{"id":"ff833d37-4b82-45db-8318-830e9c7c3300","arxiv_id":"2508.05575","paper_version":1,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":9,"one_line_summary":"A triangular trimer of TBTAP molecules on Pb(111) shows negative differential conductance, current decreasing with increasing voltage, caused by Coulomb blockade and inter-molecular capacitive coupling and reproduced by a three-impurity Anderson model.","lead":"Researchers built a triangle of three organic molecules on a lead surface and saw the current drop as voltage increased, a molecular-scale version of a tunnel diode. The effect is explained by electrons blocking each other on the molecular cluster, and the same building blocks can be rearranged to create different device functions.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"PME simulation of NDC uses Γs≈0.01 meV, two to three orders below the Γs=10–20 meV inferred from YSR fits; the authors' 'ring size is independent of Γ' defense doesn't cover the non-equilibrium NDC mechanism, which depends on Γs.","rationale":"The experimental observation of NDC in the TBTAP trimer is well supported and is not in question. The load-bearing issue is the theoretical attribution: the paper's mechanism requires weak coupling to the substrate, yet the same experiment (via YSR spectroscopy) implies strong coupling. The authors' defense addresses ring positions but not the NDC mechanism, which is a non-equilibrium population effect. This is not an external disagreement; it is an internal inconsistency between two calculations presented in the same paper, explicitly acknowledged in the text. A sensitivity analysis at intermediate and realistic Γs would settle it. The reader's weakest assumption identifies exactly this point, so I agree. The verdict should remain conditional: the experimental result stands, but the quantitative theoretical explanation requires either a proper treatment at realistic Γs or a demonstration that NDC is insensitive to Γs, not just that ring positions are.","tokens_in":21683,"tokens_out":5871,"duration_ms":59896,"concrete_test":"Repeat the PME simulation for the Fig. 3 cut at Vs = 0.79 V with Γs = 1, 5, 10, 20 meV (and Γt scaled accordingly), keeping all electrostatic parameters fixed. Determine whether the central NDC dip in the computed dI/dV and the mixed populations of |100⟩, |010⟩, |001⟩ survive. Because Γs = 10–20 meV lies outside PME validity, additionally compare with a method that treats finite Γs non-perturbatively (e.g., finite-bias NRG or Hubbard NEGF) at one representative voltage. If NDC disappears or changes qualitatively, the load-bearing assumption is falsified; if it survives, the concern is resolved.","verdict_should_be":"UNCHANGED","load_bearing_attack":"In the Supplementary Text (Pauli master equations), the authors set Γs ∼ 0.01 meV ≪ kBTexp = 0.22 meV to stay in the sequential-tunneling regime, while noting that YSR fits for the same TBTAP/Pb(111) system give Γs,exp = 10–20 meV [27]. Their justification is that 'the size and shape of the discharge rings do not depend on the value of Γ, but rather on the electrostatic details,' and that only the ring width is proportional to Γs and Γt. Even if this is true for the locus of charge-state transitions, it does not protect the central NDC mechanism. The NDC in Fig. 3B is explicitly a dynamical effect: it arises because the steady-state populations are spread over singly occupied states |100⟩, |010⟩, |001⟩, which are frozen by the absence of inter-site hopping. These populations are the solution of the master equations (Eq. S7) and depend on Γs and Γt. With Γs = 10–20 meV, the substrate relaxation rate exceeds kBT/ħ by orders of magnitude; the cluster should equilibrate to the substrate Fermi level and the non-equilibrium occupancy that produces the central NDC region should be suppressed. The paper provides no calculation or estimate at realistic Γs. Moreover, the YSR NRG calculation in Fig. S10G/S11 uses Γ = 20 meV, so the two theoretical pillars of the paper operate in mutually inconsistent coupling regimes. The central claim that PME 'quantitatively reproduces' the NDC therefore rests on an assumption the authors themselves flag as 'questionable,' and the one-sentence justification addresses ring geometry rather than the NDC mechanism.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports the assembly of a triangular trimer of TBTAP radical molecules on Pb(111) and its characterization by low-temperature STM/STS. The central experimental observation is a negative differential conductance (NDC) region between roughly 0.7 and 0.9 V, accompanied by voltage-dependent discharge rings in dI/dV maps. The authors model the trimer with a three-impurity Anderson model, solve the equilibrium low-energy physics with NRG, and simulate the finite-bias transport with a Pauli master equation (PME). They attribute the NDC to non-equilibrium occupation of singly occupied states caused by strong inter-site capacitive coupling and negligible direct hopping. A hexamer structure is also presented as a topological variant. The claim is that the NDC is purely electronic, independent of superconductivity, and quantitatively reproduced by the model.","tokens_in":22180,"tokens_out":3974,"duration_ms":46513,"significance":"If the theoretical account is correct, the work would be a valuable demonstration of a molecular-scale, gate-tunable NDC element whose functionality is controlled by cluster geometry rather than by the molecule-lead interface. The experimental data are of high quality: the manipulation is documented, the discharge-ring evolution is systematic, the chiral patterns are reproducible, and the finite Hilbert space of the trimer makes the modeling transparent. The authors also provide code links and use open-source packages, which is a strength. However, the manuscript's central quantitative claim rests on a PME calculation performed in a coupling regime that the authors themselves concede is questionable and that is inconsistent with the coupling used in their own YSR/NRG analysis. This weakens, but does not necessarily invalidate, the proposed NDC mechanism.","major_comments":[{"comment":"The PME calculation sets Γs≈0.01 meV ≪ kBTexp, explicitly outside the Γs=10–20 meV inferred from YSR fits [27] and used in the NRG calculation of Figs. S10G/S11. The authors state that the ring size is independent of Γ, but that statement concerns the equilibrium charge-state locus, not the NDC mechanism. The NDC in Fig. 3B is a dynamical steady-state effect: it requires the population to be spread over the singly occupied states |100⟩, |010⟩, |001⟩ via Eq. (S7), and those populations depend on Γs and Γt. No calculation is shown for Γs≈10–20 meV; at that coupling the substrate relaxation rate exceeds kBT/ℏ and the non-equilibrium occupancy should be strongly suppressed. Please provide a calculation in the broadened-coupling regime (or a convincing argument, e.g., an adiabatic elimination or finite-Γ master equation) showing that the NDC survives, or reframe the claim accordingly.","section":"Supplementary Text, Eq. (S7)-(S8); main text 'Negative differential conductance emerging from non-equilibrium occupancy"},{"comment":"The phrase 'quantitatively reproduces the experimental observations' is stronger than the evidence supports. The simulation maps in Fig. 2A’-H’ are in arbitrary units and the model parameters (εid=-90 meV, W=50 meV, t=0, the tip electrostatic parameters rtip, ztip, zC, zS, and the quadrupole moment Qxx) are fitted to the same discharge-ring data that the model then reproduces. The YSR NRG check (Fig. S10G/S11) is not an independent validation because it uses the same U, W, t and ε≈-104 meV inferred from the charging-ring analysis. I recommend presenting the comparison as a self-consistent parametrization or semi-quantitative reproduction and adding a sensitivity analysis for at least W and the tip parameters.","section":"Main text 'Triangular trimer assemblies and their discharging behavior'; Fig. 2 and Fig. S7"}],"minor_comments":[{"comment":"The caption lists negative sample voltages (Vs = −0.37 V, −0.39 V, ...) but the panel labels show 'VS = 0.37 V' etc. The missing minus signs should be corrected for clarity.","section":"Figure 4 caption and panels"},{"comment":"The phrase 'quantitatively reproduces' should be reconciled with the arbitrary-unit simulations in Fig. 2 and Fig. S7; otherwise the claim is misleading.","section":"Abstract and main text"},{"comment":"Several typographical/spacing errors remain, e.g., 'an molecular system', 'e fficiently', 'di fferential'. A careful proofread is needed.","section":"Throughout"}],"recommendation":"major_revision","confidential_remarks":"The experimental part appears solid and the idea of using cluster topology to engineer NDC is timely. My main concern is the internal coupling-regime inconsistency: the PME transport theory operates with Γs≈0.01 meV while the same paper's YSR analysis uses Γs≈20 meV, and the NDC mechanism depends on the steady-state populations that are sensitive to Γs. This is not a minor presentational issue; it directly affects the central theoretical claim. I recommend major revision rather than rejection because the experimental findings may well be correct, but the manuscript should either extend the transport calculation to realistic couplings or substantially soften the quantitative-reproduction claims."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Bottom line: this is a good experimental paper with a load-bearing soft spot in the transport modeling. The first realization of NDC in a laterally manipulated C3-symmetric TBTAP trimer on Pb(111) is real, and the spatial dI/dV maps showing discharge rings and their overlap, plus the chiral NDC regions, are convincing. The hexamer control experiment adds a nice topological knob. Credit where due: the authors ship their simulation code and deposit data, and the PME framework is transparent enough that the mechanism (frustrated charge states, negligible inter-site hopping, non-equilibrium occupancies of |100>, |010>, |001>) can be checked. The YSR NRG calculation is a separate consistency cross-check.\n\nThe soft spot is exactly where the stress-test lands. The PME calculation sets Gamma_s ≈ 0.01 meV << kBT = 0.22 meV to stay in sequential tunneling, while the same group's YSR fits give Gamma_s = 10–20 meV. The authors' defense — ring size and shape depend on electrostatics, not Gamma — may hold for the discharge-ring locus, but it does not protect the NDC mechanism. The NDC arises from steady-state populations spread over singly occupied states; those populations are the solution of the master equation and depend directly on Gamma_s and Gamma_t. At Gamma_s = 10–20 meV, the substrate relaxation rate vastly exceeds kBT/hbar, and the non-equilibrium occupancy that produces the central NDC should be suppressed. The paper offers no calculation at realistic Gamma_s. And the two pillars of the theory run in mutually inconsistent regimes: the PME uses 0.01 meV, the YSR NRG uses 20 meV. This is not a fatal flaw in the experimental observation, but it is a real gap in the explanation.\n\nThere is also the circularity burden: the tip-electrostatic parameters, W, epsilon_id, and the quadrupole moment are fitted to the very discharge-ring data the model then reproduces. That's less alarming than it sounds — the model has enough structure to fail elsewhere, and the YSR cross-check is independent — but the \"quantitative reproduction\" claim should be softened to \"consistent with\" unless a sensitivity analysis is provided. Error bars on the fitted parameters would help.\n\nWho is this for: STM/STS experimentalists working on molecular charge states and Coulomb blockade, and theorists interested in master-equation models of few-site correlated systems. It deserves a serious referee; the central experiment will be of interest even if the transport model needs revision. My recommendation: send to peer review, ask for a sensitivity analysis of Gamma_s and an explicit acknowledgment of the regime inconsistency, and require that the non-equilibrium NDC mechanism be demonstrated (or refuted) at Gamma_s values extracted from YSR.","headline":"Solid experimental NDC in a molecular trimer with spatial Coulomb-ring mapping; the PME 'quantitative reproduction' rests on an explicitly questionable weak-coupling assumption that the stress-test correctly identifies as unprotected for the NDC mechanism itself.","tokens_in":22678,"tokens_out":3840,"would_cite":true,"duration_ms":36214,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["73.23.Hk","73.63.-b","85.65.+h"],"model":"deepseek-v4-flash","headline":"A triangular cluster of three TBTAP molecules on lead produces negative differential conductance between 0.7 and 0.9 V, driven by inter-molecular Coulomb correlations rather than superconductivity.","keywords":["negative differential conductance","Coulomb blockade","molecular trimer","three-impurity Anderson model","Pauli master equation","capacitive coupling","Yu–Shiba–Rusinov states","scanning tunneling spectroscopy"],"falsifier":"Two checks would settle the claim. (1) Re-run the Pauli master equation with the experimentally fitted substrate coupling $\\Gamma_s = 10$\\u2013$20$ meV and compare the simulated dI/dV maps with the measured discharge rings and NDC patterns; if the rings or the NDC region change shape or vanish, the electrostatic-only assertion fails. (2) Measure the trimer on a normal-metal substrate (e.g., Ag(111)): the claim that NDC is independent of superconductivity predicts the same 0.7–0.9 V NDC there, and its absence would refute that independence.","tokens_in":21595,"feed_emoji":"🔬","tokens_out":14311,"duration_ms":121441,"temperature":0.7,"pith_summary":"The paper reports a three-molecule device that shows negative differential conductance (NDC): between 0.7 and 0.9 V the current drops as the voltage rises. The device is a C3-symmetric triangular trimer of TBTAP molecules assembled on a Pb(111) surface by lateral STM manipulation, and the authors reproduce its behavior quantitatively with a three-impurity Anderson model solved by Pauli master equations. The mechanism is capacitive: the inter-molecular coupling ($W = 50$ meV) makes the two-electron charge state of the cluster more stable than the three-electron state, and under bias the cluster is driven into a non-equilibrium mixture of singly-occupied states in which the empty site often sits far from the tip, suppressing the tunneling current. The authors argue the effect is pure electron-correlation physics, unrelated to the superconducting substrate, and that the same molecules can be re-programmed by cluster geometry: a linear tetramer behaves as a memory, the triangular trimer as a molecular analogue of a Gunn diode.","feed_headline":"Three molecules in a triangle make current fall as voltage rises","feed_subtitle":"The effect is pure electron-correlation physics, no superconductivity needed — a wiring-free route to molecular logic.","key_machinery":"The three-impurity Anderson model (TIAM): three molecular sites at $-90$ meV with on-site repulsion $U = 200$ meV, inter-site capacitive coupling $W = 50$ meV, and negligible direct hopping, weakly coupled to tip and substrate. Large $U$ forbids double occupancy, shrinking the Hilbert space from 64 to 8 states and making the dynamics solvable by Pauli master equations. The tip acts two ways: its electrostatic field (plate capacitor with spherical apex and mirror charge) shifts site energies; its tunneling amplitude decays exponentially with site distance, $t_{ti} \\propto \\exp(-\\beta|r_i - r_t|)$. Molecular orbital shape (the SOMO with a fitted quadrupole moment) reproduces the chiral pattern","core_discovery":"Central claim: the NDC between 0.7 and 0.9 V in a C3-symmetric trimer of TBTAP molecules on Pb(111) is a Coulomb-blockade effect driven by inter-molecular capacitive coupling, purely electron-correlation physics, independent of the superconducting substrate. Many sharp peaks in the dI/dV maps are intra-cluster charge rearrangements, not total-charge changes. The NDC appears between excited-state crossings, where singly occupied states (e.g., $|100\\rangle$) fall below doubly occupied states; without inter-site hopping, electrons cannot relax, occupancy spreads across the cluster, and tunneling is suppressed because the empty site is often far from the tip. The non-superconducting model reprod","pith_inferences":["If the mechanism is right, the NDC voltage window (0.7–0.9 V) should be tunable by modifying the inter-site capacitance $W$ — for example, by changing molecular spacing, linker chemistry, or substrate screening — since $W$ sets the separation between doubly and triply occupied configurations.","The non-equilibrium-occupancy mechanism predicts similar NDC in other frustrated geometries of strongly coupled charge sites with negligible inter-site hopping, such as triangular arrays of different radical molecules; the essential ingredients are capacitive coupling and threefold frustration, not the specific molecule.","The paper's assertion that discharge-ring geometry is independent of the substrate coupling $\\Gamma_s$ is directly testable: at higher temperatures the ring edges should broaden with $k_B T$ rather than with $\\Gamma_s$, and cotunneling signatures would appear if the weakly coupled assumption breaks down.","The YSR-validated parameters imply a doubly occupied, spin-carrying trimer ground state; the same platform might therefore combine NDC switching at high bias with spin-sensitive transport at low bias, extending toward the hybrid superconducting devices the authors mention as an outlook."],"forward_implications":["The same molecular building block implements different devices by geometry alone: a linear tetramer acts as a bistable memory, while the triangular trimer acts as a molecular analogue of a Gunn diode, with no change in chemistry.","Because the NDC is independent of the superconducting substrate, the trimer should function on normal-metal surfaces, cleanly separating the Coulomb-blockade physics from YSR and superconductivity physics.","The eight-state Hilbert space of the trimer makes it a fully simulable unit, offering a tractable building block for designing molecular logic and charge-state computing architectures.","Charge-state readout of such clusters cannot rely on counting dI/dV peaks as total-charge toggles: many peaks signal charge redistribution among sites at constant total charge.","The hexamer demonstrates electrostatic gating of inner sites by outer molecules, extending the design space from three-site clusters toward larger, programmable arrays."],"supporting_citations":[{"why":"Prior TBTAP dimer study: supplies the YSR-fitted substrate coupling $\\Gamma_s = 10$\\u2013$20$ meV and the dimer $W$ value the trimer fit must match, plus the bistable switching the trimer is contrasted with.","marker":"[27]"},{"why":"Earlier TBTAP-on-Ag(111) study establishing the molecules' radical nature, the on-site $U = 200$ meV, and tip-gated discharging used as the experimental baseline.","marker":"[25]"},{"why":"Open-source master-equation package (QmeQ) whose implementation the Pauli master equation code is derived from and verified against.","marker":"[42]"},{"why":"Reference for the Pauli master equation approach to sequential transport through multi-dot systems.","marker":"[52]"},{"why":"NRG solver used to compute the YSR spectra of the superconducting three-impurity model.","marker":"[41]"},{"why":"Provides the charge-gating NDC mechanism (charging opens a tunneling channel while raising the barrier) invoked for the hexamer case.","marker":"[18]"},{"why":"Theoretical prediction of symmetry-controlled negative differential resistance in a triangular molecule, the phenomenon this work realizes experimentally.","marker":"[11]"},{"why":"Imaging of discharge cascades in WS2/WSe2 moiré lattices, the comparison point for the flower-like ring patterns (there without NDC).","marker":"[36]"}],"fun_headline_variants":["Triangle of molecules: voltage up, current down","Tiny triangle of molecules shows negative resistance","No superconductivity needed: molecule triangle blocks current","Molecule triangle's Coulomb blockade bends the current–voltage curve","Three molecules in a triangle: a natural current limiter"],"cache_read_input_tokens":2816,"weakest_assumption_plain":"The load-bearing premise is that the discharge-ring geometry and the NDC pattern depend only on the local electrostatic environment, so the simulations can use a substrate coupling of $\\Gamma_s \\sim 0.01$ meV even though the YSR fit gives $\\Gamma_s = 10$\\u2013$20$ meV. The paper states this without derivation in the transport-calculations section, calling the weak-coupling assumption 'questionable'; if the electrostatic-only assertion is wrong, the quantitative reproduction o","fun_headline_variants_meta":{"raw":{"variants":["Triangle of molecules: voltage up, current down","Tiny triangle of molecules shows negative resistance","No superconductivity needed: molecule triangle blocks current","Molecule triangle's Coulomb blockade bends the current–voltage curve","Three molecules in a triangle: a natural current limiter"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001321,"raw_usage":{"total_tokens":5215,"prompt_tokens":740,"completion_tokens":4475,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":484,"completion_tokens_details":{"reasoning_tokens":4399}},"tokens_in":484,"tokens_out":4475,"duration_ms":37714,"temperature":1.0,"reasoning_tokens":4399,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-05T23:13:35.302940+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Two checks would settle the claim. (1) Re-run the Pauli master equation with the experimentally fitted substrate coupling $\\Gamma_s = 10$\\u2013$20$ meV and compare the simulated dI/dV maps with the measured discharge rings and NDC patterns; if the rings or the NDC region change shape or vanish, the electrostatic-only assertion fails. (2) Measure the trimer on a normal-metal substrate (e.g., Ag(111)): the claim that NDC is independent of superconductivity predicts the same 0.7–0.9 V NDC there, and its absence would refute that independence.","supporting_citations":[{"cited_title":"Li, et al., Individual Assembly of Radical Molecules on Superconductors: Demonstrating Quantum Spin Behavior and Bistable Charge Rearrangement","cited_arxiv_id":null,"evidence_quote":"Prior TBTAP dimer study: supplies the YSR-fitted substrate coupling $\\Gamma_s = 10$\\u2013$20$ meV and the dimer $W$ value the trimer fit must match, plus the bistable switching the trimer is contrasted with."},{"cited_title":"Li, et al., Strong signature of electron-vibration coupling in molecules on Ag(111) trig- gered by tip-gated discharging","cited_arxiv_id":null,"evidence_quote":"Earlier TBTAP-on-Ag(111) study establishing the molecules' radical nature, the on-site $U = 200$ meV, and tip-gated discharging used as the experimental baseline."},{"cited_title":"Kir ˇsanskas, J","cited_arxiv_id":null,"evidence_quote":"Open-source master-equation package (QmeQ) whose implementation the Pauli master equation code is derived from and verified against."},{"cited_title":"Goldozian, F","cited_arxiv_id":null,"evidence_quote":"Reference for the Pauli master equation approach to sequential transport through multi-dot systems."},{"cited_title":"ˇZitko, NRG Ljubljana (2021), doi:10.5281 /zenodo.4841076, https://doi.org/10","cited_arxiv_id":null,"evidence_quote":"NRG solver used to compute the YSR spectra of the superconducting three-impurity model."},{"cited_title":"Fern ´andez-Torrente, D","cited_arxiv_id":null,"evidence_quote":"Provides the charge-gating NDC mechanism (charging opens a tunneling channel while raising the barrier) invoked for the hexamer case."},{"cited_title":"Kostyrko, B","cited_arxiv_id":null,"evidence_quote":"Theoretical prediction of symmetry-controlled negative differential resistance in a triangular molecule, the phenomenon this work realizes experimentally."},{"cited_title":"Li, et al., Imaging local discharge cascades for correlated electrons in WS 2/WSe2 moir´e superlattices","cited_arxiv_id":null,"evidence_quote":"Imaging of discharge cascades in WS2/WSe2 moiré lattices, the comparison point for the flower-like ring patterns (there without NDC)."}],"review_version":1}