{"id":"8a6c48a2-b24f-471b-b5ea-1d2d64a80b74","arxiv_id":"2507.09553","paper_version":1,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"In electrodeposited Cu1-xBix nanowires, high tartaric acid concentration gives large crystallites with bismuth uniformly in the copper lattice, while low concentration gives small crystallites with bismuth at grain boundaries; heating above 100 °C drives most bismuth out of the lattice, leaving…","lead":"CuBi nanowires were made by electrodeposition, and the amount of tartaric acid in the bath controls both crystal size and where bismuth ends up: inside the copper lattice or at grain boundaries. Heating pushes bismuth out of the lattice, and only a small fraction stays behind, which matters for designing spin-based devices.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The residual 1–2 at.% Bi claim rests on a Vegard calibration built from a two-phase weighted-average volume and possibly nominal SC2/SC4 compositions; a direct post-anneal composition measurement is needed.","rationale":"The paper's qualitative conclusions—TA concentration correlates with crystallite size and Bi distribution, and heating drives Bi out of the Cu lattice—are supported by complementary methods: 4D-STEM orientation maps, EELS maps, variable-temperature SPXRD with sequential Rietveld analysis, and total scattering/PDF. The thermal out-diffusion is corroborated by two independent in situ experiments (slow and fast heating), including the direct observation of increased metallic Bi recrystallization after quenching. I therefore agree with the reader's CONDITIONAL verdict and do not see grounds for rejection. The weakest link is the quantitative residual-doping estimate. The conclusion 'approximately 1–2% Bi remains in the lattice' is not measured directly; it is inferred by comparing post-annealed lattice parameters with a Vegard calibration built from Fig. 5. That calibration has three concrete problems. First, the LC2 point used to anchor the trend is a weighted-average unit-cell volume of two distinct phases (tetragonal Bi-poor, V=47.243 ų; cubic Bi-rich, V=48.086 ų). A weighted average of two-phase volumes is not the lattice parameter of a homogeneous alloy of the average composition, so it cannot validate Vegard's law for the Cu1−xBix solid solution. Second, EELS composition was reported for LC2/LC4/LC7 and for SC7/LC7, but not for SC2/SC4; the SC2/SC4 entries in Fig. 5 therefore appear to be nominal compositions rather than EELS-measured values. If their true Bi contents differ from 2% and 4%, the conclusion that SC-series samples show negligible lattice incorporation is not quantitatively supported. Third, no independent method (e.g., DFT, atom-probe tomography, or homogeneous alloy standards) rules out other contributions to the observed volume increase, such as point defects or residual strain. Any of these issues would change the inferred residual Bi fraction. A less central but noteworthy concern is that the crystallite-size control claim rests on 4D-STEM orientation maps of essentially one nanowire per condition (SC7* and LC7*); although the growth-time difference is consistent, a statistical replication would strengthen this claim. This does not affect the thermal-stability conclusions. The proposed test—direct EELS/EDX of post-annealed nanowires—would settle whether the 1–2 at.% residual is real or an artifact of the calibration. Since the reader already conditioned acceptance on clarifying the composition/lattice-parameter relation, the verdict should remain CONDITIONAL.","tokens_in":21467,"tokens_out":11853,"duration_ms":133637,"concrete_test":"Use STEM-EELS or EDX to measure the Bi concentration in the grain interior of post-annealed SC7 and LC7 nanowires (i.e., after the heating/cooling cycles that yielded a=3.61893 Å and 3.62077 Å), and compare the measured value with the 1–2 at.% inferred from the lattice parameter via the Fig. 5 Vegard calibration. Agreement within the quoted uncertainties would validate the residual-doping claim; disagreement by more than ~1 at.% would require revising it and would also flag the SC2/SC4 nominal-composition entries in Fig. 5.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The most load-bearing quantitative claim is that 'approximately 1–2 at.% Bi remains in the Cu lattice after thermal treatment.' This number is obtained by converting post-annealed lattice parameters (SC7: 3.61893 Å; LC7: 3.62077 Å; pure Cu: 3.61446 Å) into composition using the Fig. 5 Vegard-type volume–composition trend. That trend is not independently validated. (i) The LC-series calibration includes LC2, whose plotted volume is a weighted average of a Bi-poor tetragonal phase (47.243 ų) and a Bi-rich cubic phase (48.086 ų); a weighted-average volume of a heterogeneous two-phase mixture is not a homogeneous alloy lattice parameter, so it cannot establish Vegard behavior for Cu1−xBix. (ii) The SC2 and SC4 points in Fig. 5 appear to use nominal Bi(NO3)3 concentrations rather than EELS-measured Bi content; EELS quantification is reported only for LC2/LC4/LC7 and for SC7/LC7. (iii) No independent check rules out contributions from point defects, strain, or the AAO template to the residual lattice expansion. Because the residual-doping number is a headline conclusion and is used to infer that a small stable Bi fraction persists after heating, the quantitative claim is less secure than the qualitative diffusion evidence, which rests on the directly measured increase in recrystallized metallic Bi in the total-scattering experiment.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript reports a multi-technique study of electrodeposited Cu1−xBix nanowires (x≈0, 0.02, 0.04, 0.07) with nominal small-crystallite (SC) and large-crystallite (LC) series controlled by tartaric acid concentration. The central claims are: (i) TA concentration controls crystallite size and Bi distribution, with Bi at grain boundaries in SC nanowires and homogeneously incorporated into the Cu lattice in LC nanowires; (ii) the LC series shows a Vegard-like lattice expansion with Bi content; (iii) upon heating above ~100 °C Bi diffuses out of the Cu lattice, recrystallizing as metallic Bi on cooling, with approximately 1–2 at.% Bi remaining in the lattice; and (iv) PDF analysis suggests local ordering of Bi at specific interatomic distances. The evidence combines 4D-STEM, EELS/EDS, Rietveld analysis of synchrotron SPXRD, variable-temperature SPXRD, total scattering, and PDF analysis.","tokens_in":21718,"tokens_out":8268,"duration_ms":90428,"significance":"If the results hold, the paper provides a practical synthesis-structure-thermal-budget correlation for CuBi nanowires and identifies a stable residual Bi fraction, which is directly relevant to spin Hall device applications. The study is carefully executed in several respects: Rietveld refinements include instrumental calibration against NIST LaB6 and state uncertainties; EELS composition maps are quantified with reported error bars; 4D-STEM orientation maps directly show crystallite-size control; the in situ VT SPXRD and total-scattering experiments include slow and fast/quench protocols and a control on released nanowires; and the PDF analysis includes both disordered and ordered models. The qualitative picture—TA concentration tunes crystallite size, Bi partitions between lattice and grain boundaries, and Bi begins to leave the lattice above ~100 °C—is well supported by the direct observations. The main weakness is the quantitative residual-doping estimate, which depends on a Vegard calibration that is not independently validated.","major_comments":[{"comment":"The figure caption and the text state that the Bi content in Figure 5 is 'extracted from EELS data', but EELS quantification is reported only for the LC series (LC2, LC4, LC7) and for SC7/LC7; no EELS data are shown for SC2 or SC4. If the SC2 and SC4 points are plotted against nominal Bi(NO3)3 concentrations, the claim that the SC series shows negligible lattice incorporation for those samples is not directly supported by the stated EELS calibration. Please provide EELS-determined compositions for SC2 and SC4, or relabel the horizontal axis and soften the corresponding conclusion.","section":"Composition and Bi distribution – EELS/EDS; Figure 5"},{"comment":"The LC Vegard-type calibration includes LC2 as a point whose volume is the weighted average of a Bi-poor tetragonal phase (47.2429 ų, 66.9 wt%) and a Bi-rich cubic phase (48.086 ų, 33.1 wt%). A weight-averaged volume of a two-phase mixture is not a homogeneous alloy lattice parameter, so it does not independently establish a Vegard relation for Cu1−xBix. Because the post-annealed residual Bi content ('approximately 1–2 at.%' in the Conclusions) is derived by converting the post-annealed lattice parameters through this calibration, the quantitative residual-doping estimate is not secure. Please redo the calibration using only single-phase compositions or provide a sensitivity analysis showing how the residual estimate changes when LC2 is excluded.","section":"Crystal structure – Room temperature SPXRD; Table 2; Figure 5"},{"comment":"The quantitative statement that approximately 1–2 at.% Bi remains in the Cu lattice after heating is an inference from the post-annealed lattice parameters (SC7: 3.61893 Å; LC7: 3.62077 Å) relative to pure Cu (3.61446 Å), not a directly measured composition. No EELS/EDS or refined-occupancy measurement after annealing is reported, and possible contributions from residual strain, point defects, or impurities (e.g., the observed Cr contamination) to the lattice expansion are not excluded. Please add a direct post-anneal composition measurement or explicitly present the 1–2 at.% value as an estimate whose accuracy depends on the Vegard calibration and on the absence of other expansion mechanisms.","section":"Conclusions; Thermal stability – VT SPXRD sections"}],"minor_comments":[{"comment":"In the text describing Figures 1(a,b) and the orientation maps, the sample labels 'SL7*' and the second 'SC7*' should both read 'LC7*'.","section":"Results and discussion – Morphology and microstructure"},{"comment":"The sentence 'the higher Bi-content NWs (LC4 and LC6) consist of a single homogenous cubic phase' mentions LC6, which does not appear in Table 1; this should be LC7.","section":"Crystal structure – Room temperature SPXRD"},{"comment":"The PDF model is labeled 'Cu0.93O0.07' in the text and in the Figure 8 caption; this should be 'Cu0.93Bi0.07'.","section":"Local atomic structure – PDF analysis; Figure 8"},{"comment":"For the ordered PDF model, please report the number of refined parameters and a statistical comparison (e.g., Rwp over the same r-range with the same background treatment) so that the improved fit can be assessed independently of the added degrees of freedom.","section":"Local atomic structure – PDF analysis; Figure 9"},{"comment":"The horizontal-axis error bars for Bi content are not shown even though the text quotes ±1 at.% EELS uncertainty; please add them to the figure.","section":"Figure 5"}],"recommendation":"major_revision","confidential_remarks":null},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Read it. The new part is real: tartaric acid concentration gives a handle on crystallite size in template-grown CuBi nanowires, and that size decides whether Bi sits in the Cu lattice (large grains, LC series) or at grain boundaries (small grains, SC series). The thermal stability story is also new and well-supported: Bi starts leaving the lattice above about 100 °C, migrates seriously between 150–350 °C, and some re-crystallizes as metallic Bi on cooling. The two independent in situ experiments (slow VT-SPXRD and fast TS/quench) agree, and the released-versus-embedded control was the right thing to do.\n\nWhat it does well: the microscopy and scattering are cross-validated; the paper is honest about its own limitations—occupancies cannot be refined, crystallite sizes from SPXRD are not reliable, and the TS data for SC7-B only resolve one Cu phase. The EELS maps give a plausible picture of the inhomogeneous Bi distribution.\n\nThe soft spots are real but not fatal. First, Figure 5 and the text say the Bi content axis comes from EELS, but EELS was performed only on LC2/LC4/LC7 and SC7/LC7. SC2 and SC4 are plotted at nominal concentrations. That is a labeling error; it does not break the qualitative SC-versus-LC conclusion, but it overstates the measurement basis.\n\nSecond, the residual-1–2-at% Bi number. The lattice-parameter-to-composition conversion depends on a Vegard plot that includes LC2 as a weighted-average volume of two cubic phases. A weighted average of two heterogeneous phase volumes is not a homogeneous alloy lattice parameter, so that point is a poor anchor for a Vegard line. The total-scattering mass balance—about 4.8 mol% Bi leaves the lattice, leaving about 2 mol%—is better evidence for the residual-doping claim. The paper would be stronger if that were the primary argument and the Vegard conversion a consistency check.\n\nThird, the PDF local-ordering model: the paper says the ordered model improves the fit but gives no R_w for the two models. Without quantitative comparison, that conclusion is speculative. Minor typo: the Figure 1 caption says 'SL7*' and mixes up SC7*/LC7* once.\n\nWho it is for: people working on CuBi spintronics or on electrodeposited nanowire stability will get value. It deserves a serious referee, with a request for a revision on composition labeling and the Vegard calibration. I would accept it for review.","headline":"Strong multi-technique study showing tartaric-acid controls Bi placement in Cu nanowires and Bi leaves the lattice above 100 °C; the residual-doping estimate is plausible but the Vegard calibration has warts.","tokens_in":22350,"tokens_out":6926,"would_cite":true,"duration_ms":70745,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["81.15.Pq","61.05.cp","72.25.-b"],"model":"deepseek-v4-flash","headline":"The paper shows that varying the tartaric acid concentration in the electrodeposition bath controls both the crystalline domain size and the location of bismuth in CuBi nanowires (in the copper lattice for large crystallites, at grain…","keywords":["CuBi nanowires","electrodeposition","tartaric acid","spin Hall effect","thermal stability","STEM-EELS","Rietveld analysis","pair distribution function"],"falsifier":"Measure the bismuth concentration inside the crystal grains of an annealed nanowire directly—using atom-probe tomography or a cross-sectional EELS line profile that excludes grain boundaries and the oxide shell—and compare it with the 1–2 at.% value implied by the lattice parameter. If the grain-interior bismuth is measurably different, the Vegard-based residual-doping claim fails. A complementary check is to compare the lattice expansion of annealed wires with that of wires containing the same bismuth but deliberately introduced dislocations, separating substitutional bismuth from other strain sources.","tokens_in":21263,"feed_emoji":"🧲","tokens_out":16450,"duration_ms":150981,"temperature":0.7,"pith_summary":"This paper establishes that one synthesis parameter—the concentration of tartaric acid in the electrodeposition bath—controls both the crystalline domain size of copper nanowires and where the bismuth dopant ends up. Low tartaric acid gives wires with ~200 nm crystallites and bismuth piled at grain boundaries; high tartaric acid gives ~1 µm crystallites with bismuth substituted uniformly into the copper lattice, following a Vegard-type lattice expansion. The paper also establishes that this doping is thermally fragile: above roughly 100 °C bismuth starts diffusing out of the lattice, migrates most rapidly between 150 and 350 °C, and recrystallizes as metallic bismuth on cooling, leaving only about 1–2 at.% stably in the lattice even after heating to 1000 °C. This matters because CuBi is a candidate for all-metallic spin-current generation, and these wires will heat up under operating current, so where the dopant sits and how heat moves it determines whether the spin Hall effect survives in a device.","feed_headline":"One bath ingredient decides where bismuth sits in copper nanowires","feed_subtitle":"Tartaric acid tunes crystal size and dopant site; heat past 100 °C drives most bismuth back out.","key_machinery":"The central mechanism is tartaric acid acting as a chelating agent in the electrodeposition bath: it complexes the metal ions, slows the deposition rate (growth time rises from 7500 s to 12000 s), and thereby enlarges the crystalline domains from ~200 nm to ~1 µm. The argument linking structure to composition is Vegard's law as a calibrated volume–composition curve: refined unit cell volumes from Rietveld analysis are plotted against bismuth content measured by EELS, so a lattice expansion signals in-lattice bismuth while a flat volume signals grain-boundary segregation. The thermal argument is carried by sequential Rietveld analysis of variable-temperature diffraction, which tracks the unit cell contraction of the Bi-rich phase as bismuth migrates out, and by total-scattering pair distribution function fits that locate bismuth's preferred interatomic distances in the local structure.","core_discovery":"The core discovery is a two-way coupling between electrodeposition chemistry and the structural state of bismuth in $\\mathrm{Cu}_{1-x}\\mathrm{Bi}_x$ nanowires. Raising the tartaric acid concentration from 0.33 M to 0.99–1.32 M slows deposition enough that crystallites grow from about 200 nm to about 1 µm, and the dopant follows the microstructure: in the large-crystallite wires bismuth is homogeneously incorporated into the copper lattice and the unit cell expands linearly with Bi content, while in the small-crystallite wires most bismuth sits at grain boundaries and the lattice hardly expands. Variable-temperature synchrotron diffraction shows that heat reverses the incorporation: above about 100 °C the Bi-rich phase's unit cell stops its thermal expansion and starts shrinking as bismuth leaves the lattice, with the most rapid migration between 150 and 350 °C; metallic bismuth melts near 250 °C and recrystallizes on cooling. After both slow and fast thermal cycles, including one to 1000 °C, roughly 1–2 at.% bismuth remains in the lattice, as judged from the post-anneal unit cell parameter staying about 0.12% above that of pure copper nanowires. Pair distribution function analysis adds that the incorporated bismuth is not randomly placed: fits improve when bismuth sits at distances near 4.46 Å and 6.81 Å, avoiding nearest and next-nearest neighbor sites.","pith_inferences":["If the ~1–2 at.% residue is a genuine solubility limit, a short anneal near 400 °C could serve as a universal calibration step that fixes the in-lattice bismuth content regardless of electrodeposition kinetics — a consequence the paper does not draw.","The paper does not measure spin transport, but its structural distinction invites a testable prediction: small-crystallite wires, whose bismuth resides at grain boundaries, should retain their spin Hall properties after annealing better than large-crystallite wires, whose lattice bismuth is the population heat removes.","The PDF evidence that bismuth avoids nearest-neighbor sites suggests that theories of the skew-scattering spin Hall effect in CuBi should model isolated, well-separated bismuth scatterers rather than dimers or clusters, an assumption the existing scattering calculations leave open."],"forward_implications":["Device fabrication can select the dopant architecture by selecting tartaric acid concentration: uniform in-lattice bismuth in large-crystallite wires, or grain-boundary bismuth in small-crystallite wires, at the same nominal composition.","Any thermal budget above roughly 100 °C will progressively strip bismuth from the copper lattice, so Joule-heated spintronic devices must either stay below that threshold or be designed around the stable ~1–2 at.% residual doping.","Heating and cooling cycles convert lattice bismuth into metallic bismuth particles, as shown by the growth of the rhombohedral Bi weight fraction, which means annealing offers a route to metal–metal composites inside the wire.","Because the diffusion behavior is identical for template-embedded and released nanowires, the alumina template plays no role in bismuth migration, so the thermal results transfer directly to device geometries.","The residual doping is stable up to 1000 °C, giving device makers a well-defined, reproducible starting state after a single high-temperature anneal."],"supporting_citations":[{"why":"Reports the giant spin Hall angle of about −0.24 in CuBi with ~0.5% Bi doping, the experimental benchmark the nanowire study aims to support.","marker":"[6]"},{"why":"Confirms the spin Hall effect in highly Bi-doped Cu films with interface-free X-ray measurements, grounding the device motivation for the nanowires.","marker":"[7]"},{"why":"Earlier work establishing electrodeposition of highly Bi-doped Cu nanowires that this paper extends to control crystallite size and Bi distribution.","marker":"[13]"},{"why":"Supplies the anodization procedure for the AAO templates that define the nanowire diameter and growth geometry.","marker":"[18]"},{"why":"The 4D-STEM orientation-mapping analysis software from which the ~200 nm versus ~1 µm crystallite sizes are extracted.","marker":"[26]"},{"why":"The reference lattice parameter of bulk, defect-free Cu used to judge how much the Bi-doped wires expand.","marker":"[30]"},{"why":"Vegard's law, the volume–composition relation used to convert refined unit cell volumes into in-lattice Bi content.","marker":"[34]"}],"fun_headline_variants":["Tartaric acid controls bismuth location in copper nanowires","Crystal grain size decides where bismuth hides in Cu–Bi nanowires","Heating drives bismuth out of copper lattice in nanowires","Synthesis additive sets bismuth's final seat in Cu nanowires"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"Everything quantitative about the residual doping rests on translating refined unit cell volumes into bismuth content through a Vegard-law calibration anchored to electron-microscopy compositions and not independently validated; if other lattice defects contribute to the measured expansion, or the composition assignments are off, the 1–2 at.% residue and the claim of negligible in-lattice bismuth in the small-crystallite samples would be weakened.","fun_headline_variants_meta":{"raw":{"variants":["Tartaric acid controls bismuth location in copper nanowires","Crystal grain size decides where bismuth hides in Cu–Bi nanowires","Heating drives bismuth out of copper lattice in nanowires","Synthesis additive sets bismuth's final seat in Cu nanowires"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000271,"raw_usage":{"total_tokens":1709,"prompt_tokens":1103,"completion_tokens":606,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":719,"completion_tokens_details":{"reasoning_tokens":527}},"tokens_in":719,"tokens_out":606,"duration_ms":7403,"temperature":1.0,"reasoning_tokens":527,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T17:52:57.283567+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the bismuth concentration inside the crystal grains of an annealed nanowire directly—using atom-probe tomography or a cross-sectional EELS line profile that excludes grain boundaries and the oxide shell—and compare it with the 1–2 at.% value implied by the lattice parameter. If the grain-interior bismuth is measurably different, the Vegard-based residual-doping claim fails. A complementary check is to compare the lattice expansion of annealed wires with that of wires containing the same bismuth but deliberately introduced dislocations, separating substitutional bismuth from other strain sources.","supporting_citations":[],"review_version":1}