{"id":"5c8d31c4-355d-46ac-9d1a-a677310744ad","arxiv_id":"2608.01244","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"DFT+DMFT calculations reveal pressure-driven 5f5/2-to-5f7/2 electron transfer, charge fluctuations, and Fermi-surface reconstruction in UAs2 that track the superconducting dome.","lead":"Using supercomputer simulations, this paper shows that squeezing UAs2 shifts electrons between two types of uranium orbitals, reshaping the Fermi surface exactly where superconductivity appears. The result points to charge fluctuations, not just magnetism, as a possible engine for the record-high superconducting temperature in this class of uranium compounds.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Self-doping direction likely hinges on U/J values imported from uranium oxides; no sensitivity check is provided, so the orbital-transfer mechanism may be parameter-dependent.","rationale":"The reader's weakest assumption identifies exactly the same load-bearing point: the U/J values are taken from uranium oxides and applied to a pnictide without sensitivity analysis. I agree that this is the most critical unresolved issue. The paper's ambient-pressure comparison with ARPES gives independent support for the general electronic structure, and the nonmagnetic-state calculation is appropriate for the high-pressure SC region. However, the entire self-doping picture—the core novelty—rests on the orbital-resolved occupancies, which are highly sensitive to the local interaction parameters in DFT+DMFT. Without a parameter sweep or a computed U/J for UAs2, the central claim remains a hypothesis. The conditional verdict is appropriate: accept the paper provisionally, but require this sensitivity test before the self-doping mechanism is treated as established. My stress-test does not move the verdict; it reinforces the same condition. I considered other potential concerns (interpolated structures, unquantified nesting) but they are secondary: a structural interpolation would influence the band structure, while the U/J choice directly controls the orbital occupation numbers that define self-doping. Thus I set verdict_should_be to UNCHANGED, meaning the reader's conditional verdict stands.","tokens_in":9628,"tokens_out":5425,"duration_ms":51805,"concrete_test":"Repeat the DFT+DMFT calculation at 26.8 GPa (the pressure of maximum Tc) with U=7 eV, J=0.5 eV and with U=9 eV, J=0.7 eV, and also with a rotationally invariant (full Hund) interaction instead of density-density. Track the f5/2 and f7/2 occupancies and their pressure evolution (also recompute at 45 GPa). If the 5f5/2-to-5f7/2 transfer with pressure persists for all these variants, the self-doping mechanism is robust. If it reverses or becomes negligible, the central claim is parameter-dependent, not a robust property of UAs2. A complementary check is to compute U and J from constrained random-phase approximation for UAs2 and compare with the oxide-derived values.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim—pressure-induced self-doping from the 5f5/2 to the 5f7/2 manifold—depends on the relative on-site energies and exchange splittings of the two SOC-split manifolds. The paper sets U=8 eV and J=0.6 eV 'following previous calculations for uranium oxides' (Method section) and applies them to UAs2, a pnictide, without any sensitivity check. The occupancy difference between j=5/2 and j=7/2 is a subtle quantum-chemistry balance: U shifts the average f level, J controls the Hund's coupling that splits different multiplets, and the density-density approximation neglects spin-flip and pair-hopping terms. A different U or J could change the crystal-field and exchange contributions to the f5/2–f7/2 splitting, potentially reducing, eliminating, or even reversing the pressure-driven electron transfer. The paper's Fig. 3(c) shows only the chosen-parameter result (transfer of ~0.06–0.09 e over 26.8–45 GPa), with no error bars or alternative calculations. Since the 'self-doping' is the physical mechanism offered to explain the superconducting dome, this parameter sensitivity is the most load-bearing assumption. The high-pressure structures at 26.8 and 30.8 GPa are also interpolated from volume evolution, with internal coordinates unspecified, but the U/J issue is more direct: it sets the very orbital-resolved occupancy that defines self-doping.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports a DFT+DMFT study of the pressure evolution of the correlated electronic structure of UAs2. At ambient pressure, the calculations reproduce the flat hybridization bands near the Fermi level observed by ARPES. Under pressure, the authors find an orbital-selective charge transfer from the 5f5/2 to the 5f7/2 manifold with the total 5f occupancy nearly unchanged, which they term pressure-induced self-doping. This is accompanied by enhanced charge fluctuations, a Fermi-surface reconstruction from two nested sheets to a corrugated three-dimensional cylinder, and a gradual crossover from non-Fermi-liquid to Fermi-liquid-like self-energy behavior. The authors correlate these changes with the experimentally reported superconducting dome and propose that self-doping, charge fluctuations, and Fermi-surface nesting provide the microscopic electronic-structure basis for superconductivity in UAs2.","tokens_in":10058,"tokens_out":3143,"duration_ms":37178,"significance":"If the central mechanism is correct, the paper offers a concrete and falsifiable electronic-structure picture for the pressure-induced superconducting dome in UAs2, going beyond the conventional magnetic-quantum-critical scenario. The study is technically competent: it uses a standard DFT+DMFT framework with a CT-HYB impurity solver, reproduces the ambient-pressure ARPES flat bands, and makes explicit predictions about orbital-resolved occupancies and Fermi-surface topology that could be tested by high-pressure ARPES or Compton scattering. The claimed self-doping is not fitted to the superconducting dome; it is a raw output of the calculation, and the dome comparison is an after-the-fact observation, which is methodologically appropriate. The main weakness is that the central result depends on Coulomb parameters that are imported from oxide calculations without any sensitivity analysis, and on crystallographic structures at two dome pressures that are interpolated without documented internal coordinates.","major_comments":[{"comment":"The entire self-doping mechanism in Fig. 3(c) rests on the choice U=8 eV, J=0.6 eV, taken 'following previous calculations for uranium oxides.' The occupancy difference between 5f5/2 and 5f7/2 is a delicate balance between the on-site Coulomb repulsion and Hund's coupling, and the density-density approximation omits spin-flip and pair-hopping terms. A different U or J could shift the relative energies of the two SOC-split manifolds and reduce, eliminate, or even reverse the pressure-driven transfer. The paper provides no sensitivity check, no error bars, and no independent estimate for UAs2. This is load-bearing for the main claim, so I ask the authors to repeat the calculation with at least a few physically reasonable (U,J) values, or to provide a constrained-RPA estimate, and show that the direction and approximate magnitude of the self-doping are robust.","section":"Method"},{"comment":"The high-pressure DFT+DMFT calculations at 26.8 and 30.8 GPa use lattice constants 'obtained by interpolation based on the experimental volume evolution,' but the text does not state how the internal coordinates in the Pnma structure were treated. The Fermi-surface reconstruction in Fig. 4(a) and the orbital occupancies in Fig. 3(c) are sensitive to small changes in the As and U internal positions. If the internal coordinates are taken from the 45 GPa structure or linearly interpolated, the 26.8 and 30.8 GPa results may be artifacts. Please specify and, ideally, relax the internal coordinates at each pressure or at least test the sensitivity of the FS topology and occupancies to these coordinates.","section":"Method"},{"comment":"The authors state they 'only focus on the nonmagnetic state for simplicity,' yet the ambient-pressure spectral function is compared directly with ARPES data measured in the antiferromagnetic state (TN≈274 K). The AFM exchange splitting and the associated band folding could modify the orbital-resolved DOS and the exact shape of the flat bands. Since the nonmagnetic calculation is used as the baseline for interpreting the high-pressure results, I ask the authors to clarify whether a magnetic DFT+DMFT calculation was attempted and, if not, to discuss how the AFM state could affect the ambient orbital occupancies and the pressure trend. This is a limitation of the present comparison, even if the high-pressure SC region is expected to be paramagnetic.","section":"Method/Ambient pressure"}],"minor_comments":[{"comment":"Typo: 'useful informations' should be 'useful information'.","section":"Method"},{"comment":"The pressure-dependence plot shows occupancies only at 26.8, 30.8, and 45 GPa. Adding the ambient-pressure values would help quantify the baseline and make the self-doping transfer more transparent.","section":"Fig. 3(c)"},{"comment":"The comparison with ARPES is visually convincing but the energy range of the experimental data is not clear. Indicate the experimental energy and momentum ranges in the figure or caption.","section":"Fig. 2(c)"},{"comment":"The power-law fits to ImΣ(iωn) are described only by 'extracted slopes.' Please specify the frequency window used for the fits and whether the exponent is sensitive to that window.","section":"Fig. 4(b)"},{"comment":"Ref. [9] is cited as a uranium oxide calculation, but it is a UTe2 calculation. The provenance of U=8 eV, J=0.6 eV should be cited more precisely, ideally with a table of the parameters used in Refs. [31,32].","section":"References"}],"recommendation":"major_revision","confidential_remarks":"The paper is within the scope of the journal and addresses a timely experimental finding. The central claim is interesting but rests on a parameter choice that is not justified for UAs2, and on interpolated crystal structures with unspecified internal coordinates. These are fixable with additional calculations, so I do not recommend rejection, but they are load-bearing enough to require a major revision. The nonmagnetic-state assumption is less concerning because the superconducting dome appears after the AFM collapse, but it should be explicitly discussed. The citation pattern is mostly appropriate, with the caveat that the parameter provenance should be clearer."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The paper's real new result is the orbital-selective self-doping under pressure: electrons move from the 5f5/2 to the 5f7/2 manifold while the total 5f count barely changes, and the Fermi surface evolves from nested two-sheet to a corrugated cylinder. That is a concrete, material-specific scenario connecting the superconducting dome to charge fluctuations and topology. I think that is genuinely new and worth taking seriously.\n\nWhat the paper does well: the ambient-pressure DFT+DMFT reproduces the ARPES flat bands around Γ and M, which gives some confidence that the method is working. The multi-band interpretation of the hybridization is a nice correction to earlier speculation. The pressure evolution of the self-energy exponent—from roughly 0.54 to 0.67—is consistent with the experimental NFL-to-FL crossover, and the Fermi-surface nesting argument is visually and conceptually clear.\n\nThe soft spots are real, and the main one is the one flagged in the stress-test: U=8 eV and J=0.6 eV are taken from uranium oxides and applied to a pnictide with no sensitivity check. That matters because the self-doping direction is a subtle balance between the two SOC-split manifolds; a change in U/J could reduce or even reverse the electron transfer. The transfer is only 0.06–0.09 e, so this is not a robust-looking effect. I would not call it fatal, but it is load-bearing, and a referee should demand a U/J scan or at least a justification grounded in the physics of UAs2. The high-pressure structures at 26.8 and 30.8 GPa are interpolated rather than relaxed, so the exact Fermi-surface features at those pressures carry some structural uncertainty. The nonmagnetic state is a deliberate simplification, and the authors say so; that is fine, but it means the connection to the actual AFM-adjacent superconducting state is indirect. No code or data is provided, which makes independent verification harder.\n\nAll that said, the paper is honest and coherent. The central mechanism is a hypothesis, but a good one, with an ambient-pressure benchmark and a clear falsifiable prediction: the self-doping should disappear or reverse if the f5/2–f7/2 splitting is tuned. This deserves a serious referee, not a desk reject.\n\nI would bring this to a reading group concerned with heavy-fermion superconductivity, but I would not yet cite the self-doping mechanism as established in my own work. If I were refereeing, I would ask for sensitivity analysis and structural relaxation before publication.","headline":"A plausible but parameter-sensitive DFT+DMFT case for pressure-induced self-doping in UAs2; worth refereeing, but the U/J dependence needs to be tested before the mechanism is solid.","tokens_in":10444,"tokens_out":1922,"would_cite":false,"duration_ms":21777,"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":"Pressure in UAs2 does not add electrons to uranium; it moves them between the two spin-orbit halves of the 5f shell, and that internal self-doping tracks the superconducting dome.","keywords":["uranium diarsenide","heavy-fermion superconductor","DFT+DMFT","5f electrons","pressure-induced self-doping","Fermi-surface reconstruction","charge fluctuations","mixed valence"],"falsifier":"Repeat the DFT+DMFT calculation at 26.8 and 45 GPa with $U=7$ eV and $U=9$ eV (and $J=0.4$, $0.8$ eV). If the direction or magnitude of the transfer from $j=5/2$ to $j=7/2$ does not persist, the self-doping mechanism collapses. A complementary experimental check is X-ray absorption at the uranium $M_4$/$M_5$ edges under pressure, which should show the $5f$ branching ratio shifting in the direction the calculation predicts if the story is right.","tokens_in":9587,"feed_emoji":"⚛️","tokens_out":8458,"duration_ms":81228,"temperature":0.7,"pith_summary":"UAs2 is a uranium diarsenide antiferromagnet that becomes superconducting under pressure, with a dome peaking near 26.8 GPa. This paper uses DFT+DMFT to argue that pressure does not change the total number of uranium 5f electrons; instead it transfers a fraction of an electron from the more localized $j=5/2$ manifold to the more itinerant $j=7/2$ manifold. That internal \"self-doping\" pushes the $f_{5/2}$ states from a Kondo-localized regime toward mixed valence, enhances charge fluctuations, and reconstructs the Fermi surface. The calculation finds superconductivity in exactly the pressure window where the Fermi surface is made of two disconnected, nested sheets, and its disappearance coincides with their merger into a corrugated three-dimensional cylinder. A sympathetic reader would take the paper as providing the electronic-structure mechanism behind the highest reported $T_c$ among uranium-based $5f$ superconductors.","feed_headline":"Pressure shuffles uranium electrons to tune UAs2 superconductivity","feed_subtitle":"DFT+DMFT traces UAs2's pressure dome to an electron exchange between 5f orbitals and a nested Fermi surface.","key_machinery":"The load-bearing object is the orbital-resolved occupancy of the uranium $5f$ shell, split by spin-orbit coupling into the $j=5/2$ and $j=7/2$ manifolds. DFT+DMFT, with spin-orbit coupling, an interaction $U=8$ eV, Hund's coupling $J=0.6$ eV, and an exact double-counting scheme, lets the calculation track how electrons redistribute between these manifolds as the experimental lattice parameters are compressed. The paper also uses the shape of the Fermi surface—two disconnected nested sheets versus a merged corrugated cylinder—as the structural criterion that correlates with the superconducting dome.","core_discovery":"The central claim is that pressure in UAs2 acts as a self-doping switch: the total U-$5f$ occupancy stays near 2.14, but electrons move from the strongly correlated $j=5/2$ orbitals to the $j=7/2$ orbitals. The paper shows this transfer weakens the Kondo-localized character of the $f_{5/2}$ states, drives the system toward a mixed-valence regime with increased weight in the $N=1$ and $N=3$ atomic configurations, and produces a low-energy kink in the spectral function. Around the superconducting dome the Fermi surface consists of two disconnected quasi-two-dimensional sheets with enhanced nesting; beyond the dome the sheets bend and merge into a corrugated three-dimensional cylinder. The auth","pith_inferences":["A direct X-ray absorption experiment at the uranium $M_4$/$M_5$ edges as a function of pressure could test the predicted self-doping: the $f_{5/2}$/$f_{7/2}$ branching ratio should shift systematically across the 22–31 GPa window.","If the mechanism is generic, isovalent substitutions in the dipnictide family (e.g., replacing As with Sb or P) or epitaxial strain could mimic the pressure-induced orbital transfer and stabilize superconductivity at ambient pressure.","The nonmagnetic calculation leaves the ambient antiferromagnetic order out of the equation; a spin-polarized DFT+DMFT calculation could show whether the same orbital transfer and charge fluctuations survive in the magnetically ordered state and whether they strengthen as N\\'eel order collapses.","The link drawn to hole-doped cuprates and to the second CeCu$_2$Si$_2$ dome suggests a broader pattern: charge-fluctuation-mediated pairing may be a generic route to elevated $T_c$ in $f$-electron metals when the Fermi surface is tuned into a nesting-prone geometry."],"forward_implications":["At ambient pressure the calculated flat Kondo hybridization bands near the $\\Gamma$ and $M$ points match ARPES, supporting a multi-band hybridization scenario rather than a single-band picture.","Pressure transfers roughly 0.03–0.09 electrons per uranium from $j=5/2$ to $j=7/2$ while the total $5f$ count stays constant, so the change is internal self-doping, not external charge doping.","The self-doping drives the $f_{5/2}$ electrons out of the Kondo-localized regime toward mixed valence, and the enhanced $N=1$ and $N=3$ configuration weights indicate stronger charge fluctuations.","Superconductivity occurs only while the Fermi surface has two disconnected nested sheets; its suppression at higher pressure is tied to a Lifshitz-type merger of those sheets into a three-dimensional corrugated surface.","The Matsubara self-energy exponent grows from 0.54 to 0.67 with pressure, tracking the experimental crossover from non-Fermi-liquid to Fermi-liquid-like normal-state behavior."],"supporting_citations":[{"why":"Provides the experimental pressure-temperature phase diagram, the superconducting dome, and the structural transition that the calculation is designed to explain.","marker":"[20]"},{"why":"Supplies the ARPES spectra against which the ambient-pressure DFT+DMFT spectral function is benchmarked.","marker":"[18]"},{"why":"Supplies the experimental lattice parameters for the ambient and high-pressure crystal structures used in the DFT+DMFT calculations.","marker":"[12]"},{"why":"Implements the hybridization-expansion continuous-time quantum Monte Carlo impurity solver used for the U-$5f$ correlated subspace.","marker":"[30]"},{"why":"Provides the full-potential DFT+DMFT implementation and the exact double-counting scheme that fixes the $f$-occupancy reference.","marker":"[28]"},{"why":"Defines the exact double-counting procedure that relates the DFT and DMFT charge densities, a load-bearing element for the absolute $f$ occupancies.","marker":"[33]"},{"why":"Supplies the local Coulomb interaction $U=8$ eV for the actinide $5f$ shell, the parameter on which the pressure-driven self-doping result depends.","marker":"[31]"},{"why":"Supplies the Hund's coupling $J=0.6$ eV for uranium compounds, the other interaction parameter controlling the relative position of the $j=5/2$ and $j=7/2$ manifolds.","marker":"[32]"}],"fun_headline_variants":["Pressure swaps U electrons to spark UAs2 superconductivity","Self-doping reshapes Fermi surface, boosts UAs2's Tc","UAs2: Pressure moves 5f electrons, tunes superconductivity","Pressure-driven orbital shuffle keys UAs2's superconducting dome","Uranium electron swap under pressure drives UAs2's Tc"],"cache_read_input_tokens":2816,"weakest_assumption_plain":"The calculation fixes the local Coulomb repulsion at $U=8$ eV and Hund's coupling at $J=0.6$ eV, values taken from earlier uranium-oxide work, and never checks whether the pressure-driven transfer between the $j=5/2$ and $j=7/2$ manifolds survives a change in these parameters.","fun_headline_variants_meta":{"raw":{"variants":["Pressure swaps U electrons to spark UAs2 superconductivity","Self-doping reshapes Fermi surface, boosts UAs2's Tc","UAs2: Pressure moves 5f electrons, tunes superconductivity","Pressure-driven orbital shuffle keys UAs2's superconducting dome","Uranium electron swap under pressure drives UAs2's Tc"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000173,"raw_usage":{"total_tokens":1155,"prompt_tokens":820,"completion_tokens":335,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":564,"completion_tokens_details":{"reasoning_tokens":246}},"tokens_in":564,"tokens_out":335,"duration_ms":3804,"temperature":1.0,"reasoning_tokens":246,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T00:25:50.927471+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Repeat the DFT+DMFT calculation at 26.8 and 45 GPa with $U=7$ eV and $U=9$ eV (and $J=0.4$, $0.8$ eV). If the direction or magnitude of the transfer from $j=5/2$ to $j=7/2$ does not persist, the self-doping mechanism collapses. A complementary experimental check is X-ray absorption at the uranium $M_4$/$M_5$ edges under pressure, which should show the $5f$ branching ratio shifting in the direction the calculation predicts if the story is right.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the experimental pressure-temperature phase diagram, the superconducting dome, and the structural transition that the calculation is designed to explain."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the ARPES spectra against which the ambient-pressure DFT+DMFT spectral function is benchmarked."},{"cited_title":"Gerward, J","cited_arxiv_id":null,"evidence_quote":"Supplies the experimental lattice parameters for the ambient and high-pressure crystal structures used in the DFT+DMFT calculations."},{"cited_title":"Haule, C.-H","cited_arxiv_id":null,"evidence_quote":"Provides the full-potential DFT+DMFT implementation and the exact double-counting scheme that fixes the $f$-occupancy reference."},{"cited_title":"Haule, Exact double counting in combining the dy- namical mean field theory and the density functional the- ory, Phys","cited_arxiv_id":null,"evidence_quote":"Defines the exact double-counting procedure that relates the DFT and DMFT charge densities, a load-bearing element for the absolute $f$ occupancies."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the local Coulomb interaction $U=8$ eV for the actinide $5f$ shell, the parameter on which the pressure-driven self-doping result depends."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the Hund's coupling $J=0.6$ eV for uranium compounds, the other interaction parameter controlling the relative position of the $j=5/2$ and $j=7/2$ manifolds."}],"review_version":1}