{"id":"99b72f6e-572d-44de-8269-d1d58f037652","arxiv_id":"2505.07091","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"At room temperature, the C2 filled ice converts to the hydrogen-richer C3 phase between 47 and 104 GPa, and the C3 structure is solved at 69 GPa as cubic Fd-3m with doubled H2 content.","lead":"Hydrogen-water mixtures squeezed above 47 GPa slowly transform into a hydrogen-rich filled ice called C3, which holds twice as much H2 as the earlier C2 phase. The paper determines the crystal structure of C3 at 69 GPa and maps its stability up to 104 GPa, with implications for icy planets and hydrogen storage.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"C3 composition claim rests on a fixed, unrefined 16c H2 occupancy in a 41-reflection refinement; 'XRD determines (H2O)(H2)2' needs a free-occupancy check.","rationale":"The paper is credible and the SC XRD structure solution is a real advance. My concern is narrower: the abstract's 'determines' for composition goes beyond the reported refinement, where the H2 occupancy appears to be imposed at 2 rather than refined. With 41 independent reflections and H as a weak scatterer, occupancy and displacement are strongly correlated, so the refinement alone cannot bear the full weight of the doubled-H2 claim. The volume data, theory, and prior work by Ranieri et al. all point in the same direction, so this is a condition to verify, not a reason to reject. The reader's alternative concern about phase identification at 104 GPa is real but secondary: if the 69 GPa composition is verified, the high-pressure identification by peak positions, volume, and Raman convergence remains reasonable. Hence the verdict remains CONDITIONAL, with no change needed to the reader's assessment.","tokens_in":18241,"tokens_out":8722,"duration_ms":90842,"concrete_test":"Re-refine the deposited CSD2419902 69 GPa data (or the raw CrysAlisPro frames) with the 16c H-site occupancy free to vary, using the same twin model and restraints, and also test an alternative model with H2 on 8b (partial or full). Compare R1/wR2 and the refined occupancy with its esd, and inspect the difference Fourier map at 16c. If the free occupancy is within about 2σ of 2 and the 8b model fits significantly worse, the composition claim stands. If not, the paper should downgrade 'determines' to 'consistent with' and the verdict should remain conditional.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central novelty is the doubled H2 content of C3, (H2O)(H2)2, stated as 'determined' by single-crystal XRD at 69 GPa (Table 1). In that refinement, H2 was modelled as a single H atom on 16c with occupancy fixed at 2 ('Due to rotational disorder, the hydrogen molecule was approximated as a single hydrogen atom (H2) with an occupancy of 2'). No refined occupancy or esd is reported, and the independent-reflection count is only 41 with a data/parameter ratio of about 8. Hydrogen is a weak X-ray scatterer, so a fixed 16c occupancy cannot by itself establish the composition; the volume/EOS agreement with the P41 C3 phase is supporting evidence, but that agreement is not equivalent to a diffraction determination. The 103–104 GPa cold-compressed phase identification inherits this same model because it is made by matching Fd-3m peak positions and extrapolated volumes to the 69 GPa C3 structure. Thus the most load-bearing weak point is the occupancy assumption, not the high-pressure extrapolation per se.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript reports a combined synchrotron single-crystal X-ray diffraction, Raman spectroscopy, and first-principles DFT study of the hydrogen hydrate system (H2O)(H2)n up to 104 GPa. The authors find that the C2 filled ice transforms to the C3 filled ice at room temperature gradually over 47–104 GPa in hydrogen-rich conditions, and that laser heating induces the transformation at lower pressures. At 69 GPa, a single-crystal structure solution assigns the C3 phase to cubic Fd3m, with H2 molecules in the 16c positions and a nominal composition (H2O)(H2)2, twice the H2 content of C2. The paper also reports Raman vibron softening across the transition, theoretical stability calculations placing C2 stable to 14 GPa and C3 stable above 23 GPa, and band-gap calculations for both phases.","tokens_in":18474,"tokens_out":5629,"duration_ms":52903,"significance":"If the structural assignment and composition hold, this is a significant result: it provides the first direct crystallographic evidence for the structure of the most hydrogen-dense filled ice and updates the pressure-composition stability landscape relevant to giant-planet interiors and hydrogen storage. The SC XRD work is technically impressive: 41 independent reflections, a complete structure solution with R1 = 4.06%, twinning handled by simultaneous integration, and the structure deposited in the CSD. The DFT calculations are largely independent of the experimental identification and support the phase stability picture; the hybrid-functional band-gap results are not load-bearing. The main caveat, discussed below, is that the doubled H2 content relies on a fixed, unrefined site occupancy.","major_comments":[{"comment":"The central compositional claim, (H2O)(H2)2, is stated as determined by single-crystal XRD, but the H2 molecule is modeled as a single H atom on the 16c site with the occupancy fixed at 2, and no free refinement or estimated standard deviation for this occupancy is reported. Because hydrogen is a weak X-ray scatterer and the refinement uses only 41 independent reflections, the diffraction data alone do not independently establish the full 16c occupancy. The agreement of the measured volume per formula unit with the theoretical P41 C3 volume is supporting evidence, but it is not equivalent to a diffraction determination. The authors should either report a free refinement of the 16c occupancy (with its esd) or explicitly state that the composition is inferred from the site assignment plus volume/Raman evidence rather than determined by refinement.","section":"Table 1 and Section III (SC XRD refinement at 69 GPa)"},{"comment":"The identification of the cold-compressed phase at 103–104 GPa as the C3 phase is based on matching Fd3m peak positions, extrapolated unit-cell volumes, and Raman vibron convergence with the 69 GPa laser-heated structure, not on a full structure solution at the highest pressure. The text should state this explicitly and discuss the possibility that the cold-compressed material may be compositionally intermediate, since the room-temperature transformation is described as gradual and the H2 uptake may be incomplete. This is load-bearing for the abstract claim that the C2-to-C3 transformation occurs at room temperature up to 104 GPa.","section":"Section III, cold-compression results at 103–104 GPa"}],"minor_comments":[{"comment":"Please correct 'spectradiometric' to 'spectroradiometric' and remove the doubled 'by' in 'by via absorption'.","section":"Section II.A (Methods)"},{"comment":"Pressure, volume, and Raman frequency data are presented without error bars or quantitative uncertainties; adding error estimates from the Au/ruby pressure markers and peak fits would strengthen the comparison with theory.","section":"Figures 5 and 7"},{"comment":"The maximum temperature of approximately 1500 K is stated as a visual estimate; since the lower-pressure laser-heating pathway to C3 depends on this temperature, the authors should clearly mark it as approximate and give an uncertainty range.","section":"Section II.A (Laser heating)"},{"comment":"The Supplemental Material reference contains 'Supplemental Figures S1-S10' twice in a single sentence; the duplicate should be removed.","section":"Reference 12"},{"comment":"The caption lists an excitation wavelength of 630 nm, while the text reports 660 nm; please make these consistent.","section":"Figure 6 caption"}],"recommendation":"major_revision","confidential_remarks":"The paper is within the journal's scope and the experimental dataset is substantial. The main concern is the fixed H2 occupancy in the SC XRD refinement; if the authors can provide a free-occupancy refinement or clearly reframe the composition claim as an inference, the paper would be publishable. I do not see any indication of circular reasoning between the experiment and DFT, and the self-citations are limited to methods and prior structural work."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nRead this one. The genuinely new result is the room-temperature C2-to-C3 conversion under H2-rich conditions over 47–104 GPa, plus the first single-crystal structure solution of C3 at 69 GPa with hydrogen positions. The structure refinement is credible: cubic Fd-3m, R1 = 4.06%, 41 independent reflections, twin handling, and agreement with Raman and volume data. That is real progress over the earlier powder work.\n\nThe central composition claim, (H2O)(H2)2, is the place to focus. The H2 molecule is modeled as a single atom on 16c with occupancy fixed at 2. That occupancy is not refined, and with hydrogen being a weak X-ray scatterer and only 41 reflections, the diffraction data alone do not determine the doubled H2 content. The supporting evidence—volume agreement with the predicted P41 C3 phase and the observed lattice expansion—is meaningful but not equivalent to a diffraction-based composition measurement. The stress-test note has this right. The 104 GPa cold-compressed phase identification inherits the same assumption, since it is based on peak matching and volume extrapolation rather than a full structure solution at that pressure.\n\nOther soft spots are minor: no error bars on pressure or volume, laser-heating temperatures estimated visually rather than measured, and the hybrid BLYP band-gap calculations use empirically fitted mixing parameters. None of these affect the core structural assignment, which is independent of the DFT.\n\nWho gets value: high-pressure experimentalists and planetary-interior modelers. The paper deserves a serious referee, but the referee should ask for a free refinement of the H2 occupancy (or at least an explicit discussion of why it cannot be refined) before the composition is stated as determined. The cold-compression and laser-heated samples may also deserve separate treatment.\n\nRecommendation: send to peer review with a request for that occupancy check. If the occupancy holds, this is a clean contribution to the hydrate phase diagram.\n\nBest,\n[Your name]","headline":"A solid experimental determination of the C3 filled-ice structure and its room-temperature formation path, with the composition resting on a fixed H2 occupancy that deserves a free-refinement check.","tokens_in":19020,"tokens_out":535,"would_cite":true,"duration_ms":6839,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"A water–hydrogen ice doubles its molecular H2 content at high pressure.","keywords":["hydrogen hydrate","filled ice","high pressure","single-crystal X-ray diffraction","C3 phase","Fd-3m structure","hydrogen storage","planetary interiors"],"falsifier":"Measure the single-crystal structure of the cold-compressed sample near 100 GPa and refine the hydrogen positions directly; if the H2 molecules do not sit at 16c with the (H2O)(H2)2 composition, or if the C2 reflections persist at higher pressures than reported, the central claim fails.","tokens_in":18074,"feed_emoji":"🧊","tokens_out":6439,"duration_ms":58052,"temperature":0.7,"pith_summary":"This paper establishes that the hydrogen-filled ice C2 transforms into a hydrogen-richer phase, C3, at room temperature across 47–104 GPa, and that C3 contains twice as much molecular hydrogen as C2. At 69 GPa, synchrotron single-crystal X-ray diffraction solves the C3 structure as cubic, space group Fd-3m, with H2 molecules occupying the 16c sites, giving the composition (H2O)(H2)2. The paper also shows that heating C2 above roughly 1500 K drives the same transition at pressures as low as 47 GPa, and that C3 remains metastable on decompression to 40 GPa. These results matter because they set new stability and composition limits for hydrates, with direct consequences for hydrogen storage and for models of the interiors of Neptune-like planets.","feed_headline":"Hydrogen-rich ice packs twice the H2 by 104 GPa","feed_subtitle":"Single-crystal X-rays identify the C3 phase as (H2O)(H2)2, a recipe for ice planets and hydrogen storage.","key_machinery":"The load-bearing identity is the crystallographic site switch between the two filled ices: in the cubic space group Fd-3m, C2 places molecular hydrogen on the 8b sites while C3 places it on the 16c sites, which are twice as numerous, so occupancy of the same water network doubles the H2 content to (H2O)(H2)2. The argument is carried by single-crystal X-ray diffraction on multiply twinned crystals, which returned 41 independent reflections and structure solution from difference Fourier maps (R1 = 4.06%), together with Raman measurements of the H2 intramolecular vibron, whose downward shift marks the denser hydrogen packing. First-principles calculations using hydrogen-ordered Pna21 and P41 proxy structures supply the enthalpy ordering and the theoretical volume and vibron curves that the experiment is compared against. The cubic experimental C3 structure is interpreted as a disordered average of the lower-symmetry tetragonal ordered form, analogous to ice VII versus ice VIII.","core_discovery":"At 69 GPa, laser heating of the C2 filled ice produced single crystals of a new phase whose structure the paper solves completely by single-crystal X-ray diffraction. The C3 phase is cubic, space group Fd-3m, with oxygen atoms on the 8b sites of the same hydrogen-bonded water network as C2; the difference is where molecular hydrogen sits. In C2 the H2 molecules occupy the 8b positions (one H2 per formula unit), while in C3 they occupy the 16c positions, twice as many sites, refined as a single scatterer of occupancy 2, yielding (H2O)(H2)2. Refinement gives R1 = 4.06% from 41 independent reflections, with hydrogen positions taken directly from difference Fourier maps. The paper further claims that cold compression under hydrogen-rich conditions converts C2 into the same C3 phase gradually over 47–104 GPa, with the C2 diffraction peaks disappearing by 103 GPa, and that laser heating above about 1500 K makes the transition abrupt at pressures as low as 47 GPa. Density functional calculations reproduce the volume relations and the softening of the H2 vibron, and place thermodynamic stability of C3 above about 23 GPa.","pith_inferences":["If C3 is representative of deep ice mantles, planetary models may need to include a hydrogen reservoir that is twice as concentrated as previously assumed at the same depths; the paper's 104 GPa stability supports that possibility but does not itself model planetary interiors.","The cubic Fd-3m symmetry may be a time and space average of ordered tetragonal domains; low-temperature experiments on the same composition could reveal an ordered analog analogous to ice VIII.","The gradual pressure-driven uptake suggests that intermediate pressures below 47 GPa might also convert C2 to C3 given longer times or higher H2 fugacity, a testable kinetic prediction.","Comparing the 1.94 Å H2–H2 distance at 100 GPa with bulk hydrogen suggests the hydrate cavities keep H2 near the onset of metallization physics, and the saturating band-gap trend could be probed by optical conductivity measurements."],"forward_implications":["C3 doubles the molecular-hydrogen capacity of the filled ice relative to C2, so hydrogen storage in hydrate form can hold more H2 at a given pressure.","The C2-to-C3 transformation at 300 K proceeds slowly and only reaches completion after weeks at 103 GPa, implying a kinetically limited reaction that can be completed by laser heating above about 1500 K.","Once formed, C3 survives decompression to 40 GPa, which defines a practical window for recovering hydrogen-rich hydrate to moderate pressures.","Hydrogen-bond symmetrization in the water sublattice is delayed from about 32–40 GPa in C2 to about 78–87 GPa in C3, showing that extra H2 stiffens the water framework against symmetrization.","The band gap of C3 narrows rapidly with pressure up to about 100 GPa and then levels off, keeping the hydrate insulating at the highest pressures studied."],"supporting_citations":[{"why":"Establishes the C2 filled ice structure, Fd-3m with H2 in the 8b sites, which this paper's C2 data build on.","marker":"[6]"},{"why":"Documents structural changes of filled ice up to 50 GPa, anchoring the prior stability range of C2.","marker":"[7]"},{"why":"Predicts the C3 structure and stability; provides the hydrogen-ordered Pna21 and P41 models used for comparison.","marker":"[10]"},{"why":"Reports the first observation of the most H2-dense filled ice and the C2–C3 transition, which this paper extends and structurally solves.","marker":"[11]"},{"why":"Provides the equation of state of bulk hydrogen used to compare volumes of the hydrate versus separate ice plus H2.","marker":"[31]"},{"why":"Supplies the Raman signature and pressure scale for hydrogen-bond symmetrization used to interpret C2 and C3.","marker":"[32]"}],"fun_headline_variants":["C3 hydrate phase stores twice the H2","High-pressure ice packs double the hydrogen","New filled ice phase doubles H2 content","Hydrogen-rich hydrate gets twice as dense at 104 GPa","C3 ice phase: two H2 per water at high pressure"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The identification of the cold-compressed material at 103–104 GPa with the laser-heated C3 structure solved at 69 GPa rests on matching Fd-3m diffraction peak positions, volume extrapolation, and vibron convergence rather than on a full structure solution at the highest pressure, and the laser-heating temperatures were estimated visually rather than measured.","fun_headline_variants_meta":{"raw":{"variants":["C3 hydrate phase stores twice the H2","High-pressure ice packs double the hydrogen","New filled ice phase doubles H2 content","Hydrogen-rich hydrate gets twice as dense at 104 GPa","C3 ice phase: two H2 per water at high pressure"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000192,"raw_usage":{"total_tokens":1382,"prompt_tokens":1019,"completion_tokens":363,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":635,"completion_tokens_details":{"reasoning_tokens":287}},"tokens_in":635,"tokens_out":363,"duration_ms":4641,"temperature":1.0,"reasoning_tokens":287,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T22:25:12.860074+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the single-crystal structure of the cold-compressed sample near 100 GPa and refine the hydrogen positions directly; if the H2 molecules do not sit at 16c with the (H2O)(H2)2 composition, or if the C2 reflections persist at higher pressures than reported, the central claim fails.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Establishes the C2 filled ice structure, Fd-3m with H2 in the 8b sites, which this paper's C2 data build on."},{"cited_title":"Mao, H.-k","cited_arxiv_id":null,"evidence_quote":"Documents structural changes of filled ice up to 50 GPa, anchoring the prior stability range of C2."},{"cited_title":"V os, L.W","cited_arxiv_id":null,"evidence_quote":"Predicts the C3 structure and stability; provides the hydrogen-ordered Pna21 and P41 models used for comparison."},{"cited_title":"Hirai, S","cited_arxiv_id":null,"evidence_quote":"Reports the first observation of the most H2-dense filled ice and the C2–C3 transition, which this paper extends and structurally solves."},{"cited_title":"Sheldrick, SHELXT - Integrated space-group and crystal-structure determination, Acta Crystallographica Section A, 71 (2015) 3-8","cited_arxiv_id":null,"evidence_quote":"Provides the equation of state of bulk hydrogen used to compare volumes of the hydrate versus separate ice plus H2."},{"cited_title":"Dolomanov, L.J","cited_arxiv_id":null,"evidence_quote":"Supplies the Raman signature and pressure scale for hydrogen-bond symmetrization used to interpret C2 and C3."}],"review_version":1}