{"id":"5c3689eb-11f5-47c7-b07d-58997a99ea89","arxiv_id":"2607.07363","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"unknown","formal_verification":"none","parameter_count":3,"one_line_summary":"Probe-particle-model simulations show that constant-force AFM isosurfaces preserve molecular framework and recover tilt angles for non-planar molecules better than constant-height imaging.","lead":"This paper simulates constant-force AFM images of non-planar molecules and shows they retain structural information lost in conventional constant-height imaging. It matters for researchers trying to image 3D molecular structures on surfaces, where flat-scan methods fail.","discovery_kind":"unclear","skeptic_critique":{"model":"glm-5.2","headline":"Electrostatics omission is a real but unevenly distributed limitation: likely negligible for the quantitative benzene/pyrrole tilt recovery at 250 pN (repulsive regime), but potentially load-bearing for qualitative contrast predictions on ZnTPP and CO-FePc where the distinguishing features are elect","rationale":"The reader correctly identifies the most significant limitation, but overstates its load-bearing nature for the quantitative claims. The 1° tilt recovery for benzene and pyrrole at 250 pN is in the repulsive regime where electrostatics is a minor perturbation; this result is likely robust. The concern is more acute for the qualitative contrast predictions on ZnTPP and CO-FePc, where the distinguishing molecular features (metal center, CO ligand) have significant electrostatic character that is entirely absent from the simulation.\n\nThe paper is honest about its scope ('short-range force contrast arising from the molecular structures') and about the 2H-TPP saddle geometry limitation (5° error). The C60 results are qualitative but consistent with the overall framework. The lack of shipped code is a minor reproducibility concern given that the extraction algorithm is described clearly.\n\nThe CONDITIONAL verdict is appropriate. The paper provides useful target contrasts for the AFM community within its stated scope, but the omission of electrostatics for metal-containing molecules means the specific qualitative predictions for ZnTPP and CO-FePc should be treated as preliminary until validated with a more complete force model. The core finding — that force isosurfaces retain structural information lost in constant-height imaging — is sound and does not depend on electrostatics being included, since it is a geometric consequence of following the force field rather than scanning at fixed height.","tokens_in":9703,"tokens_out":5095,"duration_ms":303179,"concrete_test":"Re-run the ZnTPP simulation (saddle geometry) with electrostatic interactions included in the PPM, using point charges on the Zn center and coordinating nitrogens from a standard DFT-derived charge partitioning (e.g., Bader or RESP charges). Extract the 250 pN force isosurface and compare it to the LJ-only result in Figure 5. If the central cross-like feature changes qualitatively — in height, lateral extent, or shape — the electrostatics omission is load-bearing for the metalloporphyrin claims and the paper should include at least one electrostatics-included case to validate its target contrasts.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The reader correctly identifies the omission of electrostatic interactions as the paper's most significant simplification. However, the concern's severity varies across the paper's claims. The quantitative tilt recovery (1° for benzene and pyrrole, Figure 3) uses a 250 pN repulsive threshold, placing the isosurface at ~2.5–3 Å from the nearest atoms where Pauli/LJ repulsion dominates and electrostatic contributions from a weakly polar probe (CO-terminated tip) are small perturbations. These results are likely robust.\n\nThe concern lands more forcefully on the qualitative claims for ZnTPP (Figure 5) and CO-FePc (Figure 6). For ZnTPP, the paper claims the Zn atom 'produces a raised central feature' with 'contrast extending towards the four surrounding nitrogen atoms,' giving a 'cross-like structure.' But the Zn²⁺ center and coordinating nitrogens carry significant partial charges. At the distances where 250 pN repulsion is felt (~2.5–3 Å), the electrostatic force between a charged metal center and a CO-terminated tip (which has a permanent dipole of ~0.1 D and polarizability) could be comparable to the LJ contribution. A rough estimate: a charge of +1e interacting with an induced dipole on the probe at 3 Å gives forces on the order of 50–200 pN, which is not negligible relative to 250 pN. If electrostatics were included, the central feature could become more or less pronounced, or shift laterally, changing the predicted 'target contrast.' Similarly, for CO-FePc, the CO ligand's dipole (~0.1 D) and the Fe center's charge state could modify the isosurface shape around the protruding ligand, potentially altering the claim that constant-force imaging 'reveals submolecular contrast within the phthalocyanine core.'\n\nThe paper provides no sensitivity analysis or bound on these effects. The framing ('based on the short-range force contrast arising from the molecular structures') is honest about the scope but does not address whether the specific qualitative contrasts predicted","agreement_with_reader":"partial"},"referee_report":{"model":"glm-5.2","summary":"This manuscript uses Probe Particle Model (PPM) simulations to extract force isosurfaces from 3D force fields above molecules of increasing structural complexity, from tilted benzene and pyrrole to 2H-TPP, ZnTPP, CO-FePc, and C60. The central claim is that constant-force (force-isosurface) imaging retains submolecular and 3D structural information that is lost in constant-height imaging of non-planar molecules. For tilted benzene and pyrrole, the authors show quantitative recovery of tilt angles to within 1 degree. For larger systems, they identify characteristic contrast patterns associated with different adsorption geometries, central metal features, and curved molecular surfaces. The progression from simple to complex systems is well-designed, and the honest reporting of the 5-degree systematic underestimation for the 2H-TPP saddle geometry is commendable.","tokens_in":10431,"tokens_out":1437,"duration_ms":250738,"significance":"The paper provides falsifiable, molecule-specific target contrasts for force-isosurface imaging that could guide future 3D force-mapping experiments. The quantitative tilt-angle recovery for benzene and pyrrole (Figure 3, within 1 degree) is a clean validation result. The systematic comparison of constant-height versus constant-force imaging across a range of molecular complexities is a useful contribution to the field. The simulation methodology is standard and sound, and the parameter choices (250 pN threshold, 4 N/m stiffness, CO-like probe) are drawn from established PPM practice. The paper does not ship reproducible code or machine-checked proofs, but the methodology is sufficiently described for replication.","major_comments":[{"comment":"§'Simulating constant-force images': The omission of electrostatic interactions is stated but its implications are not analyzed. This is a load-bearing simplification for the ZnTPP (Figure 5) and CO-FePc (Figure 6) results, where the distinguishing features involve charged or polar species (Zn2+ center, CO ligand dipole). At the ~2.5-3 Å probe-sample separations corresponding to 250 pN repulsion, electrostatic forces from a CO-terminated tip interacting with a +1e metal center are plausibly on the order of 50-200 pN, which is not negligible relative to the 250 pN threshold. The paper should either (a) provide a sensitivity analysis or order-of-magnitude bound on how electrostatics would modify the predicted contrast for ZnTPP and CO-FePc, or (b) explicitly scope the claims for these two systems as purely LJ-based predictions that may not hold quantitatively in experiment. The benzene/pyr","section":null},{"comment":"§'Differentiating between adsorption geometries of 2H-TPP', Figure 3c,d: The 5-degree systematic underestimation of the phenyl-ring angle for the saddle geometry is attributed to 'the influence of the surrounding molecular structure on the local force isosurface.' This attribution is plausible but not demonstrated. A simple test would be to fit the plane to a smaller or shifted region of the phenyl-ring isosurface to check whether the shoulder from the macrocycle is the actual cause, or whether the discrepancy arises from the plane-fitting procedure itself. Without this, the reader cannot assess whether the 5-degree error is a fundamental limitation of force-isosurface imaging for crowded molecules or an artifact of the fitting protocol. This matters because the paper's central quantitative claim — that molecular orientation can be recovered from force isosurfaces — is directly tested by","section":null},{"comment":"§'Metal-centred porphyrins and phthalocyanines', Figure 5: The claim that the Zn atom 'produces a raised central feature' with 'contrast extending towards the four surrounding nitrogen atoms' is presented without comparison to a 2H-TPP constant-force image at the same setpoint. The text references Albrecht et al. [12] showing that CuTPP produces a filled centre compared to 2H-TPP, but the reader cannot visually confirm this distinction from the figures provided, since Figure 4 (2H-TPP) and Figure 5 (ZnTPP) use different colour scales and height ranges. A direct side-by-side comparison with matched scales would substantially strengthen the claim that the central feature is attributable to the metal atom rather than to the saddle geometry itself.","section":null}],"minor_comments":[{"comment":"Figure 2: The line profiles are labelled with force values in pN but the relationship between the profile positions and the molecular structure is not always clear. Adding molecular skeleton overlays or tick marks at atom positions would help readers interpret the profiles.","section":null},{"comment":"Figure 3c,d: The 2H-TPP data points are plotted alongside benzene and pyrrole but use different marker shapes (diamonds vs hexagons/pentagons). A legend entry or caption note clarifying which diamond colour corresponds to which geometry would help readers interpret the plot.","section":null},{"comment":"§'Differentiating between adsorption geometries of 2H-TPP': The phenyl-ring angles for the inverted geometry are reported as 3.9, 3.9, 3.8, and 3.9 degrees with error ±0.5 degrees. It is unclear whether this error is a fitting uncertainty, an RMS residual, or something else. This should be specified.","section":null},{"comment":"§'Metal-centred porphyrins and phthalocyanines': The C60 colour scales in Figure 7 are described as 'focused on the uppermost face' and 'adjusted to lower heights,' but the specific height ranges used for each panel are not stated. This should be specified.","section":null},{"comment":"The paper would benefit from a brief discussion of the experimental accessibility of the simulated force thresholds. The 250 pN setpoint is within the range of PeakForce-style methods, but it would help to note whether such thresholds are achievable under typical UHV NC-AFM conditions.","section":null},{"comment":"Reference [18] is cited for both the CO-FePc experimental work (Chen et al.) and the PPM methodology description. These appear to be different works and should be disambiguated.","section":null}],"recommendation":"major_revision","confidential_remarks":"The reader's report and stress-test note correctly identify electrostatics omission as the key concern. My assessment is that this concern is unevenly distributed: it is likely negligible for the benzene/pyrrole quantitative results (repulsive regime, weakly polar probe) but potentially load-bearing for ZnTPP and CO-FePc qualitative contrast predictions. The paper's claims for these two systems are presented without sufficient hedging. A major revision requesting either a sensitivity bound or explicit scoping of the ZnTPP/CO-FePc claims would address this. The 2H-TPP saddle fitting issue is also worth raising but is less critical since the authors already acknowledge the discrepancy honestly."},"author_rebuttal":{"model":"glm-5.2","summary":"We thank the referee for a careful and constructive report. The referee correctly identifies the main contributions of the paper and raises three substantive points regarding (1) the omission of electrostatic interactions, (2) the untested origin of the 5-degree underestimation in the 2H-TPP saddle geometry, and (3) the lack of a matched-scale comparison between ZnTPP and 2H-TPP. We address each point below.","responses":[{"response":"The referee is correct that electrostatic forces are not negligible at the probe-sample separations corresponding to 250 pN repulsion, and we should have been more explicit about the scope and limitations this introduces. We agree with option (b): the claims for ZnTPP and CO-FePc should be explicitly scoped as Lennard-Jones-based predictions. We will add a paragraph to the methodology section stating that electrostatic contributions from the Zn2+ centre and CO ligand dipole are expected to modify the predicted contrast at the ~2.5-3 Å separations involved, and that the ZnTPP and CO-FePc results should be interpreted as purely LJ-based predictions that identify which structural features are expected to dominate the short-range force contrast. We will also add an order-of-magnitude estimate of the electrostatic force from a CO-terminated tip interacting with a +1e metal centre at these separations to make clear that this is a non-negligible effect. The benzene, pyrrole, 2H-TPP, and C60 results, which do not involve charged or strongly polar species, are less affected by this omission.","revision_made":"yes","referee_comment":"Omission of electrostatic interactions not analyzed, especially for ZnTPP and CO-FePc where charged/polar species are involved."},{"response":"The referee makes a fair point: our attribution of the 5-degree underestimation to the shoulder from the macrocycle is plausible but untested. We will perform the suggested test by fitting the plane to smaller and shifted regions of the phenyl-ring isosurface, specifically excluding the region where the macrocycle shoulder is expected to contribute. This will allow us to determine whether the discrepancy is reduced when the neighbouring structure is excluded, or whether it arises from the plane-fitting procedure itself. We will report the results of this test in the revised manuscript, either confirming or revising our attribution. If the discrepancy persists with reduced fitting regions, we will reframe the discussion to acknowledge that the origin is not fully established.","revision_made":"yes","referee_comment":"The 5-degree underestimation for 2H-TPP saddle geometry is attributed to surrounding molecular structure but not demonstrated; a test fitting to a smaller or shifted region is needed."},{"response":"This is a reasonable request and we will implement it. We will produce a side-by-side comparison of the 2H-TPP and ZnTPP constant-force images at the same 250 pN setpoint with matched colour scales and height ranges, so that the reader can directly assess whether the central raised feature in ZnTPP is attributable to the metal atom rather than to the saddle geometry itself. This comparison will be added as a new figure or panel in the revised manuscript.","revision_made":"yes","referee_comment":"ZnTPP central feature claim lacks a side-by-side comparison with 2H-TPP at the same setpoint and matched colour scales."}],"tokens_in":9622,"tokens_out":722,"duration_ms":136015,"standing_objections":[]},"desk_editor":{"model":"glm-5.2","letter":"The main thing to know: this paper systematically shows that constant-force AFM imaging (via force isosurfaces extracted from simulated 3D force fields) retains structural information that constant-height imaging loses on non-planar molecules. The quantitative tilt-angle recovery for benzene and pyrrole (within 1°, Figure 3) is clean and convincing. The progression from simple tilted molecules to 2H-TPP, ZnTPP, CO-FePc, and C60 is well-designed and gives the AFM community useful target contrasts for force-feedback experiments. The honest reporting of the 5° systematic underestimation in the 2H-TPP saddle geometry is a good sign — they don't paper over it. The CO-FePc result, where constant-force imaging recovers the phthalocyanine core hidden behind the protruding CO ligand, is a genuinely nice demonstration. What is new here relative to Schuler et al. [10] and Hapala et al. [6] is the systematic head-to-head comparison of constant-height vs constant-force across a series of molecules with increasing non-planarity, plus the quantitative angle recovery from force isosurfaces. That is a real contribution. The soft spot is the omission of electrostatics, and the reader is right to flag it. But the severity is uneven. For benzene and pyrrole at 250 pN (repulsive regime, ~2.5–3 Å), electrostatic contributions from a CO-terminated probe are small perturbations to the Pauli/LJ-dominated force. The 1° tilt recovery is probably robust. The concern lands harder on ZnTPP and CO-FePc, where the distinguishing features involve charged metal centers and polar ligands. A rough estimate puts electrostatic forces at 50–200 pN at these distances — not negligible relative to 250 pN. The qualitative claims (Zn produces a raised central feature; constant-force reveals the FePc core) are likely directionally correct because they rest on geometry, not electrostatics. But the specific contrast patterns — the cross-like structure around Zn, the depressions in the isoindole sections — could shift if electrostatics were included. The paper provides no sensitivity analysis or bound on this. Minor issues: no code or data shipped, and the C60 results are purely qualitative with no quantitative metrics. Neither is fatal for a simulation study of this type, but shipping the extraction code would be straightforward and would help. Overall: the core findings hold within the scope of what is simulated. The paper is written for AFM experimentalists who need predicted contrasts for force-mapping experiments, and it delivers that. It deserves a serious referee. The referee should push the authors to include electrostatics for at least one molecule (ZnTPP is the obvious candidate) and to discuss how it would affect the qualitative contrasts they predict.","headline":"Solid simulation study with a real but unevenly distributed electrostatics gap; deserves a serious referee","tokens_in":10584,"tokens_out":1500,"would_cite":true,"duration_ms":123381,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"glm-5.2","headline":"Constant-force AFM recovers 3D molecular structure lost in conventional imaging","keywords":[],"falsifier":"If adding electrostatic interactions to the probe-particle-model simulations substantially changes the shape of the force isosurface for pyrrole, ZnTPP, or CO-FePc — for instance by shifting the contour above polar regions enough to obscure the lower-lying features the paper claims are revealed — then the predicted contrasts and tilt-angle recovery would not hold for these molecules in real experiments.","tokens_in":9950,"feed_emoji":"🔬","tokens_out":1003,"duration_ms":127938,"temperature":0.7,"pith_summary":"The paper argues that extracting a force isosurface — the set of points in space where the tip–sample interaction force equals a chosen threshold — from a simulated 3D force field above a molecule preserves structural information that conventional constant-height imaging loses when the molecule is non-planar. In constant-height AFM, the tip scans a fixed geometric plane; for tilted or 3D molecules, features lying below that plane vanish from the image while features above it may destabilise the tip. The force isosurface instead lets the tip rise and fall to follow the molecular force field, maintaining contact with both high and low parts of the structure simultaneously. Using probe-particle-model simulations with Lennard–Jones interactions, the authors show that for tilted benzene and pyrrole the full molecular ring remains visible across tilt angles from 0 to 30 degrees and the tilt angle can be recovered to within 1 degree by fitting a plane to the isosurface. For larger non-planar systems — two adsorption geometries of 2H-TPP, ZnTPP, CO-FePc, and C60 — the force isosurface distinguishes conformations, reveals lower-lying structural features obscured by protruding ligands, and resolves contrast across curved molecular surfaces. The paper provides simulated target contrasts that experimentalists can compare against when performing 3D force-mapping or force-feedback measurements.","feed_headline":"Constant-force AFM recovers 3D molecular structure lost in conventional imaging","feed_subtitle":"Force isosurfaces preserve tilt angles, hidden ligands, and curved molecular faces that fixed-height scans miss — if electrostatics don't干扰","key_machinery":"The Probe Particle Model simulates a flexible CO-like tip apex interacting with molecular coordinates through Lennard–Jones potentials. A 3D force field is computed on a grid above each molecule. A custom extraction algorithm searches each vertical column from large tip–sample separation toward the molecule and records the highest z-position where the force threshold (typically 250 pN) is exceeded, producing the outermost force isosurface. Plane fitting to selected regions of this isosurface recovers molecular tilt angles.","core_discovery":"The central object is the force isosurface: the height contour defined not by a fixed plane but by a constant force threshold in the 3D tip–sample interaction field. The paper demonstrates that this isosurface retains submolecular and three-dimensional structural information — molecular rings, tilt angles, central metal atoms, lower-lying cores beneath protruding ligands, and adjacent faces on curved molecules — that is progressively lost in constant-height images as molecular non-planarity increases. For simple tilted fragments the isosurface recovers orientation to within 1 degree; for complex molecules it distinguishes adsorption geometries and reveals features otherwise hidden, though it","pith_inferences":[],"forward_implications":["Experimental 3D force-mapping protocols can use these simulated isosurfaces as reference templates to verify that measured force contours contain genuine structural information rather than artefacts of tip relaxation or feedback instability.","Force-feedback AFM modes (PeakForce, off-resonance) that already acquire force curves during imaging could extract isosurface contrast directly, potentially bringing 3D molecular structure determination to ambient or room-temperature conditions without requiring full UHV frequency-shift tomography.","For molecules with protruding functional groups — axial ligands, metal centres, or adsorbed species — force-isosurface imaging could simultaneously resolve the protrusion and the underlying molecular framework, which is structurally important for understanding catalytic sites or on-surface reaction intermediates.","The systematic underestimation of tilt angles in the saddle geometry of 2H-TPP demonstrates that neighbouring molecular features distort the local isosurface, which means quantitative structural recovery in complex molecules will require simulation-guided interpretation rather than direct geometric reading of the contour."],"fun_headline_variants":["Constant-force AFM surfaces recover 3D molecular structure that flat scans lose","Force isosurfaces beat constant-height AFM for non-planar molecule imaging","Constant-force AFM captures molecular tilt and hidden structure lost in flat scans","Force isosurfaces preserve 3D molecular features that constant-height imaging misses","Constant-force AFM surfaces reveal tilted rings and buried ligands in 3D molecules"],"cache_read_input_tokens":0,"weakest_assumption_plain":"The simulations include only Lennard–Jones (van der Waals and Pauli repulsion) interactions and omit electrostatic forces entirely. Several molecules studied — pyrrole with its polar N–H bond, ZnTPP with a central metal ion, and CO-FePc with a CO ligand carrying a significant dipole — have electrostatic contributions that could reshape the force isosurface and alter the specific contrasts and quantitative angle recovery reported.","fun_headline_variants_meta":{"raw":{"variants":["Constant-force AFM surfaces recover 3D molecular structure that flat scans lose","Force isosurfaces beat constant-height AFM for non-planar molecule imaging","Constant-force AFM captures molecular tilt and hidden structure lost in flat scans","Force isosurfaces preserve 3D molecular features that constant-height imaging misses","Constant-force AFM surfaces reveal tilted rings and buried ligands in 3D molecules","Simulations map where constant-force AFM beats constant-height for 3D molecules","Force isosurfaces retain submolecular detail across non-planar molecule geometries","Constant-force AFM contours recover molecular orientation to within one degree","Force isosurfaces expose 3D molecular structure invisible to constant-height scans","Simulations define target contrasts for 3D force-mapping of non-planar molecules"]},"model":"glm-5.2","effort":"low","cost_usd":0.0,"raw_usage":{"total_tokens":1389,"prompt_tokens":526,"completion_tokens":863,"prompt_tokens_details":null},"tokens_in":526,"tokens_out":863,"duration_ms":24341,"temperature":1.0,"reasoning_tokens":726,"cache_read_input_tokens":0,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-07-09T13:00:30.199225+00:00","model_set":{"reader":"glm-5.2"},"falsifier":"If adding electrostatic interactions to the probe-particle-model simulations substantially changes the shape of the force isosurface for pyrrole, ZnTPP, or CO-FePc — for instance by shifting the contour above polar regions enough to obscure the lower-lying features the paper claims are revealed — then the predicted contrasts and tilt-angle recovery would not hold for these molecules in real experiments.","supporting_citations":[],"review_version":1}