{"id":"42814a7c-43f4-4173-9e93-194b4eb5febd","arxiv_id":"2507.10146","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"A whispering-gallery-mode microcavity with plasmonic gold nanoparticles detects and distinguishes single molecules of GABA, glutamate and dopamine, and tells GABA apart from glutamate by their binding behavior.","lead":"This paper reports a light-based sensor that detects single neurotransmitter molecules (GABA, glutamate, dopamine) down to extremely low concentrations of 10 attomolar. It may give neuroscientists a label-free tool to watch neurotransmitter molecules interact in real time.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The single-molecule and 10 aM claims hinge on identifying each resonance step as one analyte molecule, but the paper reports no blank-buffer control and no concentration-dependent event-rate scaling, so this identification is unverified.","rationale":"The reader's weakest assumption is exactly the load-bearing point: each discrete WGM step is interpreted as a single neurotransmitter binding event, and this interpretation is not independently calibrated. The paper provides plausible supporting evidence — Poisson inter-event statistics, DFT interaction energies, SERS verification, and a qualitative polarizability trend — but none of these rules out alternative sources of step-like shifts. The Methods section explicitly describes a semi-manual, visually guided step-selection procedure, and no negative control or concentration-scaling test is presented. Because a 10 aM solution in the experimental volume contains only thousands of molecules, even a small number of spurious events could masquerade as detection. My concern does not contradict the paper's internal consistency; rather, it identifies an unverified identification step that is central to the abstract's strongest claims. The appropriate verdict remains CONDITIONAL: the claims are not disproven, but they require an explicit calibration experiment before acceptance. I agree with the reader's framing and would not change the verdict.","tokens_in":16589,"tokens_out":2944,"duration_ms":38315,"concrete_test":"Run the identical AuNS-functionalized WGM sensor and data-processing pipeline on (1) filtered phosphate buffer alone (no neurotransmitter) and (2) a dilution series of GABA and glutamate from 10 aM to 1 fM (e.g., 0, 10 aM, 100 aM, 1 fM), using freshly prepared, filtered solutions and fresh sensor assemblies as in Methods. Count step and spike events with the same semi-manual 3σ protocol and report event rates. If the buffer-only rate overlaps the 10 aM analyte rate, or if the event rate is not monotonic and approximately linear with concentration over at least two orders of magnitude, the single-molecule calibration fails.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The core claim that the sensor detects and discriminates individual neurotransmitter molecules at 10 aM depends on the assertion in Experimental Results that 'each step corresponding to a single-molecule event.' The supporting evidence is (i) step-like resonance shifts, (ii) Poissonian inter-event intervals (Supp. Fig. S3), and (iii) DFT-computed binding energies. None of these independently establishes that a step is one molecule rather than a cluster, an aggregate, a surface rearrangement, or a buffer/contamination artifact. Methods 3b states that steps were 'first located by visual inspection of the time trace,' and the amplitude was then fit between two manually identified linear segments. No buffer-only negative control is reported anywhere, and event rates are never compared across a dilution series to test the expected linear concentration dependence. This matters quantitatively: in the 300 µL PDMS chamber, a 10 aM solution contains only about 1.8×10^3 analyte molecules, so a few impurity events or noise features passing the 3σ threshold could produce the same traces. The single-molecule interpretation is the pivot for all three headline claims: 10 aM sensitivity, single-molecule detection, and the polarizability correlation. If a step is not one molecule, those claims lose support regardless of the DFT and SERS corroboration.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript reports an optoplasmonic whispering-gallery-mode (WGM) sensor, in which gold nanorods or nanostars are immobilized on a microsphere resonator, for label-free detection of the neurotransmitters GABA, glutamate, and dopamine. The authors claim single-molecule detection down to 10 aM, discrimination between structurally similar GABA and glutamate through pH-controlled surface chemistry, and a correlation between mean WGM step heights and DFT-computed molecular polarizabilities. Supporting material includes DFT interaction energies for neurotransmitter–phosphate–Au(111) complexes, SERS spectra, FDTD/FEM field simulations, and survivor-function kinetic analyses of transient spike events. The central detection and discrimination narrative rests on interpreting discrete step-like resonance shifts as single-molecule binding events.","tokens_in":16815,"tokens_out":2254,"duration_ms":29056,"significance":"If the central claims hold, the work would be a meaningful advance in label-free single-molecule biosensing of small neurotransmitters, with potential relevance to neuroscience applications. The paper has genuine strengths: the polarizability values are computed independently by DFT rather than fitted to the WGM shifts, so the reported correlation is not circular; the DFT and SERS characterization of the phosphate-mediated adsorption mechanism is a useful addition; and the kinetic survivor-function analyses provide quantitative on/off rates. However, the single-molecule and 10 aM claims are currently under-supported by controls and by the semi-manual event-selection procedure. The paper needs additional verification before the headline claims can be accepted as established.","major_comments":[{"comment":"The assertion that each step in the WGM resonance trace corresponds to a single-molecule event is load-bearing for the 10 aM sensitivity, the single-molecule detection, and the polarizability correlation, but it is not supported by the reported controls. No buffer-only negative control is described anywhere, and no concentration-dependent event-rate scaling is shown to verify that the number of detected events varies linearly with analyte concentration. Methods 3b states that step events were 'first located by visual inspection of the time trace,' and the step amplitude was then obtained from two manually identified linear segments. Given that a 10 aM solution in the 300 µL chamber contains on the order of 1.8×10^3 analyte molecules, contamination or drift features passing the 3σ threshold could produce similar traces. The authors should add blank-buffer controls, analyte-free chamber controls, and a dilution series in which event rates are compared with expectation; without these, the step-signal interpretation remains unverified.","section":"Experimental Results and Methods 3b"},{"comment":"The step-height analysis relies on a semi-manual visual selection of steps before fitting. This creates a risk of selection bias, particularly because the log-normal step-height distributions are broad and overlap substantially between molecules. The manuscript does not report any reproducibility measure for the step identification, such as inter-operator agreement, a blinded analysis, or an automated step-detection algorithm applied independently. Since the polarizability correlation and the mean step-height comparisons in Fig. 1e are based on these manually selected steps, the authors should quantify the selection uncertainty or provide an automated detection routine with a defined threshold and demonstrate that the conclusions are unchanged.","section":"Methods 3b, Fig. 1d"},{"comment":"The claimed correlation between mean WGM step height and molecular polarizability rests on only three analytes (GABA, glutamate, dopamine). With n=3, the correlation is anecdotal rather than statistically established, and the error bars on the mean step heights are large (e.g., GABA 4.5±2.4 fm, glutamate 5.2±1.7 fm, dopamine 5.7±1.6 fm). The manuscript should either report a proper statistical test with a confidence interval, include additional small molecules to increase the number of independent data points, or explicitly soften the claim to a qualitative trend. As written, the phrasing 'strongly correlates' overstates the support.","section":"Fig. 1e, Discussion"},{"comment":"Dopamine measurements were performed in PBS at pH 6.5, whereas GABA and glutamate were measured in phosphate buffer at pH 7.4. This pH/buffer difference is a confound for the comparison of mean step heights across the three molecules: the surface charge, ionic strength, and protonation states differ between dopamine and the other two analytes. The manuscript should address whether the larger mean step height for dopamine could be caused by the different buffer/pH conditions rather than by molecular polarizability, and ideally include a measurement of at least one neurotransmitter under both pH conditions to separate the effects.","section":"Experimental Results, Fig. 1e"}],"minor_comments":[{"comment":"The caption contains two entries labeled 'd,' one for the histograms of step heights and one for the amplitude-of-step histograms; the labeling should be corrected to sequential letters (d and e), and the subsequent panels should be renumbered accordingly.","section":"Fig. 1 caption"},{"comment":"The time axes in Fig. 1b,c are described as relative, with traces aligned to the first detected event and displayed in a 20 s window. It would improve clarity to state explicitly in the caption whether the same 20 s window is used for all traces and whether the number of displayed events is limited by the window rather than by the full measurement duration.","section":"Fig. 1 caption"},{"comment":"The background noise σ is estimated as the minimum standard deviation across windows of N points, with N ranging from 10 to 1,000. This procedure can underestimate the noise level because the minimum of many window estimates is itself a downward-biased estimator. The authors should justify this choice or use a robust noise estimator such as the median absolute deviation.","section":"Methods 3a"},{"comment":"The statement that '3-MPA functionalization enables selective differentiation' but 'introduces slightly higher noise' is not quantified. Reporting the noise level before and after functionalization would help the reader assess the stated trade-off between selectivity and sensitivity.","section":"Discussion"},{"comment":"The Poissonian-interval analysis is presented as corroboration of single-molecule binding, but the histogram of time intervals alone does not distinguish single-molecule events from aggregates or surface rearrangements. The text should note this limitation explicitly rather than presenting the Poisson statistics as sufficient evidence.","section":"Supplementary Fig. S3"}],"recommendation":"major_revision","confidential_remarks":"The paper is within the scope of the journal and the experimental platform is interesting. The main concern is that the headline claims—single-molecule detection at 10 aM and the polarizability correlation—are not yet robust to the absence of negative controls, the lack of concentration-scaling data, and the semi-manual step selection. These issues are fixable with additional experiments and analyses, so I recommend major revision rather than rejection. I have no concerns about the authors' novelty disclosure or citation practices."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"One thing you should know up front: the GABA vs glutamate discrimination part is the real contribution here, and it holds up reasonably well. The authors use 3-MPA functionalization at pH 9.7 so that GABA hydrogen-bonds and produces steps while glutamate repels and spikes, and they show the two signatures side by side in an equimolar mixture. That is a clever, testable strategy, and it is new relative to the earlier Baaske work on nucleic acids. The DFT and SERS corroboration for the phosphate-bridging interaction is a plus, and the authors are candid about the trade-offs: low event counts at attomolar levels, noise from 3-MPA, the dopamine polymerization problem.\n\nWhat I do not buy yet is the 10 aM single-molecule detection. The paper reports no buffer-only negative control anywhere, and no concentration series showing that the event rate scales linearly with analyte concentration. Without that, 'each step corresponds to a single molecule' is an assumption, not a demonstrated fact. In a 300 µL chamber at 10 aM there are only about 1.8×10^3 molecules, so a handful of impurity hits or surface rearrangements could produce the same traces. The Poissonian inter-event statistics in Supp S3 are suggestive but do not distinguish one molecule per step from rare cluster events. The step extraction is also semi-manual—visual location followed by linear fits—which risks selection bias, though the 3σ threshold helps.\n\nThe polarizability correlation rests on only three molecules, and the step-height distributions overlap substantially; calling it a 'strong correlation' is a stretch. To be fair, the polarizabilities come from DFT, not from fitting the WGM shifts, so there is no circularity. Still, it is an interesting pointer, not strong evidence.\n\nOverall, this is a serious experimental paper from a group that knows this platform. The differentiation strategy deserves a real referee. The 10 aM claim needs revision: add negative controls, show event-rate scaling over at least 10 aM–1 µM, and ideally blind the step detection. If those come in, the paper could be quite valuable. For now, I would not cite the 10 aM sensitivity, but I would point colleagues to the pH-tuned discrimination idea.","headline":"Solid pH-driven GABA/glutamate discrimination strategy, but the 10 aM single-molecule claim needs a buffer control and concentration scaling before it can be taken at face value.","tokens_in":17380,"tokens_out":2811,"would_cite":false,"duration_ms":33220,"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":"This paper claims that a whispering-gallery-mode microcavity decorated with plasmonic gold nanostars detects individual neurotransmitter molecules at concentrations down to 10 attomolar and distinguishes the structurally similar molecules…","keywords":["opto-plasmonic WGM sensor","single-molecule detection","neurotransmitter sensing","gold nanostars","plasmonic hotspot","molecular polarizability","label-free biosensor","GABA glutamate discrimination"],"falsifier":"A decisive test would be to record the sensor response over a dilution series spanning, for example, 10 aM to 1 µM and check whether the rate of step events scales linearly with concentration, as expected for single-molecule binding; if the event rate saturates nonlinearly or if steps split into sub-steps when the data acquisition rate is increased, the one-step-per-molecule assignment would be falsified.","tokens_in":16393,"feed_emoji":"🧠","tokens_out":3452,"duration_ms":41044,"temperature":0.7,"pith_summary":"The paper aims to show that a microsphere-based whispering-gallery-mode (WGM) optical sensor, enhanced by attached gold nanostars, can detect single neurotransmitter molecules—GABA, glutamate, and dopamine—at concentrations as low as 10 attomolar. It further claims that the sensor can discriminate between structurally similar neurotransmitters such as GABA and glutamate by tuning the pH and surface chemistry so that one analyte produces permanent step-like resonance shifts and the other produces transient spike-like shifts. The authors also report that the average size of the resonance shift caused by one binding event correlates with the molecule's computed polarizability, suggesting the sensor reads an intrinsic electronic property of the analyte. If true, this would provide a label-free, real-time method for studying neurotransmission at the single-molecule level, with potential applications in neuroscience research and diagnostics.","feed_headline":"Single-molecule sensor sees neurotransmitters at 10 attomolar","feed_subtitle":"Hybrid WGM-plasmonic platform tells GABA from glutamate and links signal size to polarizability.","key_machinery":"The central object is the optoplasmonic WGM sensor: a silica microsphere (about 85–90 µm in diameter) supporting a whispering-gallery mode, a circulating optical resonance, whose evanescent field couples to localized surface plasmons of CTAB-capped gold nanostars (and nanorods) immobilized on the microsphere surface. The argument is carried by the step-like and spike-like shifts of the resonance wavelength that occur when a neurotransmitter molecule binds or transiently interacts at a plasmonic hotspot; the shift magnitude is taken to be proportional to the real part of the molecule's polarizability. Supporting machinery includes density functional theory calculations of interaction energies on a Au(111) slab in the presence of phosphate and carbonate anions, and kinetic analysis of inter-event time and dwell-time distributions using a Poissonian survivor-function model.","core_discovery":"The central claim is that a WGM microcavity hybridized with plasmonic nanospiked gold antennas detects and discriminates among GABA, glutamate, and dopamine at the single-molecule level, with detection demonstrated at concentrations down to 10 aM. Discrete step-like shifts in the WGM resonance are interpreted as irreversible single-molecule binding events, and transient spike-like shifts as weak, short-lived interactions. The discrimination between GABA and glutamate is achieved by functionalizing gold nanorods with 3-mercaptopropionic acid and operating at pH 9.7, where GABA forms stable hydrogen bonds producing steps and glutamate experiences electrostatic repulsion producing spikes; the two signals appear simultaneously in an equimolar mixture. The average step height scales with the molecular polarizability computed by density functional theory, and DFT also explains the role of phosphate anions from the buffer as a hydrogen-bonding bridge between the gold surface and the protonated amine of the neurotransmitters.","pith_inferences":["If each step truly is one molecule, the platform could be used to count molecules and estimate absolute concentrations from event rates, a capability that ensemble sensors do not offer; this extension is not demonstrated in the paper.","The same pKa-based discrimination strategy could be generalized to other small biomolecules with ionizable groups and similar sizes, which is a testable extension beyond the three neurotransmitters studied.","The authors note that at 10 aM and 1 fM the limited number of spike events made statistical analyses not ideally smooth, implying the single-molecule and 10 aM claim rests on a modest number of events; an independent dilution series with predicted event-rate scaling would strengthen that inference."],"forward_implications":["If the sensor truly detects single molecules, it can track neurotransmitter release and reuptake in real time with millisecond resolution, matching the timescale of synaptic events.","Detection at 10 aM is orders of magnitude more sensitive than typical ensemble techniques with limits around 10 nM, opening the possibility of monitoring sub-femtomolar neurochemical dynamics.","The correlation between step height and molecular polarizability means the sensor reports an intrinsic electronic property, potentially allowing identification of small molecules without labels.","The pH- and interface-tuned discrimination of GABA from glutamate works even in an equimolar mixture, suggesting a route to multiplexed detection of chemically similar analytes.","Gold nanostars increase the number of detectable events by 35–52% compared with nanorods, providing a design principle for future plasmonic enhancement in single-molecule sensing."],"supporting_citations":[{"why":"Provides the precedent and method for interpreting discrete step-like resonance shifts in a WGM biosensor as single-molecule binding events.","marker":"[31]"},{"why":"Supplies the hybrid quantum-classical polarizability model that predicts the resonance shift magnitude is sensitive to the analyte's electronic polarizability.","marker":"[32]"},{"why":"States the standard relation that the WGM resonance shift is proportional to the real part of the interacting molecule's polarizability tensor.","marker":"[37]"},{"why":"Documents the speciation and adsorption of phosphate on gold electrodes, which supports the proposed phosphate-bridge interaction between buffer anions and protonated neurotransmitter amines.","marker":"[26]"},{"why":"Justifies the use of a log-normal distribution to model step-height variability arising from a multiplicative combination of underlying factors.","marker":"[29]"},{"why":"Establishes the optimal coupling technique for exciting whispering-gallery modes in microsphere resonators, which underlies the sensor's operation.","marker":"[41]"},{"why":"Supports the choice of the M06-2X density functional for accurately describing noncovalent interactions and adsorption on metal surfaces in the DFT calculations.","marker":"[33]"}],"fun_headline_variants":["Attomolar neurotransmitter detection with hybrid plasmonic sensor","Single-molecule neurochemical sensing at 10 attomolar","WGM sensor distinguishes GABA and glutamate at single-molecule level","Ultra-sensitive biosensor reads single neurotransmitters","Hybrid WGM-plasmonic sensor detects single molecules at 10 aM"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The central claim rests on the assumption that each discrete step in the WGM resonance trace corresponds to the binding of exactly one neurotransmitter molecule to a plasmonic hotspot, an interpretation inferred from step-like shifts and Poissonian inter-event statistics rather than from an independent single-molecule calibration.","fun_headline_variants_meta":{"raw":{"variants":["Attomolar neurotransmitter detection with hybrid plasmonic sensor","Single-molecule neurochemical sensing at 10 attomolar","WGM sensor distinguishes GABA and glutamate at single-molecule level","Ultra-sensitive biosensor reads single neurotransmitters","Hybrid WGM-plasmonic sensor detects single molecules at 10 aM"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000685,"raw_usage":{"total_tokens":3102,"prompt_tokens":937,"completion_tokens":2165,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":553,"completion_tokens_details":{"reasoning_tokens":2080}},"tokens_in":553,"tokens_out":2165,"duration_ms":16107,"temperature":1.0,"reasoning_tokens":2080,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T17:38:18.252558+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A decisive test would be to record the sensor response over a dilution series spanning, for example, 10 aM to 1 µM and check whether the rate of step events scales linearly with concentration, as expected for single-molecule binding; if the event rate saturates nonlinearly or if steps split into sub-steps when the data acquisition rate is increased, the one-step-per-molecule assignment would be falsified.","supporting_citations":[],"review_version":1}