{"id":"2d3f8a6e-24d4-44d0-b287-682af618db0a","arxiv_id":"2412.17195","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Traveling-solvent floating-zone growth yields hexagonal boron nitride crystals with Raman and photoluminescence quality comparable to established flux-grown crystals, opening a route to larger bulk samples.","lead":"Researchers grew hexagonal boron nitride crystals with a traveling-solvent floating-zone furnace, producing boules a few millimeters across. The method could make larger and purer h-BN crystals for graphene devices, photonics, and quantum sensing.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The feasibility claim is supported, but the 'on par with CBF' quality claim hinges on the untested assumption that the high stacking-fault PL ratio in TSFZ sample 1 is a polishing artifact rather than intrinsic.","rationale":"The reader's weakest assumption is correct. This is the only quality metric where TSFZ sample 1 differs from CBF by an order of magnitude, and the explanation is purely hypothetical. The paper's own Discussion flags this uncertainty. Since the feasibility claim (h-BN grows via TSFZ, Raman linewidths comparable, low carbon defects) is well supported by microCT, XRD, SEM/EDS, Raman, and PL, the concern does not warrant rejection; it warrants a conditional acceptance pending the polishing-control measurement. The reader's verdict of CONDITIONAL is appropriate, so no change is needed. The proposed test directly settles the ambiguity and would either confirm the artifact hypothesis or require the authors to moderate the 'on par' statement to exclude stacking-fault density.","tokens_in":22595,"tokens_out":3742,"duration_ms":33775,"concrete_test":"Measure PL on the same TSFZ sample (e.g., sample 1 or a replicate grown under identical optimized conditions) on a cleaved or as-grown surface with no polishing, then polish that same surface and remeasure PL under identical conditions. If the 5.48/5.77 eV ratio increases substantially after polishing, the artifact hypothesis is confirmed; if it remains ~20.9, the high stacking-fault density is intrinsic and the 'on par' crystallinity claim should be retracted. A complementary check is to polish a CBF-grown crystal and see whether its ratio rises toward 20.9.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim that TSFZ-grown h-BN is 'on par with, or superior to' CBF crystals rests on the PL stacking-fault ratio. Table 1 reports PL(5.48 eV)/PL(5.77 eV) = 20.9 for the optimized TSFZ sample 1 versus 2.4 and 1.1 for CBF samples. The paper assumes this tenfold excess is an extrinsic effect of cutting and polishing, citing one dendritic flake (sample 8, ratio 2.7) measured as-synthesized and prior work on deformation-induced changes in the excitation spectrum [73]. However, sample 8 was grown under different conditions (B feed, FeB0.2 flux) and is not a bulk piece, so it does not control for growth parameters. The authors explicitly state 'the intrinsic stacking fault density characteristic of the TSFZ technique is currently unclear.' If the high ratio is intrinsic, the TSFZ crystals have an order-of-magnitude higher stacking-fault density than CBF crystals, directly contradicting the 'on par' crystallinity claim in the Conclusions. This is the single most load-bearing assumption: it is untested, and the paper provides no control measurement on the same sample before and after polishing.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports the first demonstration of hexagonal BN (h-BN) crystal growth by the traveling-solvent floating-zone (TSFZ) method, using Fe or FeCr-based fluxes under N2 pressures of 1–50 bar. Boules 3–5 mm in diameter and 2–10 mm long were grown, with microCT showing grains up to roughly 1 mm × 2 mm × 0.5 mm and variable amounts of trapped flux. For the optimized growth (Fe flux, 5–7 bar, 0.1 mm/h), Raman spectroscopy gives an intralayer E2g linewidth of 7.7(2) cm−1 and an interlayer shear-mode linewidth of 1.0(1) cm−1, comparable to values reported for crucible-based flux (CBF) crystals; photoluminescence shows sharp phonon-assisted free-exciton features and strongly reduced carbon-related emission near 3.9–4.1 eV. The authors conclude that TSFZ is a viable route to larger, high-quality h-BN crystals.","tokens_in":22862,"tokens_out":5987,"duration_ms":50640,"significance":"If the feasibility result holds, this is a meaningful advance for bulk h-BN synthesis: TSFZ is crucible-free, offers zone-refining, and potentially enables larger crystals than the typical millimeter-scale CBF plates. The paper's strengths include the use of complementary characterization (microCT, powder and polished-surface XRD, SEM/EDS, Raman, PL), direct benchmarking against CBF crystals grown at Kansas State rather than only literature values, and unusually candid acknowledgment of unresolved issues, including flux inclusions, the pressure/rate confound, and the stacking-fault question. The central feasibility claim—that TSFZ can grow h-BN grains with Raman linewidths comparable to CBF material—is supported by the data. However, the stronger 'on par with, or superior to CBF' claim in the Conclusions rests on an untested assumption about the origin of the enhanced stacking-fault-related PL, and the evidence base for the optimized-growth comparison is a single boule.","major_comments":[{"comment":"The central quality benchmark for the optimized TSFZ material is the PL ratio PL(5.48 eV)/PL(5.77 eV) = 20.9 for sample 1, versus 2.4 and 1.1 for the CBF samples. The manuscript attributes this order-of-magnitude excess to cutting and polishing, citing the unpolished dendritic flake (sample 8, ratio 2.7) and deformation studies [73]. This control is not adequate: sample 8 was grown with a FeB0.2 flux rather than the optimized Fe flux and is a thin flake rather than a sectioned bulk boule, so it does not isolate the effect of polishing from the effect of growth conditions. Because the text itself states that 'the intrinsic stacking fault density characteristic of the TSFZ technique is currently unclear,' the claim in §6 that the optimized procedure yields crystals 'on par with, or superior to' CBF crystals in crystallinity is not yet supported. The authors should either provide a same-sample before/after polishing comparison on the optimized material or explicitly limit the quality claim to the Raman linewidth and carbon-defect PL results.","section":"§5, Table 1, Fig. 6"},{"comment":"The Conclusions state that TSFZ-grown h-BN is 'on par with, or superior to' CBF crystals 'in terms of purity and crystallinity,' but the PL data in §5 show that the TSFZ samples have significantly higher stacking-fault-related emission than the CBF samples, and the Discussion explicitly leaves open whether this is intrinsic. This is in tension with the statement in the same Conclusions that 'the TSFZ samples reported here mostly show higher stacking fault densities than reported CBF samples.' The crystallinity claim should be separated into (i) the intralayer Raman linewidth, which is supported by the data, and (ii) the stacking-fault density, which is unresolved.","section":"§6 Conclusions"},{"comment":"The optimized-growth quality comparison rests on spectroscopy of a single boule (sample 1) for PL, and on a different, unpolished piece of the same boule for Raman. No replicate of the optimized Fe-flux, 5–7 bar, 0.1 mm/h growth is spectroscopically characterized. Given the large run-to-run variation visible in the microCT and Raman data across the other samples, the statement that the 'currently optimized growth procedure produces' crystals of this quality would be materially strengthened by a second optimized-growth sample. I do not consider this a fatal flaw for a feasibility demonstration, but it is part of the evidence needed for the 'on par' claim.","section":"§4 and §5, Table 1"}],"minor_comments":[{"comment":"The abstract and Introduction report the interlayer E2g FWHM as 1.0(1) cm−1, while the Fig. 5a caption reports 1.1(1) cm−1; please reconcile these values.","section":"Abstract and Fig. 5a"},{"comment":"The sentence 'the intralayer mode is substantially broadened compared to samples 1 and 3' appears to be a typo; sample 3 is the sample with the broadened mode, so the comparison should presumably be to samples 1 and 2.","section":"§5 Discussion"},{"comment":"The phrase 'The signal corresponding signal in sample3 shows no resolvable peaks' contains a duplicated word.","section":"§4 Results"},{"comment":"The comparison in Fig. 6b would benefit from a legend or caption explicitly identifying which CBF sample has natural isotope abundance and which is isotopically enriched, since the text refers to both.","section":"Fig. 6b"}],"recommendation":"major_revision","confidential_remarks":"I see no circularity or citation-practice concern: the quality benchmarks are external (literature values and directly measured CBF samples) and the paper contains no fitting that would make the claims self-referential. The main issue is the overstatement in the Conclusions relative to the unresolved stacking-fault density. The feasibility claim itself is solid and the manuscript is close, but the 'on par with, or superior to' crystallinity claim needs either a polishing-control experiment or a carefully softened formulation."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The short version: this is the first TSFZ growth of h-BN, and the feasibility claim holds. The quality claim is one notch stronger than the evidence, but the authors know it and say so.\n\nWhat's new: no one has applied traveling-solvent floating zone to h-BN before. They grew 3-5 mm boules, got grains up to 1x2 mm, and backed it with microCT, XRD, SEM/EDS, Raman, and PL. That's a convincing demonstration. The best sample's Raman linewidths (7.7 cm-1 intralayer, 1.0 cm-1 interlayer) track CBF values, and the carbon-related PL is lower than in the natural-abundance CBF comparison. The parameter mapping (Fe vs FeCr flux, B vs h-BN feed, pressure, geometry) is useful and honestly presented.\n\nThe soft spots: the 'on par with or superior to CBF' claim in the conclusions rests in part on the PL stacking-fault ratio of 20.9 versus 2.4 and 1.1 for CBF. The paper assumes cutting and polishing inflate that ratio, and cites one dendritic flake (sample 8) as a control, but that flake was grown under different conditions, so it doesn't isolate the polishing effect. The authors explicitly say the intrinsic stacking fault density is unclear. That's the load-bearing uncertainty. A referee should ask them to measure an as-synthesized TSFZ surface, or polish a CBF sample and show the ratio jumps. Also, the 50-bar growth confounds pressure and growth rate; they acknowledge it, so it's minor. And only one optimized boule gets the full spectroscopic benchmark; sample-to-sample scatter isn't quantified.\n\nNone of this sinks the paper. The feasibility claim is new, real, and well-supported. The quality parity claim is plausible but unproven at the margin. For the h-BN growth community this is a useful result; it gives a concrete route to larger crystals and a clear set of next experiments. The authors are careful with the literature and flag their own uncertainty. I'd send it to peer review, ideally with a request for the polishing control. I'd cite it if I worked on h-BN synthesis.","headline":"Solid first demonstration of TSFZ growth of h-BN; feasibility holds, but the 'on par with CBF' claim outruns the stacking-fault evidence.","tokens_in":23403,"tokens_out":2953,"would_cite":true,"duration_ms":24421,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["81.10.Fq"],"model":"deepseek-v4-flash","headline":"Traveling-solvent floating zone grows high-quality hexagonal BN crystals.","keywords":["hexagonal boron nitride","traveling-solvent floating zone","bulk crystal growth","iron flux","Raman spectroscopy","photoluminescence","crystal quality","two-dimensional materials"],"falsifier":"Measure the photoluminescence of a TSFZ-grown h-BN crystal that is never cut or polished, using an as-grown or naturally cleaved surface, and compare the intensity ratio of the stacking-fault emission (5.48 eV) to the phonon-assisted free-exciton emission (5.77 eV); a ratio near 20 rather than the CBF values near 1–2 would show the stacking faults are intrinsic to TSFZ growth.","tokens_in":22321,"feed_emoji":"💎","tokens_out":7997,"duration_ms":63748,"temperature":0.7,"pith_summary":"The paper reports that hexagonal BN (h-BN) crystals can be grown by traveling-solvent floating zone (TSFZ), a crucible-free directional solidification method. With an iron flux, moderate nitrogen pressure (about 6 bar), and a slow growth rate of 0.1 mm/h, the resulting boules contain grains up to about 1 mm × 2 mm × 0.5 mm whose Raman linewidths match the best crystals made by the established crucible-based flux (CBF) method. The claim matters because CBF growth yields small, thin plates, while TSFZ is a continuous growth route that could in principle produce larger bulk crystals while also benefiting from zone refining of impurities. The paper also reports that the best TSFZ sample shows sharper phonon-assisted free-exciton photoluminescence and less carbon-related defect emission near 3.9–4.1 eV than a natural-isotope CBF comparison crystal.","feed_headline":"New melt-zone route grows high-quality hexagonal BN crystals","feed_subtitle":"Boules grown from an iron flux rival the best existing crystals in Raman and photoluminescence quality.","key_machinery":"The central object is the traveling-solvent floating zone (TSFZ) itself: a molten iron flux is held between a feed rod and a seed rod without a crucible; the feed rod dissolves into the flux, and h-BN recrystallizes onto the seed as the molten zone is slowly translated. This arrangement combines a steep temperature gradient, continuous directional solidification, and zone refining (impurities stay in the liquid rather than entering the recrystallized solid). The paper also relies on the Sieverts'-law relation between nitrogen pressure and nitrogen solubility in the flux, and on the identity of Raman and PL features (the $E_{2g}$ intralayer and interlayer phonon modes, phonon-assisted free excitons, and the 3.9–4.1 eV carbon-defect band) as metrics of crystal quality.","core_discovery":"On its own terms, the paper's central claim is that TSFZ is a viable route to high-quality h-BN single crystals. The optimized procedure uses an Fe flux, a polycrystalline h-BN feed, moderate $N_2$ pressure ($P \\approx 5$–$7$ bar), and a growth rate of 0.1 mm/h; the grown boules have diameters of 3–5 mm and lengths of 2–10 mm, with the largest crystalline grains reaching roughly 1 mm × 2 mm × 0.5 mm. X-ray diffraction, Raman spectroscopy, and photoluminescence show that the h-BN lattice itself is of high quality: the intralayer $E_{2g}$ mode has an average linewidth of 7.7(2) cm$^{-1}$ at 1365.46(4) cm$^{-1}$ and the interlayer shear mode 1.0(1) cm$^{-1}$ at 51.78(9) cm$^{-1}$, values comparable to CBF-grown crystals, and the PL spectrum has sharp phonon-assisted free-exciton peaks with minimal carbon-defect signal. The paper openly notes that the measured stacking-fault-to-exciton PL ratio (20.9, versus 2.4 and 1.1 for CBF samples) is higher, and that whether this is intrinsic to TSFZ or an artifact of cutting and polishing the samples is not yet settled.","pith_inferences":["If the stacking-fault excess is indeed polishing-induced, the practical payoff is large: TSFZ may already produce material whose intrinsic defect density matches CBF, and the technique's inherent scalability becomes the main remaining challenge.","A direct way to test the paper's open question would be growing at 50 bar with the same 0.1 mm/h rate used at low pressure; the present 50 bar runs were limited to 0.4 mm/h, so pressure and rate effects are conflated.","The occasional columnar, c-axis-oriented growth hints that, with the right seeding and interface control, TSFZ could grow thick h-BN ingots with an ideal shape for exfoliation, a form factor CBF cannot deliver.","Because TSFZ can use either h-BN or boron feed rods, one could isotopically enrich or dope the feed to engineer the crystal's phonon and thermal properties."],"forward_implications":["Because TSFZ is continuous, it offers a path to bulk h-BN crystals larger than the roughly 1 mm by 100-micrometer plates typical of CBF growth.","If the polishing explanation for the elevated stacking-fault PL is correct, polished TSFZ crystals would be expected to show CBF-comparable defect signals when measured on as-grown or cleaved surfaces.","The lower 4.1 eV carbon-defect emission in the TSFZ sample than in a natural-isotope CBF sample suggests TSFZ's crucible-free geometry and zone refining reduce carbon contamination.","Using a boron feed rod in TSFZ demonstrates that nitrogen can be supplied from the gas, which implies isotopic control of the grown h-BN is possible.","The finding that Fe flux outperforms Fe-Cr in TSFZ, opposite to the trend in CBF, means flux chemistry interacts with growth geometry and needs separate optimization."],"supporting_citations":[{"why":"Establishes the crucible-based flux (CBF) method with a high-pressure barium-boron-nitrogen solvent, the benchmark that TSFZ aims to match.","marker":"[38]"},{"why":"Provides the iron-flux CBF crystal whose Raman FWHM (7.6 cm$^{-1}$) and PL ratio (1.1) are the atmospheric-pressure comparison values in Table 1.","marker":"[45]"},{"why":"Documents large-scale Fe-Cr flux growth at atmospheric pressure, the CBF variant whose PL spectra are directly compared with the best TSFZ sample.","marker":"[40]"},{"why":"Supplies the zone-refining principle that underlies the paper's claim that TSFZ reduces extrinsic contamination.","marker":"[52]"},{"why":"Shows that deformation changes the excitation spectrum of h-BN crystals, evidence used to suggest cutting and polishing may inflate the stacking-fault PL signal.","marker":"[73]"},{"why":"Gives Sieverts' law, the relation between N$_2$ pressure and nitrogen solubility in the flux that motivates the moderate-pressure growth strategy.","marker":"[68]"},{"why":"Assigns the $E_{2g}$ intralayer and interlayer phonon modes used as the Raman quality metrics.","marker":"[59]"},{"why":"Identifies the 3.9–4.1 eV emission band with a carbon-related defect, used as the purity metric.","marker":"[64]"}],"fun_headline_variants":["Traveling-solvent zone grows high-quality h-BN","Fe-flux zone melting yields high-quality h-BN crystals","New TSFZ method produces high-purity h-BN boules"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The paper assumes the high stacking-fault photoluminescence of its best sample is caused by cutting and polishing before measurement, not by the TSFZ growth itself; if the stacking faults are intrinsic, the claim that TSFZ matches CBF quality is unsupported.","fun_headline_variants_meta":{"raw":{"variants":["Traveling-solvent zone grows high-quality h-BN","Fe-flux zone melting yields high-quality h-BN crystals","New TSFZ method produces high-purity h-BN boules"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001358,"raw_usage":{"total_tokens":5629,"prompt_tokens":1184,"completion_tokens":4445,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":800,"completion_tokens_details":{"reasoning_tokens":4387}},"tokens_in":800,"tokens_out":4445,"duration_ms":32352,"temperature":1.0,"reasoning_tokens":4387,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T05:42:28.983405+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the photoluminescence of a TSFZ-grown h-BN crystal that is never cut or polished, using an as-grown or naturally cleaved surface, and compare the intensity ratio of the stacking-fault emission (5.48 eV) to the phonon-assisted free-exciton emission (5.77 eV); a ratio near 20 rather than the CBF values near 1–2 would show the stacking faults are intrinsic to TSFZ growth.","supporting_citations":[{"cited_title":"Hexagonal boron nitridecrystalgrowthfromiron,asinglecomponentflux","cited_arxiv_id":null,"evidence_quote":"Provides the iron-flux CBF crystal whose Raman FWHM (7.6 cm$^{-1}$) and PL ratio (1.1) are the atmospheric-pressure comparison values in Table 1."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Documents large-scale Fe-Cr flux growth at atmospheric pressure, the CBF variant whose PL spectra are directly compared with the best TSFZ sample."},{"cited_title":"Techniques of zone melting and crystal growing","cited_arxiv_id":null,"evidence_quote":"Supplies the zone-refining principle that underlies the paper's claim that TSFZ reduces extrinsic contamination."},{"cited_title":"Band-edge luminescence of deformed hexagonal boron nitride single crystals","cited_arxiv_id":null,"evidence_quote":"Shows that deformation changes the excitation spectrum of h-BN crystals, evidence used to suggest cutting and polishing may inflate the stacking-fault PL signal."},{"cited_title":"Near band-gap photoluminescence properties of hexagonal boron nitride","cited_arxiv_id":null,"evidence_quote":"Gives Sieverts' law, the relation between N$_2$ pressure and nitrogen solubility in the flux that motivates the moderate-pressure growth strategy."}],"review_version":1}