{"id":"45fcd9da-769f-4bc0-a605-098a6ba7afaf","arxiv_id":"2502.02429","paper_version":3,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"Extended Lagrangian quantum molecular dynamics with an electrode bias shows solvent-mediated outer-sphere O2 reduction at high bias and adsorbed inner-sphere reduction at low bias in a proof-of-concept simulation.","lead":"This paper simulates oxygen reduction on nitrogen-doped graphene in water using an accelerated quantum molecular dynamics method and a tunable electronic bias on the electrode. The simulations show two different reaction pathways at different bias settings, but each pathway is currently supported by single short trajectories.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Low-bias inner-sphere run is confounded: it used a modified geometry with O2 pre-positioned near NG plus a steering force, so the mechanism switch may reflect initial geometry/steering, not applied bias.","rationale":"The reader's weakest_assumption correctly identifies the geometry/steering confound in the low-bias inner-sphere simulation. I agree that this is the most load-bearing concern: the abstract makes a causal claim about bias controlling the mechanism, but the only inner-sphere observation is generated under a protocol that simultaneously changes the starting O2 position and applies a steering force. The high-bias and zero-bias simulations use the original geometry, so the comparison conflates bias with geometry. This is not an internal inconsistency in the implementation—the individual trajectories appear to show the reported events—but it is a failure of the comparison design for the stated conclusion. The static onsite-shift bias (Eq. 6) is an additional physical approximation, but the paper is transparent about its limitations, and even a more rigorous constant-potential method would leave the geometry confound unresolved. The proposed test directly isolates the bias variable by rerunning the low-bias simulation in the original geometry without steering. If the inner-sphere pathway persists, the claim gains support; if not, the central claim must be weakened to a proof-of-concept that different pathways can be observed under different combined conditions. This does not change the reader's CONDITIONAL verdict: the paper remains a promising methodological demonstration whose headline mechanistic interpretation requires additional controlled simulations. The absence of deposited inputs and the lack of DFTB/ORR benchmarking reinforce the conditional status but are not the primary load-bearing concern.","tokens_in":18033,"tokens_out":2905,"duration_ms":28715,"concrete_test":"Re-run the low-bias protocol (µe shift -0.9 eV on NG, same thermostat, same β, same 20 ps window) with the original unmodified starting geometry and no steering potential, using at least two independent velocity seeds. If O2 adsorbs and follows the inner-sphere pathway, the confound is resolved; if it reduces via solvent or remains intact, the mechanism-switch claim is not supported by bias alone.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central mechanistic claim—that applied bias alone flips ORR from outer-sphere to inner-sphere—rests on a single comparison with a confounded protocol. The outer-sphere and zero-bias results come from the original system (O2 initially distant from NG; Figs. 3–5, S1). The inner-sphere observation comes from a 'modified version of the main test system' with 'the O2 molecule located closer to the NG surface' and a steering method 'to induce diffusion of the oxygen molecule towards the NG sheet' (Simulations With Applied Bias, paragraph 3). Thus the low-bias run differs from the high-bias run in at least two ways: applied bias (-0.9 eV vs +2 eV) and initial/steered geometry. The abstract and conclusion attribute the mechanism difference solely to bias ('Which mechanism we see depends on the bias applied'), but the data cannot separate bias from geometry/steering. The bias method itself (Eq. 6) is a static onsite shift without a counter electrode or electron reservoir—acknowledged by the authors—so the physical meaning of the bias is also approximate; however, that limitation is secondary because even a perfect bias implementation would not remove the geometry confound. This is the load-bearing weakness: if the inner-sphere pathway is induced by the closer O2 placement or the steering force, the central claim of bias-controlled mechanism selection collapses to 'two different simulations showed two different mechanisms.'","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper presents an XL-BOMD-based DFTB framework for simulating the oxygen reduction reaction (ORR) at a nitrogen-doped graphene (NG)/water interface, including a proposed electrochemical biasing scheme in which the onsite energies of the electrode atoms are shifted (Eq. 6). Simulations without applied bias show O2 cleavage and a two-step transfer of four electrons from NG to the solution, interpreted as an outer-sphere mechanism with solvent-mediated electron transfer. A +2 eV shift is reported to accelerate this outer-sphere reaction, while a -0.9 eV shift, applied to a modified system with O2 initially closer to the NG surface and with a steering force, yields an inner-sphere pathway with O2 adsorption and partial reduction. The authors conclude that the observed ORR mechanism depends on the applied bias and that this is the first application of XL-BOMD to heterogeneous electrocatalysis.","tokens_in":18326,"tokens_out":7757,"duration_ms":72567,"significance":"If the bias-dependence claim held, the work would be a valuable proof of concept for atomistic electrocatalysis: it combines explicit solvation, an explicit electrode, picosecond-scale trajectories, and a simple bias protocol that can distinguish outer- and inner-sphere ORR pathways. The paper has real strengths: the trajectory data are internally consistent with O2 cleavage, two electron-injection events, and HOMO-LUMO gap closure at the reported times; control systems and extended-system simulations provide useful supporting evidence; and the preconditioning convergence analysis is a legitimate methodological contribution. However, the central mechanism-switch claim is currently supported by a single confounded comparison, and the quantitative overpotential statements rest on an approximate bias model whose calibration is not validated. The limitations are partly acknowledged in the text, but the abstract and conclusion state the bias-dependence claim more strongly than the evidence supports.","major_comments":[{"comment":"The central mechanistic claim—that applied bias alone determines whether ORR is outer-sphere or inner-sphere—is not supported by the reported comparison. The high-bias and zero-bias runs start from the original geometry with O2 initially far from the NG sheet, whereas the low-bias (-0.9 eV) run uses a 'modified version of the main test system' with O2 'located closer to the NG surface' and includes a steering method 'to induce diffusion of the oxygen molecule towards the NG sheet.' The two conditions therefore differ in initial geometry and steering in addition to bias. Because the abstract and conclusion attribute the mechanism difference solely to bias, a control experiment is required: either the -0.9 eV shift on the original geometry, or the modified/steered geometry at +2 eV, or both. Without such controls, the data are equally consistent with the interpretation that the closer initial placement or the steering force, rather than the bias, produced the adsorbed pathway. The reported -0.9 eV threshold and the derived 1.69 V overpotential are also tied to this modified/steered setup, so they are not transferable to the original system.","section":"Simulations With Applied Bias, paragraph 3"},{"comment":"The mechanism-switch claim rests on a single trajectory per condition. The paper itself notes that 'these simulations do depend on the initial randomized velocities established by the 300 K Langevin thermostat,' and with only one trajectory at +2 eV, one at 0 eV, and one at -0.9 eV, the difference between the observed pathways could reflect stochasticity rather than a bias-controlled transition. Repeated independent runs with different thermostat seeds and initial velocities should be reported, with the number of trajectories showing each mechanism. This is particularly important because the inner-sphere trajectory is the only run in which adsorption occurs. The text also reports that further downshifts did not produce ORR within 20 ps, so the claimed low-bias inner-sphere regime is a narrow window; this should be made explicit.","section":"Simulations With Applied Bias and Results and Discussion"},{"comment":"The static onsite-energy shift is not equivalent to an electrochemical overpotential in the grand-canonical sense; the simulation has no counter electrode, no electron reservoir, and no constant-potential constraint, as the authors partly acknowledge in the discussion of the outer-sphere simulations. The quantitative statements that a -0.9 eV shift 'corresponds to' an overpotential of 1.69 V assume a rigid-band mapping from V_i to the electrode potential that is not validated and does not account for the charge redistribution and solvent response during the reaction. The manuscript should either provide a validation of this mapping (for example, by computing the potential of the biased electrode relative to a reference) or present the bias values only as a model parameter and weaken the overpotential language.","section":"Eq. (6) and 'Applying an Electrochemical Potential Bias'"}],"minor_comments":[{"comment":"Table 1 and the surrounding text describe HOO− twice with different mean charges (-0.65 and -0.46) and different oxidation states; clarify which oxygen is underlined/bolded and define the two entries, since the subsequent species assignments (e.g., in Figure 11) rely on these reference values.","section":"Mulliken Charge Analysis, Table 1"},{"comment":"The terms 'high bias' and 'low bias' are defined only by the sign of the onsite shift, yet the -0.9 eV value is a smaller shift than the -2 eV value that produced no ORR; define the convention explicitly to avoid confusion.","section":"Simulations With Applied Bias"},{"comment":"Equation (6) uses the same symbol H_iα,jβ as Eq. (1), which is the unshifted Hamiltonian; use a different symbol (e.g., H^b or a tilde) to avoid confusion.","section":"Applying an Electrochemical Potential Bias"},{"comment":"The steering protocol is only cited to a ChemRxiv preprint; the force constant, direction, and duration of the steering should be reported in the main text or supporting information so the inner-sphere run is reproducible.","section":"Simulations With Applied Bias"},{"comment":"The paper states that the outer-sphere mechanism proceeds without adsorption of O2 to NG, but no O2-NG distance or adsorption order parameter is provided; adding such a quantity would make the mechanistic assignment more quantitative.","section":"Results and Discussion, outer-sphere mechanism"}],"recommendation":"major_revision","confidential_remarks":"The stress-test concern about circularity does not really land: the observed mechanism is a trajectory outcome, not a fitted derivation of the result from its own parameters. The decisive issue is the confound between applied bias and the modified/steered geometry in the inner-sphere run. I would not reject the manuscript on the basis of the approximate bias model alone, but the central mechanistic claim needs either the missing control simulations or a substantial softening of the abstract and conclusion."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nHere's my read on arXiv:2502.02429. The headline is that XL-BOMD, a fast approximate quantum MD method, is applied to oxygen reduction on nitrogen-doped graphene in explicit water, with an onsite-energy shift used as an electrode bias. That is genuinely new as an application, and the paper shows real observations: O2 bond cleavage around 6 ps, two-step electron transfer, solvent-mediated outer-sphere reduction, and a separate inner-sphere trajectory at lower bias. The authors are also upfront about the method's limitations—no steady state, no electron reservoir, no constant bias, sensitivity to initial velocities.\n\nThe soft spot is the central mechanism-switch claim. The outer-sphere and zero-bias runs use O2 initially far from the sheet; the inner-sphere run uses a modified system with O2 placed closer and a steering force pushing it toward the surface. So the difference between 'outer sphere under high bias' and 'inner sphere under low bias' is confounded by geometry and steering. The abstract says 'which mechanism we see depends on the bias applied,' but the data cannot separate bias from initial configuration. This is a load-bearing flaw, but it's fixable: run the high-bias case with the same modified geometry, or run low bias with the original geometry, and repeat with different velocities.\n\nThe bias method itself is also approximate—a static onsite shift without a counter electrode or electron reservoir—but the authors acknowledge that, and it's secondary to the geometry confound. There's no shipped code or inputs, which limits reproducibility, and DFTB/Mulliken charges are not benchmarked against higher-level methods for ORR, so the quantitative overpotentials (1.69 V etc.) should be taken as rough.\n\nWho's it for? People working on computational electrochemistry, especially those interested in explicit-solvent dynamics and outer-sphere charge transfer. It's a promising proof-of-concept, not a settled result. I'd send it to peer review because the framework is potentially important and the core issue is addressable with controlled simulations. The referees should ask for the confound to be dealt with, but the paper deserves their time.","headline":"A promising proof-of-concept for XL-BOMD in electrocatalysis, but the central bias-controlled mechanism switch is confounded by a different starting geometry and steering in the low-bias run.","tokens_in":18880,"tokens_out":2503,"would_cite":false,"duration_ms":23147,"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":"A single biasing scheme flips oxygen reduction between two mechanisms on nitrogen-doped graphene, resolved atom-by-atom in explicit water.","keywords":["extended Lagrangian Born-Oppenheimer molecular dynamics","XL-BOMD","oxygen reduction reaction","nitrogen-doped graphene","electrochemical bias","outer sphere mechanism","inner sphere mechanism","density functional tight binding"],"falsifier":"Run two simulations that swap the two protocols: take the original outer-sphere starting geometry (O2 far from NG) and apply the -0.9 eV shift, and take the inner-sphere modified geometry (O2 near NG, with the steering method) with no bias or a +2.0 eV shift; if the far geometry still reduces via solvent-mediated outer-sphere charge transfer and the near geometry still adsorbs O2 and dissociates it, then the mechanism is controlled by geometry rather than by the applied bias.","tokens_in":17826,"feed_emoji":"⚡","tokens_out":2057,"duration_ms":22010,"temperature":0.7,"pith_summary":"This paper claims that extended Lagrangian Born-Oppenheimer molecular dynamics (XL-BOMD), a fast approximation to quantum molecular dynamics, can simulate electrocatalytic reactions at a solid-liquid interface with explicit solvent, including the effect of an applied electrochemical bias. Using nitrogen-doped graphene (NG) as the electrode and aqueous O2 as the reactant, the authors report that the oxygen reduction reaction (ORR) proceeds by an outer-sphere mechanism under high applied bias, where electrons transfer through the solvent without O2 adsorbing to the surface, and by an inner-sphere mechanism under low bias, where O2 adsorbs to NG before direct electron transfer. If this result holds, XL-BOMD offers a practical route to time-resolved, atomistic insight into potential-dependent electrocatalysis that static DFT with the computational hydrogen electrode cannot provide, because it explicitly includes both electrode and electrolyte and observes charge-transfer events dynamically. A sympathetic reader would care because this is the first reported application of XL-BOMD to heterogeneous electrocatalysis, and the observed mechanism switch is directly tied to the magnitude of the applied bias.","feed_headline":"Bias flips oxygen reduction pathway on doped graphene","feed_subtitle":"Solvent-mediated vs adsorbed electron transfer in explicit atomistic simulations of the ORR.","key_machinery":"The load-bearing mechanism is extended Lagrangian Born-Oppenheimer molecular dynamics (XL-BOMD) within the self-consistent charge density functional tight-binding (SCC-DFTB) framework, where the net Mulliken charges on atoms are treated as auxiliary dynamical variables that oscillate around the exact ground-state density through a harmonic oscillator, avoiding repeated self-consistent field iterations. The key object is the shadow potential energy surface $U_{BO}(R,n)$, whose error scales quadratically with the residual $\\|q[n]-n\\|$, and the associated kernel $K=J^{-1}$, approximated by a preconditioned rank-$m$ update, which keeps the dynamical charges close to the Born-Oppenheimer ground state during reactive events. The bias is applied by shifting the onsite orbital energies of the electrode atoms through Eq. (6), $H_{i\\alpha,j\\beta} = H_{i\\alpha,j\\beta} + \\frac{1}{2}(S_{i\\alpha,j\\beta}V_j + V_i S_{i\\alpha,j\\beta})$, which displaces the electron chemical potential of the NG sheet relative to the solution, effectively controlling the driving force of the reduction. Reaction events are identified by HOMO-LUMO gap closures near zero and by spikes in the XL-BOMD residual norm, which coincide with the two electron-injection waves.","core_discovery":"The central claim is that a controllable potential bias applied to an NG electrode, implemented by shifting the onsite orbital energies of the electrode atoms according to Eq. (6), changes the observed ORR mechanism: high bias (no applied shift, effectively a high reductive overpotential of about 2.59 V) drives a solvent-mediated outer-sphere pathway in which O2 is reduced to 4 OH− through two successive two-electron injections without surface adsorption, while a negative 0.9 eV shift of the electrode's electron chemical potential (lowering the overpotential to about 1.69 V) enables an inner-sphere pathway in which O2 adsorbs to NG, forms a transient peroxo species, and dissociates into one adsorbed oxygen and two OH−. The paper also claims that nitrogen doping raises the electron chemical potential and density of states near the Fermi level relative to pure graphene, explaining NG's higher activity, and that the electron holes left on NG after oxidation localize on carbons directly bonded to nitrogen. The authors position this as the first demonstration of XL-BOMD applied to heterogeneous electrocatalysis, showing stable picosecond-scale charge transfer across an explicit NG–water interface with fully converged electronic structure at each step.","pith_inferences":["The claim that the mechanism depends on applied bias implicitly assumes that the applied bias, not the modified starting geometry, selects the pathway; a direct test would be to run the inner-sphere geometry without bias and the outer-sphere geometry with a -0.9 eV shift, and check whether the mechanism follows the bias or the geometry.","The paper's static onsite-energy shift is not a constant-potential, grand-canonical scheme because the simulation has no electron reservoir or counter electrode, so the observed 'high bias' vs 'low bias' labels should be read as relative driving-force shifts rather than absolute experimental potentials; a constant-potential extension would be a natural test of whether the mechanism-switch threshol","Because the authors note that the final step (formation of additional OH− from the adsorbed oxygen) is not seen in the simulated time span, the inner-sphere pathway may be kinetically incomplete; longer trajectories or enhanced sampling of the adsorbed-oxygen intermediate could determine whether the inner-sphere mechanism also ultimately produces 4 OH−.","If the solvent-mediated outer-sphere route is real, its rate should depend strongly on water structure and hydrogen-bonding fluctuations; a testable extension is to vary the solvent (e.g., adding more water layers, changing temperature, or using a different polar solvent) and see whether the electron-transfer timing and the HOMO-LUMO gap closures shift accordingly."],"forward_implications":["If the central claim is correct, XL-BOMD with the onsite-energy bias method becomes a standard tool for explicitly simulating electrocatalytic reactions at solid-liquid interfaces with full solvent and electrode representation, avoiding the prohibitive cost of repeated SCC iterations in traditional QMD.","The observed mechanism switch implies that the applied bias controls whether ORR proceeds through a solvent-mediated outer-sphere pathway or an adsorption-dependent inner-sphere pathway on NG, meaning that simulations without a properly tuned bias may miss the inner-sphere regime entirely.","The paper's finding that nitrogen doping raises the electron chemical potential and DOS near the Fermi level relative to pure graphene provides a direct electronic-structure rationale for NG's superior ORR activity over pure graphene, consistent with the observed absence of ORR in the pure-graphene control.","The successful observation of a water-mediated outer-sphere electron transfer with explicit solvent dynamics suggests that XL-BOMD can resolve non-adsorptive charge-transfer steps that static DFT with the computational hydrogen electrode cannot describe by construction.","If the bias-scan approach generalizes, experimentally relevant quantities such as the overpotential at which inner-sphere adsorption begins can be extracted directly from simulation by varying the onsite energy shift, providing a parameter-free route to mechanistic phase diagrams of electrocatalysts."],"supporting_citations":[{"why":"Supplies the extended Lagrangian formulation that treats the electronic degrees of freedom as auxiliary dynamical variables, enabling efficient and stable propagation without repeated SCF iterations.","marker":"[19]"},{"why":"Provides the general framework of XL-BOMD from DFT to charge relaxation models, establishing the residual-error-based accuracy and stability criteria used here.","marker":"[20]"},{"why":"Defines the SCC-DFTB Hamiltonian and charge fluctuation formalism that underlies the DFTB-based electronic structure used in all simulations.","marker":"[26]"},{"why":"Identifies DFTB+ as an implementation of SCC-DFTB, giving the baseline speed and accuracy comparison for the method.","marker":"[27]"},{"why":"Supplies the LATTE Hamiltonian package with a transferable parametrization that the authors use for the NG-water-O2 simulations.","marker":"[34]"},{"why":"Describes the graph-based quantum response theory and shadow Born-Oppenheimer MD approach that forms the basis of the GPMDK code used for the simulations.","marker":"[35]"},{"why":"Provides the steering method used to induce diffusion of O2 toward the NG surface in the inner-sphere simulations, establishing the protocol for the low-bias pathway.","marker":"[56]"},{"why":"Gives the computational hydrogen electrode approach that the paper contrasts with, highlighting that static DFT cannot describe non-adsorptive charge transfer.","marker":"[14]"},{"why":"Supplies the absolute electrode potential reference value (4.44 V) used to convert the electron chemical potential differences into effective overpotentials.","marker":"[55]"},{"why":"Supports the use of HOMO-LUMO gap closures as an indicator of reaction events, which the authors use to identify the two electron-injection steps.","marker":"[52]"}],"fun_headline_variants":["Bias flips oxygen reduction pathway on nitrogen-doped graphene","Solvent mediates oxygen reduction at high bias on doped graphene","Electrochemical bias switches ORR mechanism on N-doped graphene","Bias tunes oxygen reduction: solvent vs adsorbed pathway on NG","Nitrogen-doped graphene bias flips ORR from outer to inner sphere"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The central mechanistic comparison assumes that the applied onsite-energy shift, and not the altered initial geometry with O2 placed closer to the surface and steered toward it, is what determines which reduction pathway appears, so if the starting O2-surface separation or the steering force alone caused the inner-sphere adsorption, the mechanism-switch claim would collapse.","fun_headline_variants_meta":{"raw":{"variants":["Bias flips oxygen reduction pathway on nitrogen-doped graphene","Solvent mediates oxygen reduction at high bias on doped graphene","Electrochemical bias switches ORR mechanism on N-doped graphene","Bias tunes oxygen reduction: solvent vs adsorbed pathway on NG","Nitrogen-doped graphene bias flips ORR from outer to inner sphere"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000526,"raw_usage":{"total_tokens":2549,"prompt_tokens":962,"completion_tokens":1587,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":578,"completion_tokens_details":{"reasoning_tokens":1502}},"tokens_in":578,"tokens_out":1587,"duration_ms":11813,"temperature":1.0,"reasoning_tokens":1502,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-09T12:11:53.810239+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Run two simulations that swap the two protocols: take the original outer-sphere starting geometry (O2 far from NG) and apply the -0.9 eV shift, and take the inner-sphere modified geometry (O2 near NG, with the steering method) with no bias or a +2.0 eV shift; if the far geometry still reduces via solvent-mediated outer-sphere charge transfer and the near geometry still adsorbs O2 and dissociates it, then the mechanism is controlled by geometry rather than by the applied bias.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the extended Lagrangian formulation that treats the electronic degrees of freedom as auxiliary dynamical variables, enabling efficient and stable propagation without repeated SCF iterations."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the general framework of XL-BOMD from DFT to charge relaxation models, establishing the residual-error-based accuracy and stability criteria used here."},{"cited_title":"Self-consistent-charge density-functional tight-binding method for simulations of complex materials properties","cited_arxiv_id":null,"evidence_quote":"Defines the SCC-DFTB Hamiltonian and charge fluctuation formalism that underlies the DFTB-based electronic structure used in all simulations."},{"cited_title":"DFTB+, a Sparse Matrix-Based Implementation of the DFTB Method","cited_arxiv_id":null,"evidence_quote":"Identifies DFTB+ as an implementation of SCC-DFTB, giving the baseline speed and accuracy comparison for the method."},{"cited_title":"J.; Coe, J","cited_arxiv_id":null,"evidence_quote":"Supplies the LATTE Hamiltonian package with a transferable parametrization that the authors use for the NG-water-O2 simulations."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Describes the graph-based quantum response theory and shadow Born-Oppenheimer MD approach that forms the basis of the GPMDK code used for the simulations."},{"cited_title":"A.; Moxley, M","cited_arxiv_id":null,"evidence_quote":"Provides the steering method used to induce diffusion of O2 toward the NG surface in the inner-sphere simulations, establishing the protocol for the low-bias pathway."},{"cited_title":"K.; Rossmeisl, J.; Logadottir, A.; Lindqvist, L.; Kitchin, J","cited_arxiv_id":null,"evidence_quote":"Gives the computational hydrogen electrode approach that the paper contrasts with, highlighting that static DFT cannot describe non-adsorptive charge transfer."},{"cited_title":"The absolute electrode potential: an explanatory note","cited_arxiv_id":null,"evidence_quote":"Supplies the absolute electrode potential reference value (4.44 V) used to convert the electron chemical potential differences into effective overpotentials."},{"cited_title":"Mind the gap! Mat","cited_arxiv_id":null,"evidence_quote":"Supports the use of HOMO-LUMO gap closures as an indicator of reaction events, which the authors use to identify the two electron-injection steps."}],"review_version":1}