{"id":"6fad689b-ccad-40eb-8c58-377445ce7d5d","arxiv_id":"2607.14915","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":5,"one_line_summary":"For the first time, a point-contact germanium detector's intrinsic spatial position resolution is mapped: non-uniform, anisotropic, energy-dependent, and strongest near the point-contact.","lead":"This paper measured how the electrical pulse shape of a point-contact germanium detector changes with the position of a particle hit, using two crossed collimated beams and a validated simulation to map the detector's intrinsic spatial position resolution. It then used those simulated pulses to trace where environmental background radiation entered the detector at a nuclear power plant site.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The quantitative resolution map in Fig. 12 rests on unvalidated Siggen model assumptions; only the qualitative anisotropy/non-uniformity is experimentally supported.","rationale":"The reader's weakest assumption—that the Siggen-simulated pulse shapes are faithful enough across the whole detector volume to make the Fig. 12 resolution map equal to the real detector's capability—is the same load-bearing concern I identify. The paper's own admitted impurity-model deviation, the lack of uncertainty propagation, and the absence of any independent position-tagging measurement make the quantitative resolution numbers conditional, while the qualitative findings (fast-bulk sensitivity, Z/R anisotropy, top blind region, energy dependence) are supported by the observed t1-50 agreement and PSCS pulse-shape comparisons at the scanned points. I would therefore not change the reader's CONDITIONAL verdict. I credit the paper for its clear exposition of the PSCS method, the full Geant4/Siggen chain, the explicit treatment of pulse-shape degeneracy as a limitation, and the validation at the five measured positions; those are real supporting elements. But the central quantitative claim remains a simulation-derived number whose controlling inputs have not been independently verified, so the appropriate status is conditional acceptance pending a direct experimental test of the resolution criterion.","tokens_in":15360,"tokens_out":5115,"duration_ms":56038,"concrete_test":"Run an experimental PSCS scan at two additional intersection points in the fast bulk region whose separation matches the predicted resolution (e.g., R27-Z6 vs R27-Z9, and R27-Z9 vs R27-Z12) using the 662 keV collimated source. For each pair, compute the experimental chi2 distribution between PSCS-extracted pulses under real noise and check whether the overlap fraction below chi2_limit is <=5% as the simulation predicts. In parallel, recompute Fig. 12 with a quadratic or position-dependent impurity profile fitted to the existing five PSCS points and with measured baseline-noise PSD instead of ad hoc random noise. If the experimental overlap fraction differs from <=5% by more than statistical uncertainty, or if the recomputed resolution changes by more than 3 mm or reverses the Z/R ordering, then the claimed intrinsic resolution map is not robust to model assumptions.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's central quantitative claim—the intrinsic position-resolution map in Fig. 12—is computed entirely from Siggen-simulated pulses, not from any direct position-tagging measurement. The experimental validation in Sec. IV is limited to five collimated scan points (R=3–27 mm, Z=6–18 mm) and to comparisons of t1-50 histograms plus PSCS-extracted mean pulses; this establishes gross pulse morphology, not the pair-wise distinguishability margins that define resolution in Sec. V. The resolution at each grid point is determined by the distance at which simulated adjacent pulses, each superposed with random noise, become separable at the 5% overlap level (Eqs. 5–6). Every input to that calculation—the impurity concentration model ('linear gradient', Sec. III.B), the field/weighting-potential solution, the carrier-drift model, and the injected noise amplitude—directly scales the chi2 separation, so any model error propagates into the reported mm-level numbers. Sec. IV.B explicitly concedes 'minor deviations between the linear impurity concentration model adopted in the simulation and the actual gradient profile of the crystal.' Because the claimed Z-vs-R anisotropy and the energy dependence are statements about small differences between neighboring simulated pulses, they are precisely the quantities most sensitive to unvalidated model details. In addition, the resolution search is performed only along R and Z; the degeneracy documented in Sec. VI.C (Fig. 15, Fig. 13a) implies that a diagonal neighbor may be closer in pulse-shape space than the nearest axial/radial neighbor, so the map may not represent true full-volume localization capability. The qualitative picture (fast bulk sensitive, top bulk blind, energy dependence) is well supported; the quantitative intrinsic-resolution values are not.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript characterizes the position-dependent pulse-shape response of the CDEX-1B p-type point-contact germanium detector and, for the first time, attempts a quantitative evaluation of its intrinsic single-site spatial position resolution. The experimental method is an optimized pulse-shape comparison scan (PSCS) using two collimated 137Cs beams, with event selection by self-screening and cross-screening. A Geant4+Siggen simulation chain is built and compared with experimental t1-50 distributions and mean pulse shapes at a limited set of collimated positions. The simulated pulses are then used to compute, for each grid point and direction, the distance to the nearest position whose pulse shape can be statistically separated at the 5% overlap criterion, yielding the resolution maps in Fig. 12. The paper also applies the resulting pulse-shape database to reconstruct environmental-background events from the Sanmen measurement, finding a bottom-accumulation pattern consistent with an approximate Geant4 background model, and discusses the R-Z degeneracy in pulse-shape matching.","tokens_in":15732,"tokens_out":3831,"duration_ms":43725,"significance":"If the quantitative resolution map is reliable, this is a useful result: it provides a direct estimate of the ultimate position-discrimination power of pPCGe detectors, guides pulse-shape-database grid construction, and supports background-origin studies for CDEX/RECODE and future ton-scale arrays. The paper has real strengths: the PSCS variant is tailored to the single-channel point-contact geometry; the simulation framework is full-chain (Geant4 transport, Siggen fields, electronics response, noise); and the validation shown in Figs. 9-10 demonstrates that the simulation reproduces the key qualitative features of the measured pulse morphology, including the kink feature and the Z-vs-R anisotropy. The background tracing application is a constructive demonstration of the method. However, the central quantitative claim—the mm-level intrinsic resolution values in Fig. 12—is computed entirely from simulated pulse shapes, with no independent position-tagging validation and no propagation of model uncertainties. The manuscript itself admits deviations between the linear impurity-concentration model and the real crystal (Sec. IV.B). This is the main gap between what is experimentally est","major_comments":[{"comment":"The central quantitative result—the intrinsic spatial position resolution map—is a pure simulation product. The resolution at each point is obtained from the χ2-noise and χ2-ab distributions in Eqs. (5)-(6), using simulated signals with injected noise. Every input to that calculation (impurity-gradient model, field/weighting-potential solution, drift model, noise amplitude, electronics response) directly scales the χ2 separation, so model error propagates into the reported mm values. The experimental validation in Sec. IV is limited to t1-50 histograms and mean pulse shapes at a small set of collimated positions; it tests gross pulse morphology, not the pairwise distinguishability margins that define resolution. The manuscript explicitly states 'minor deviations between the linear impurity concentration model adopted in the simulation and the actual gradient profile of the crystal' (Sec.","section":"Sec. V, Fig. 12"},{"comment":"The quantity called 'intrinsic spatial position resolution' is defined as the nearest distinguishable position along the R or Z axis at fixed other coordinate. But Sec. VI.C and Fig. 15 show a strong R-Z degeneracy in pulse-shape features: pulses at (larger R, smaller Z) can closely resemble pulses at (smaller R, larger Z). Therefore a 1D resolution value along a coordinate axis does not describe the actual localization capability in the 2D detector volume; a position may be confused with an off-axis position lying on the degenerate contour. This should be stated explicitly in Sec. V, and ideally quantified (e.g., the extent of degenerate contours) so that the Fig. 12 map is not interpreted as a true 2D position-resolution map.","section":"Sec. V vs. Sec. VI.C"},{"comment":"The discrimination criterion in Sec. V depends on the simulated noise amplitude and on the normalization of the pulses, but the manuscript does not report how the noise amplitude was calibrated to the real CDEX-1B electronics or how sensitive the 5% overlap threshold is to that calibration. Since Fig. 11 is presented as a schematic, the actual χ2_noise distribution should be compared with the measured pulse-to-pulse variability at a fixed collimated position. This would provide a direct, position-tagged check of the central resolution criterion rather than only of mean pulse shapes. Without this calibration step, the absolute mm values in Fig. 12 remain model-dependent.","section":"Sec. IV.B, Eqs. (5)-(6)"},{"comment":"The position-reconstruction validation uses the same simulation database and the same χ2 matching logic as the resolution map. It demonstrates that the simulated database can reproduce the macroscopic collimation-line morphology, but it is not an independent verification of the mm-level resolution claim: the reconstruction does not have an external position tag with known mm accuracy. The 'upturn' near R<10 mm in Fig. 13(a), attributed to degeneracy, is itself a sign that the effective localization accuracy is position-dependent and can be worse than the local R/Z resolution values. Please discuss how the Fig. 12 resolution values relate to the actual reconstruction accuracy observed in Fig. 13.","section":"Sec. VI.A, Fig. 13"}],"minor_comments":[{"comment":"After Eq. (5), define precisely what is meant by 'normalized' pulse shapes: are the pulses amplitude-normalized before calculating χ2? Is there time alignment? The number of sampling points N and the time window should be specified because they enter the χ2 scale and therefore the 95% threshold.","section":"Eq. (5)"},{"comment":"The caption describes the bar lengths as the resolution; using 'error bars' in the caption is confusing because they are not statistical uncertainties. Rephrase to avoid misinterpretation and add a statement that no systematic uncertainties are shown.","section":"Fig. 12"},{"comment":"There are minor copyediting issues: 'volumn' in Fig. 1, '0νββdecay' missing space, 'ground-breakingly' is informal, and the t1-50 notation appears with inconsistency (e.g., 't1−50%' vs. 't1−50').","section":"General"},{"comment":"The retained-event fraction (1/30 to 1/15) and final 40-60 events are described as optimized but no uncertainty or sensitivity analysis is given for the final extracted pulse shapes. Since these choices affect the mean pulse shapes used in validation, it would be helpful to state their effect on the extracted PSCS responses.","section":"Sec. III.A"}],"recommendation":"major_revision","confidential_remarks":"The paper reports a potentially useful characterization of a pPCGe detector, and the qualitative picture (position-dependent pulse shapes, Z/R anisotropy, energy dependence, blind spots) is well supported. The main issue is that the quantitative resolution map in Fig. 12 is simulation-only and lacks systematic uncertainties or an independent position-tagging check. This is fixable within the manuscript's scope by adding conservative systematic bands, calibrating the noise model against measured pulse variability, and rephrasing the claims so that the simulation-limited character of the absolute numbers is explicit. I would not reject, but the revised version must address these points before the quantitative claims can be accepted. I see no concern about novelty or scope for a detector-physics journal."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Two things you should know about this paper. First, it does something new: it gives the first quantitative intrinsic spatial position resolution map for a p-type point-contact germanium detector, across energies from 10 keV to 2039 keV, and then uses a simulated pulse-shape database to trace real environmental background events back to the bottom of the detector. Second, the quantitative part rests entirely on the simulation. The resolution numbers in Fig. 12 come from Siggen-generated pulses with injected noise, not from any direct position-tagging measurement. So the map is a carefully made model prediction, not a measured capability.\n\nWhat the paper does well: the physics and the method are laid out clearly. The optimized PSCS approach—self-screening then cross-localization—is a sensible adaptation of the existing technique to the point-contact geometry, and the experimental-simulation agreement at the scanned points (Figs. 9 and 10) is genuinely good, especially for the rising edge and the kink feature. The qualitative conclusions—resolution best near the point contact, better along Z than R, energy dependent, essentially blind in the top bulk—are robustly supported by that agreement. The authors also deserve credit for openly discussing the linear impurity gradient model's deviation from the real crystal, and for devoting a whole section to the pulse-shape degeneracy that limits reconstruction. That is honest engagement with their own method's limitations. The background tracing at the Sanmen site is a nice application, and the consistency with their Geant4 background model supports the qualitative picture, though the model is admittedly approximate.\n\nThe soft spots are real but proportionate. The resolution values in Fig. 12 carry no uncertainty bars; the error bars show the resolution magnitude, not its uncertainty. The discrimination criterion (5% overlap of chi-square distributions) is uncalibrated against experiment. The validation is limited to a handful of scan points and compares gross pulse morphology, not the pairwise distinguishability margins that define the resolution metric. Any error in impurity profile, drift model, or injected noise amplitude propagates directly into the mm-level numbers. The degeneracy analysis in Sec. VI.C also implies a diagonal neighbor could be closer in pulse-shape space than the nearest R or Z neighbor, so the map may not capture true full-volume localization capability. These are caveats, not showstoppers: the qualitative backbone holds up.\n\nWho is this for? Detector physicists working on CDEX, LEGEND, or other point-contact germanium programs, and anyone building pulse-shape databases for background rejection. It deserves a serious referee. The main revision should reframe Fig. 12 as a simulation-based expectation, add uncertainty estimates, and make the model dependence explicit in the abstract and conclusions. I would accept this for review and push for those changes.","headline":"Genuinely new quantitative resolution map for pPCGe detectors, but the numbers are simulation-derived model estimates, not measured values; the qualitative picture is solid.","tokens_in":16298,"tokens_out":2728,"would_cite":true,"duration_ms":26740,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["29.40.Wk","95.35.+d","23.40.-s"],"model":"deepseek-v4-flash","headline":"The intrinsic spatial position resolution of a point-contact germanium detector is non-uniform, anisotropic, and energy-dependent, reaching millimeter precision near the electrode while vanishing in the upper bulk.","keywords":["p-type point-contact germanium detector","intrinsic spatial position resolution","pulse shape analysis","background tracing","CDEX-1B","rare-event search","dark matter","neutrinoless double beta decay"],"falsifier":"Place two collimated sources (or two collimator positions) so their beams intersect at a separation equal to the claimed intrinsic resolution at 662 keV and compare the recorded pulse-shape distributions: if their mutual chi-square distribution overlaps the noise-only distribution by more than 5%, the map overstates the resolution. Similarly, a collimated low-energy source scanning radially across the bulk should show no position dependence if the 10 keV map is right.","tokens_in":15282,"feed_emoji":"⚛️","tokens_out":5430,"duration_ms":50541,"temperature":0.7,"pith_summary":"This paper tries to establish that the CDEX-1B p-type point-contact germanium detector has a finite, position-dependent spatial position resolution for single-site events, and that this resolution is good enough to trace real background events back to their origin. The authors claim the resolution is non-uniform and anisotropic: finest in the 'fast bulk' region near the point contact, better along the axial direction than the radial direction, coarser at low energies, and effectively absent in the top bulk region. To reach this conclusion they combine two collimated scans with pulse-shape simulations and a statistical criterion for when two nearby positions are distinguishable. If the claim holds, it gives rare-event experiments a quantitative basis for building pulse-shape databases, subtracting backgrounds, and designing future ton-scale detector arrays.","feed_headline":"Point-contact germanium detector locates events to millimeters","feed_subtitle":"Best near the electrode, axial beats radial, and the top bulk region is a blind spot for position.","key_machinery":"The cross-scanning localization method, a pulse-shape comparison scan adapted to the point-contact geometry: two collimated cesium-137 beams enter from orthogonal directions, and only events whose pulse shapes match across the two samples are taken to come from the geometric intersection of the beams. The resolution criterion is a chi-square comparison between noise-only variations of a pulse at one position and variations between pulses at adjacent positions; the intrinsic resolution in each direction is the minimum distance at which fewer than 5% of adjacent-position comparisons fall below the one-sided 95% noise threshold. The underlying physics this exploits is the weighting-potential ge","core_discovery":"The paper's central discovery is a two-dimensional map of the intrinsic single-site spatial position resolution of the CDEX-1B detector, evaluated at 10, 100, 662, and 2039 keV. 'Intrinsic' here means the resolution set by the physics of signal formation and electronic noise, with experimental smearing from collimator width and event selection removed. The map shows three systematic features: resolution is best near the point contact and degrades outward; axial (Z) resolution is substantially better than radial (R); and the upper bulk has almost no resolving power. The authors also use a non-uniform grid built from this map to reconstruct the positions of environmental-background single-site","pith_inferences":["The resolution framework could be inverted as a design tool: electrode geometry, crystal dimensions, and electronics noise can be varied in simulation to predict which changes improve position resolution where rare-event backgrounds actually concentrate.","An obvious but unstated test is to repeat the collimated scanning at low and high energies (10 keV and 2039 keV) with appropriate sources; the paper's energy dependence is calculated from simulation and currently verified only at 662 keV.","If the same protocol were applied to other p-type point-contact detectors before deployment, their individual impurity profiles and noise characteristics would produce slightly different resolution maps, suggesting per-detector calibration rather than a universal value.","The bottom-edge background accumulation implies that passive or active shielding design for future reactors and ton-scale arrays should target radiation from below; the paper does not draw that engineering conclusion explicitly."],"forward_implications":["Pulse-shape databases can be built on non-uniform grids that match the resolution map: fine near the point contact, coarse in the bulk, rather than wasting grid points where pulses are indistinguishable.","Background events in a real deployment can be localized: the roughly 610-720 keV Compton single-site events traced to the bottom outer ring of the detector, matching an environmental model with radiation entering from below.","Dark-matter search analyses at low energy (about 10 keV) should treat the detector as essentially two-region (bulk vs fast bulk) with no useful radial position information.","Any position estimate from pulse shape must respect the anisotropy: axial information is more reliable than radial, so orientation-aware localization or error assignment is needed.","Rise-time-only matching has a hard limit in the degenerate regions; higher-dimensional features or machine-learning classifiers are needed to approach the physical resolution."],"fun_headline_variants":["Germanium detector maps single-site positions to millimeters","Point-contact Ge detector: axial beats radial in position","CDEX detector resolves event positions, but top bulk is blind","Best near electrode, blind in bulk: Ge detector position map","Intrinsic position resolution of point-contact Ge, mapped"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The resolution map is computed from simulated pulse shapes, and the simulation has been checked against real measurements at only a handful of collimated beam positions; if the crystal's true impurity profile or the electronics noise differs from the model, every resolution number shifts.","fun_headline_variants_meta":{"raw":{"variants":["Germanium detector maps single-site positions to millimeters","Point-contact Ge detector: axial beats radial in position","CDEX detector resolves event positions, but top bulk is blind","Best near electrode, blind in bulk: Ge detector position map","Intrinsic position resolution of point-contact Ge, mapped"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000405,"raw_usage":{"total_tokens":1904,"prompt_tokens":665,"completion_tokens":1239,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":409,"completion_tokens_details":{"reasoning_tokens":1160}},"tokens_in":409,"tokens_out":1239,"duration_ms":8935,"temperature":1.0,"reasoning_tokens":1160,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-02T00:42:59.542868+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Place two collimated sources (or two collimator positions) so their beams intersect at a separation equal to the claimed intrinsic resolution at 662 keV and compare the recorded pulse-shape distributions: if their mutual chi-square distribution overlaps the noise-only distribution by more than 5%, the map overstates the resolution. Similarly, a collimated low-energy source scanning radially across the bulk should show no position dependence if the 10 keV map is right.","supporting_citations":[],"review_version":1}