{"id":"bd8468e0-02ee-463a-aa60-ab8e64c651dd","arxiv_id":"2608.05305","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"Subsurface AlF3 intercalation in graphite creates a surface blister that switches adsorption from repulsive to cooperative and raises diffusion barriers, yielding a self-limiting sorption mechanism.","lead":"Quantum simulations show that an AlF3 molecule lodged just beneath a graphite surface pushes the top layer into a blister that binds additional molecules more strongly while slowing their movement. The authors use this effect to explain the two-step, self-limiting uptake of AlF3 by graphite seen in earlier experiments.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Finite-size image interactions in the 7×6×1 supercell could create the apparent cooperative adsorption; convergence checks are needed.","rationale":"The reader's weakest assumption already identifies the finite-size problem: the blister does not decay inside the 7×6×1 cell, so periodic images contribute to the computed energies without being quantified. This is precisely the most load-bearing concern because the paper's central claim is a crossover in adsorption energetics and a barrier increase, both of which are small energy differences sensitive to strain and electrostatics. A direct supercell-size convergence test is the natural and decisive check. My assessment does not change the reader's CONDITIONAL verdict: the mechanism is plausible, but the robustness of key magnitudes is unverified, so the condition on convergence remains. The experimental component assignment, while a secondary concern, is less load-bearing because the computational mechanism would stand or fall on its own energetic and kinetic predictions even if the AES decomposition were questioned.","tokens_in":15267,"tokens_out":1770,"duration_ms":18114,"concrete_test":"Repeat the DFT-D3 adsorption calculations for n=1 and n=4 on blistered graphite in at least two larger supercells, e.g., 9×8×1 and 11×9×1, keeping the same intercalant and adsorption setup and same k-point density. Check that Eads per molecule changes by less than ~0.1 eV and that the cooperative signature (Eads(n=4) < Eads(n=1) on the blister, opposite to pristine) persists. Additionally, recompute the P3 NEB barrier in the larger cell; if it shifts by more than ~50 meV, the kinetic-trap conclusion requires revision.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central computational observation, that subsurface AlF3 turns graphite adsorption from repulsive to cooperative (Section 3.3) and raises diffusion barriers (Section 3.4), rests on energy differences of 0.2–0.5 eV that are never checked for finite-size convergence. In the 7×6×1 supercell (17.29×12.83 Å), the interlayer spacing at the cell edges is reported as ~3.56 Å versus 3.35 Å pristine, meaning the blister deformation does not decay within the cell. Consequently, the intercalant and the surface molecules interact with periodic images of both the strain field and the intercalant's charge reservoir (0.552 e, Section 3.5). These unquantified image contributions could artificially stabilize or destabilize adsorption and barriers, potentially manufacturing the crossover from weakening (−1.035 to −0.830 eV on pristine) to strengthening (−0.853 to −1.498 eV on blister) that is the paper's headline. The NEB P3 barrier of 251 meV, the main source of kinetic trapping, is especially sensitive to the periodic elastic field. Without a larger-cell test, the mechanism's energetic and kinetic claims are not robust to the model's most obvious approximation.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports a DFT-D3 study of AlF3 adsorption and diffusion on pristine graphite and on a graphite surface containing a subsurface intercalated AlF3 molecule in a 7×6×1 supercell. The central claim is that the intercalant creates a blister-like deformation that turns the surface into a coverage-activated cooperative adsorption center and a kinetic trap: on the blistered surface, the adsorption energy per molecule strengthens from -0.853 eV at n=1 to -1.498 eV at n=4, whereas on pristine graphite it weakens from -1.035 to -0.830 eV; simultaneously, the P3 diffusion barrier rises from 103.9 to 251.0 meV and the computed 2D diffusion coefficient drops from 6.4×10^-9 to 2.8×10^-9 m^2/s at 300 K. The authors connect these results to a previously reported biexponential AES sorption kinetics and propose a self-limiting intercalation mechanism.","tokens_in":15387,"tokens_out":6623,"duration_ms":57424,"significance":"If the reported crossover is robust, the paper offers a novel and potentially important structure-property relationship for intercalation-induced surface modification in graphite, with direct relevance to AlF3-coated carbon electrodes. The computational work is internally consistent: reproducing Eqs. (2)-(3) from Table 3 gives the stated D values, and the authors compare two clearly defined structural models without fitting parameters to the target kinetics. The paper also tests two explicit hypotheses (molecular clustering vs. intercalation-induced conditioning), which is a methodological strength. However, the central quantitative claim rests on a single supercell size, and the reported elastic deformation does not decay within that cell. This finite-size issue is load-bearing because the cooperative binding and barrier increases are energy differences of 0.2-0.5 eV; without convergence checks, the paper's headline conclusions remain conditional.","major_comments":[{"comment":"The finite-size convergence of the supercell is not addressed. The interlayer spacing at the supercell edges is reported as ~3.56 Å versus 3.35 Å pristine, indicating that the blister deformation does not fully decay within the 7×6×1 cell. The cooperative adsorption strengthening (n=1: -0.853 eV; n=4: -1.498 eV, Table 2 and Figure 2) and the NEB P3 barrier increase (103.9 to 251.0 meV, Table 3) rely on energy differences of 0.2-0.5 eV. Because the intercalant and the adsorbed molecules interact with periodic images of both the strain field and the intercalant's charge reservoir (0.552 e, Table 4), the claimed crossover from repulsive to cooperative adsorption could be an artifact of the image interactions. I request convergence tests with at least one larger lateral supercell (e.g., 10×10×1 or 12×10×1) for the n=1 and n=4 adsorption energies and for the P3 barrier.","section":"Section 3.3 and Figure 3"},{"comment":"The effective diffusion coefficient in Eq. (3) sums over the three pathways P1-P3 with one term per pathway, without accounting for the multiplicity of equivalent jump directions on the hexagonal graphite lattice. On a surface with ABAB stacking, each type of jump can occur along several symmetry-equivalent directions; omitting these multiplicities changes D by an integer factor. The claim that Dblister/Dpristine = 2.3 at 300 K depends on this choice. The authors should state the multiplicities used for each pathway and verify that the same multiplicities apply to both the pristine and blistered surfaces; otherwise the reported mobility reduction may be quantitatively incorrect.","section":"Equation (3) and Table 3"}],"minor_comments":[{"comment":"The phrase 'spontaneous formation and adsorption of AlF3 dimers' is contradicted later in the same section by the statement that dimers did not form spontaneously and required manual construction. Please rephrase to describe the investigation of preformed dimer configurations.","section":"Section 3.2"},{"comment":"The text claims that 'the charge on the intercalated Al atom remains essentially constant at ~0.96 e across all surface coverages,' but Table 4 lists only the surface Al charges (qAl) for the blistered-surface rows; the intercalant charge is not reported for n=1-4. Either include these values in Table 4 or qualify the claim as coming from separate analysis.","section":"Section 3.5 and Table 4"},{"comment":"The α and β hollow sites and the precursor P0 are introduced without a structural illustration; adding them to Figure 4's inset would improve clarity, especially since the NEB pathways are described relative to these sites.","section":"Section 3.4"},{"comment":"There is a typo in 'Volmer–Weber' rendered as 'V olmer–Weber'; also, the experimental data in Figure 1 are reproduced from Ref. [17] and this should be stated explicitly in the figure caption, not only in Section 2.1.","section":"Section 3.1"}],"recommendation":"major_revision","confidential_remarks":"This is a conditional accept in spirit: the central mechanism is plausible and the calculations are clean, but the finite-size issue is too central to ignore. If the authors can provide a faster-converged check (e.g., a 10×10×1 cell for the key adsorption energies and the P3 barrier), the paper should be publishable. The diffusion-coefficient multiplicity issue is also fixable and could be handled as a minor revision if the ratio is robust. The paper's reliance on the prior experimental assignment of the AES components to intercalation is worth an explicit caveat in the conclusions."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nThe paper has a real mechanism worth discussing: a subsurface intercalated AlF3 turns the graphite surface into a coverage-activated cooperative adsorption center and a kinetic trap. That specific claim—the repulsive-to-cooperative crossover and the raised NEB barriers on the blistered surface—is new relative to the group's earlier blister papers. The arithmetic is internally consistent; I recomputed D from Table 3 and got their values. The electronic-structure picture (charge reservoir, extended perturbation) is coherent.\n\nThe soft spots are real but not fatal. The 7×6×1 supercell is the big one. The interlayer spacing at the cell edge is still 3.56 Å versus 3.35 pristine, so the blister deformation does not decay. That means the 0.2–0.5 eV energy differences driving the crossover include unquantified image interactions from both the strain field and the intercalant's 0.552 e charge. Without a larger-cell test, the headline crossover could be an artifact. The paper reports no convergence checks and no code/data, and the SI is unreleased. That needs fixing.\n\nThe experimental connection is oversold. A 2.3-fold D reduction at 300 K, with nanosecond hop times, does not by itself explain a biexponential AES signal with components separated by orders of magnitude. The mapping from computed diffusion to the fast/slow exponential assignment is loose, and that assignment is inherited from their Ref. [17], which I cannot access. The authors do report the frozen-substrate ratio of 52, which at least flags the sensitivity.\n\nWho is this for? Computational surface scientists working on intercalation and adsorption in layered materials. They will get a plausible mechanism and a testable prediction (preferential accumulation around blisters). It deserves a serious referee, but the referee should push for convergence tests and a tempered discussion of the experimental link.","headline":"A plausible coverage-dependent adsorption mechanism with a real finite-size problem; worth refereeing, but the experimental link is oversold.","tokens_in":16048,"tokens_out":2652,"would_cite":false,"duration_ms":23679,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"Subsurface AlF3 turns graphite surface into a kinetic trap","keywords":["graphite intercalation","AlF3 adsorption","surface blister deformation","cooperative adsorption","diffusion barrier","kinetic trapping","density functional theory","Auger electron spectroscopy"],"falsifier":"Low-temperature STM tracking of individual AlF3 molecules on a graphite surface with a known subsurface intercalant would settle the claim: the model predicts hop times above the blister of ~2.7–5.3 ps for short jumps and ~16.5 ns for the hexagon-crossing jump, versus 1.6–1.8 ps and 55.8 ps on pristine terraces, and a desorption-energy crossover from weakening to strengthening with coverage; observing no such barrier increase or no coverage strengthening would refute it.","tokens_in":14955,"feed_emoji":"🧲","tokens_out":9150,"duration_ms":64542,"temperature":0.7,"pith_summary":"The paper sets out to explain the microscopic origin of the two-step, self-limiting uptake of AlF3 on graphite observed in Auger experiments. It claims that the two steps are not independent—surface adsorption and subsurface intercalation are coupled through a feedback loop: once a single AlF3 molecule has intercalated beneath the top graphene layer, it pushes that layer up into a local blister, and the blister converts the surface into a cooperative adsorption center and a kinetic trap. The key evidence is a reversal of coverage dependence: on pristine graphite, adsorption per molecule weakens from −1.035 to −0.830 eV as coverage rises to four molecules, whereas above the blister it strengthens from −0.853 to −1.498 eV, while the diffusion coefficient drops from 6.4×10⁻⁹ to 2.8×10⁻⁹ m²/s at 300 K. The paper concludes that the blister-induced trapping makes further intercalation self-limiting and identifies the intercalant as a stable electronic reservoir that deepens the surface potential landscape. If the picture is right, it turns substrate crystallinity, deposition flux, and defect density into tunable parameters for controlling the adsorption–intercalation balance in carbon-based electrodes.","feed_headline":"Subsurface AlF3 turns graphite surface into a kinetic trap","feed_subtitle":"A subsurface blister flips AlF3 adsorption from repulsive to cooperative and slows diffusion, explaining two-step uptake.","key_machinery":"The load-bearing object is the intercalation-induced blister: a localized out-of-plane deformation of the top graphene layer, purely elastic, expanding the interlayer spacing from 3.35 Å to 5.07 Å above the molecule and decaying to ~3.56 Å at the cell edges. The paper shows that this curvature, together with intercalation-induced charge redistribution, drives three coupled effects: spontaneous dimerization of surface monomers, cooperative (coverage-strengthening) adsorption energies, and raised diffusion barriers. The electronic mechanism is identified through charge-density differences and Mulliken populations: the intercalated AlF3 maintains a nearly constant charge (~0.96 e) while acting as a reservoir that deepens the surface potential wells, so that above the blister the charge-accumulation regions of four surface molecules merge into one continuous lobe spanning the blister.","core_discovery":"The central discovery is that a subsurface intercalated AlF3 molecule does not merely expand the graphite lattice; it creates a specific surface condition—a blister—that reverses the sign of the coverage dependence of adsorption and suppresses lateral diffusion. At the DFT-D3 level, with one intercalated AlF3 molecule in a 7×6×1 Bernal graphite supercell, the interlayer spacing expands from 3.35 Å to 5.07 Å above the intercalant and relaxes to ~3.56 Å at the cell edges, with no C–C bond breaking. Adsorption energies per AlF3 molecule on this blistered surface strengthen with coverage (−0.853, −1.171, −1.281, −1.498 eV for n=1–4), while pristine graphite shows monotonic weakening (−1.035 to −0.830 eV); the blistered surface also spontaneously dimerizes monomers that do not dimerize on pristine graphite. Nudged elastic band calculations show diffusion barriers rise from 12.4–103.9 meV on pristine graphite to 25.8–251.0 meV on the blister, reducing the effective 2D diffusion coefficient from 6.4×10⁻⁹ to 2.8×10⁻⁹ m²/s at 300 K, with the ratio growing to ~8 at 100 K. Charge-density difference and Mulliken analysis show the intercalant transfers 0.552 e to graphite, perturbs 61 carbons, and keeps charge transfer per surface molecule 1.7–2.1× higher than pristine at all coverages, with surface molecules' accumulation regions merging into a single lobe at n=4. The paper interprets this as the intercalant acting as a stable electronic reservoir that deepens the surface potential landscape, and proposes that this blister-induced trapping is the mechanism behind the experimentally observed crossover from a fast, defect-gated intercalation component to a slow, substrate-independent adsorption component in the biexponential AES kinetics.","pith_inferences":["The mechanism is likely not specific to AlF3: any subsurface guest that produces a comparable elastic blister in the top graphene layer should generate similar cooperative adsorption and kinetic trapping, which could be tested computationally for ions such as AlCl4−.","Because the blister deformation does not fully decay in the 7×6×1 cell, the 0.2–0.5 eV cooperative-binding differences probably include image contributions; a supercell-size convergence test would either confirm or revise the crossover magnitude.","A direct experimental falsification would be measuring the hop time of individual molecules above a known subsurface blister by low-temperature STM and comparing with the predicted ~16.5 ns hexagon-crossing time.","The self-limiting picture suggests a design rule: pre-intercalating graphite with AlF3 could intentionally cap further molecular uptake, which could be tested by comparing AlF3 deposition on fresh versus pre-intercalated substrates."],"forward_implications":["The two exponential components in the measured Auger attenuation curves are explained as a fast, defect-gated intercalation channel followed by a slow adsorption channel, with blister trapping preventing further intercalation.","Molecular clustering on pristine graphite is ruled out as an independent thermodynamic pathway; dimer formation becomes spontaneous only after a blister exists.","The mobility gap between pristine and blistered graphite grows at low temperature, from a factor of 2.3 at 300 K to about 8 at 100 K, so low-temperature experiments maximize the contrast.","Preferential AlF3 accumulation around intercalation-induced blisters is predicted, providing a spatial signature to look for in local-probe microscopy.","Deposition flux and substrate defect density become tunable parameters for controlling the balance between surface adsorption and interlayer intercalation in AlF3-modified carbon electrodes."],"supporting_citations":[{"why":"Supplies the experimental biexponential AES kinetics and the flux/defect dependence that the mechanism is built to explain.","marker":"[17]"},{"why":"Prior DFT study establishing AlF3 intercalation stability, charge transfer to graphite, and the top-site adsorption geometry used as the reference.","marker":"[18]"},{"why":"Molecular dynamics simulations showing intercalated AlF3 forms lateral clusters and blister-like elastic deformations, motivating the blister model.","marker":"[19]"},{"why":"STM, XPS, and RBS experiments confirming the presence of surface deformations upon AlF3 intercalation.","marker":"[20]"},{"why":"Supplies the DFT-D3 van der Waals correction used for interlayer forces and molecule–surface binding.","marker":"[25]"},{"why":"Provides the climbing-image nudged elastic band method used to compute the diffusion barriers on pristine and blistered surfaces.","marker":"[28]"},{"why":"Supplies the attempt frequency prefactor (10^12 Hz) used in the Eyring–Kramers jump-rate expression.","marker":"[33]"},{"why":"Provides the dipole–dipole repulsion model explaining the coverage-dependent weakening on pristine graphite that the blister reverses.","marker":"[35]"},{"why":"Gives the random-walk relation between hopping rates and the effective 2D diffusion coefficient.","marker":"[36]"}],"fun_headline_variants":["Subsurface AlF3 blister flips adsorption to cooperative and traps diffusers","Graphite blister from intercalated AlF3 traps surface motion","Intercalated AlF3 creates cooperative adsorption and kinetic trap on graphite","AlF3 intercalation turns graphite blister into a kinetic trap for adsorbates","Blistering graphite: AlF3 intercalation drives cooperative adsorption and trapping"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that a single intercalated AlF3 molecule in a periodic 7×6×1 supercell adequately represents an intercalation-conditioned graphite surface, even though the blister deformation does not fully decay before the cell edge (spacings reach ~3.56 Å versus 3.35 Å pristine), leaving image interactions present but unquantified.","fun_headline_variants_meta":{"raw":{"variants":["Subsurface AlF3 blister flips adsorption to cooperative and traps diffusers","Graphite blister from intercalated AlF3 traps surface motion","Intercalated AlF3 creates cooperative adsorption and kinetic trap on graphite","AlF3 intercalation turns graphite blister into a kinetic trap for adsorbates","Blistering graphite: AlF3 intercalation drives cooperative adsorption and trapping"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000958,"raw_usage":{"total_tokens":4210,"prompt_tokens":1204,"completion_tokens":3006,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":820,"completion_tokens_details":{"reasoning_tokens":2906}},"tokens_in":820,"tokens_out":3006,"duration_ms":17124,"temperature":1.0,"reasoning_tokens":2906,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-08T15:40:46.387470+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Low-temperature STM tracking of individual AlF3 molecules on a graphite surface with a known subsurface intercalant would settle the claim: the model predicts hop times above the blister of ~2.7–5.3 ps for short jumps and ~16.5 ns for the hexagon-crossing jump, versus 1.6–1.8 ps and 55.8 ps on pristine terraces, and a desorption-energy crossover from weakening to strengthening with coverage; observing no such barrier increase or no coverage strengthening would refute it.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the experimental biexponential AES kinetics and the flux/defect dependence that the mechanism is built to explain."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Prior DFT study establishing AlF3 intercalation stability, charge transfer to graphite, and the top-site adsorption geometry used as the reference."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Molecular dynamics simulations showing intercalated AlF3 forms lateral clusters and blister-like elastic deformations, motivating the blister model."},{"cited_title":"Rodríguez, I","cited_arxiv_id":null,"evidence_quote":"STM, XPS, and RBS experiments confirming the presence of surface deformations upon AlF3 intercalation."},{"cited_title":"Grimme, J","cited_arxiv_id":null,"evidence_quote":"Supplies the DFT-D3 van der Waals correction used for interlayer forces and molecule–surface binding."},{"cited_title":"Henkelman, B","cited_arxiv_id":null,"evidence_quote":"Provides the climbing-image nudged elastic band method used to compute the diffusion barriers on pristine and blistered surfaces."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the attempt frequency prefactor (10^12 Hz) used in the Eyring–Kramers jump-rate expression."},{"cited_title":"Deshlahra, J","cited_arxiv_id":null,"evidence_quote":"Provides the dipole–dipole repulsion model explaining the coverage-dependent weakening on pristine graphite that the blister reverses."},{"cited_title":"Ala-Nissila, R","cited_arxiv_id":null,"evidence_quote":"Gives the random-walk relation between hopping rates and the effective 2D diffusion coefficient."}],"review_version":1}