet given a physical centre-of-mass velocity, computed from the speed of sound in supercooled water, and with a thermostat that removes the added kinetic energy; if penetration appears below 1.5M or the threshold shifts with the initialization scheme, the claimed v* is an artifact of the Mach-number boost.","tokens_in":8141,"feed_emoji":"💧","tokens_out":7812,"duration_ms":78813,"temperature":0.7,"pith_summary":"Using molecular dynamics, this paper asks whether supercooled water nanodroplets can be kept out of nanoscale grooves carved into graphite. It reports a critical impinging velocity, identified in its simulations as about 1.5 Mach, below which no water penetrates the groove and above which penetration grows with impact speed and groove width. The droplets that do enter eventually retract, and the time they need to pull back is governed mainly by droplet temperature rather than by impact velocity or groove size. If these findings hold, groove geometry could be engineered so that water stays on the surface at practical impact speeds, a design consideration for ice-phobic nanostructured substrates.","feed_headline":"Nanogrooves keep water out below a threshold impact speed","feed_subtitle":"Simulations show groove size sets the cutoff while temperature, not speed, controls how fast droplets retreat.","key_machinery":"The argument is carried by the penetration depth H, a density-weighted measure of how far water oxygen atoms have entered the groove (Eq. 4), tracked over 10 ns simulations of a supercooled TIP4P/Ice water droplet striking fixed zig-zag graphene pillars with groove widths d = 2.0 nm and 4.7 nm. A collective impinging velocity expressed in Mach numbers (M, multiples of the speed of sound, though the paper does not state which sound speed is used) is imposed on the thermal velocities of the equilibrated droplet; the threshold where wetting behavior fundamentally changes is then identified as 1.5M, below which penetration is not observed. Retraction is quantified by comparing dynamic and static wetting profiles to define a retracting time. The force-field calibration, especially the water-carbon Lennard-Jones parameters drawn from Werder et al., sets the hydrophobicity that drives retraction.","core_discovery":"The central claim is that dynamic wetting of a nanostructured graphite surface has a velocity threshold. In the simulations, impinging speeds of 0.5M and 1.0M produce wetting profiles indistinguishable from each other and from the static case, while 1.5M marks a fundamental change: above it, maximum penetration depth grows with both impinging velocity and groove width (2.0 nm versus 4.7 nm). Below the threshold, grooves are not penetrated, and the authors infer that grooves smaller than a size-related cutoff can stop water ingress even at speeds around 0.5M, attributing the resistance to interfacial tension. The time required for the droplet to retract from the groove is then shown to depend most strongly on temperature, with 250 K droplets retracting markedly more slowly than 265 K droplets, an effect the authors trace to the TIP4P/Ice model's viscosity behavior.","pith_inferences":["A testable extension is that the critical velocity should scale with interfacial tension divided by groove width, reflecting a race between impact inertia and capillary resistance; the paper does not derive this scaling.","The Mach-number boosts used here correspond to speeds of hundreds of metres per second, far above typical practical droplet impacts, so the design rule that narrower grooves block water may hold for much wider grooves at real-world speeds.","The strong temperature control of retraction time points to shear viscosity of supercooled water as the underlying transport property; one could test this by varying viscosity independently of temperature in the water model.","If the threshold is real, groove width could be chosen as a design parameter for anti-icing surfaces, but this would need confirmation at larger scales and with other water models before becoming an engineering rule."],"forward_implications":["Below the critical impinging velocity, water does not enter surface grooves, so nanostructured groove geometry can act as a barrier to wetting at impact speeds near 0.5M.","Above the threshold, maximum penetration depth increases with both impinging velocity and groove width, so larger textures are more easily wetted under fast impacts.","Retraction time is controlled mainly by droplet temperature, meaning colder supercooled droplets linger longer inside grooves regardless of impact speed.","After a sufficient retracting time, the dynamic wetting profile approaches the static wetting profile, so the impact outcome eventually relaxes to equilibrium wetting.","The calibrated hydrophobic graphite surface reproduces experimentally reported contact angles of 90-95 degrees, supporting the use of this model for ice-phobic surface design."],"supporting_citations":[{"why":"Provides the TIP4P/Ice water model whose accurate melting point makes the supercooled-droplet results trustworthy.","marker":"[13]"},{"why":"Sets the oxygen-carbon Lennard-Jones parameters that reproduce the 90-95 degree contact angle on graphite, calibrating the hydrophobicity that drives retraction.","marker":"[6]"},{"why":"Confirms the melting temperatures of TIP4P/Ice used to justify the 250 K and 265 K supercooled states.","marker":"[18]"},{"why":"Previous fracture study supplying the Lorentz-Berthelot mixing rule used for one set of water-graphite interaction parameters.","marker":"[22]"},{"why":"Earlier impact-ice work from which the collective impinging-velocity initialization is carried over.","marker":"[32]"},{"why":"The LAMMPS molecular dynamics package that runs all simulations.","marker":"[23]"},{"why":"Particle-particle particle-mesh method used for long-range electrostatic interactions between water molecules.","marker":"[24]"},{"why":"SHAKE algorithm that keeps water molecules rigid during the impact simulations.","marker":"[25]"}],"fun_headline_variants":["Nanogrooves block water ingress below a critical speed threshold","Groove size sets the impact velocity cutoff for water penetration","Temperature, not speed, controls droplet retraction from nanogrooves","Supercooled droplets stay out of nanogrooves below a speed cutoff"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing assumption is that adding a single speed expressed as a Mach number to every water molecule's thermal motion is a well-defined model of droplet impact, since the paper gives no sound speed, conversion formula, or control of the added kinetic energy.","fun_headline_variants_meta":{"raw":{"variants":["Nanogrooves block water ingress below a critical speed threshold","Groove size sets the impact velocity cutoff for water penetration","Temperature, not speed, controls droplet retraction from nanogrooves","Supercooled droplets stay out of nanogrooves below a speed cutoff"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000416,"raw_usage":{"total_tokens":2126,"prompt_tokens":904,"completion_tokens":1222,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":520,"completion_tokens_details":{"reasoning_tokens":1147}},"tokens_in":520,"tokens_out":1222,"duration_ms":11855,"temperature":1.0,"reasoning_tokens":1147,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T15:47:24.387813+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Repeat the simulations with the droplet given a physical centre-of-mass velocity, computed from the speed of sound in supercooled water, and with a thermostat that removes the added kinetic energy; if penetration appears below 1.5M or the threshold shifts with the initialization scheme, the claimed v* is an artifact of the Mach-number boost.","supporting_citations":[],"review_version":1}