{"id":"49c794e7-16e4-49b2-af43-149471d91aa3","arxiv_id":"2505.18468","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"A generalized GW-BSE framework with off-diagonal angular momentum terms reproduces the non-monotonic Rydberg series of exciton g-factors in monolayer WSe2.","lead":"The authors present a first-principles method for computing exciton g-factors in magnetic fields, including off-diagonal spin and orbital coupling that previous calculations ignored. They use it to explain the experimentally puzzling non-monotonic g-factors of the Rydberg series in monolayer WSe2.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The paper never runs a diagonal-only control: with identical GW-BSE envelopes and full k-dependent band g-factors, the Rydberg series might still be non-monotonic, so off-diagonal terms are not shown to be necessary.","rationale":"The reader's weakest assumption concerns omitted non-local band g-factor corrections (footnote 61). That is a real quantitative uncertainty, but it is partly mitigated by the authors' cross-check between Quantum ESPRESSO and the all-electron WIEN2k code, which gives nearly identical orbital matrix elements in Table I. The more load-bearing gap is logical: the paper's own central claim asserts that off-diagonal terms are necessary for the non-monotonic Rydberg series, yet no calculation is shown that removes those terms while keeping the full ab initio k-dependent diagonal g-factors. The 2-band parabolic model is not an adequate null hypothesis because its monotonic behavior is dictated by the chosen quadratic form. This gap does not invalidate the generalized formalism or the low-energy validations, but it directly affects the headline interpretation. The verdict should remain conditional, pending the diagonal-only control; hence UNCHANGED relative to the reader's CONDITIONAL verdict.","tokens_in":31994,"tokens_out":6934,"duration_ms":68732,"concrete_test":"Recompute the 1s-4s A-exciton g-factors using Eq. (1) restricted to diagonal elements in both the single-particle basis (only g^z_alpha alpha k) and the exciton basis (only S = S'), keeping the same G0W0-BSE amplitudes, 90x90 k-grid, and scissor-shifted band g-factors. If the 1s-4s sequence remains non-monotonic, the off-diagonal mechanism is not necessary and the central attribution should be weakened; if it becomes monotonic, the concern is dismissed.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim is that off-diagonal spin and orbital angular momentum matrix elements are necessary to reproduce the non-monotonic 1s-4s A-exciton g-factors. The evidence for this is a comparison between the full generalized GW-BSE result and a parabolic two-band model (Fig. 4b). That model is monotonic by construction, since it uses g_X(k) = g0 + g2 k^2 with positive g2. It therefore does not control for the k-dependence of the full diagonal band g-factors. When off-diagonal single-particle and excitonic elements are set to zero, Eq. (1) reduces to Eq. (S12), g_z(S) = sum_{vck} |A^S_{vck}|^2 (g^z_ck - g^z_vk). The A-exciton Rydberg wavefunctions in Fig. S6 sample increasingly large k, so the non-monotonic pattern could arise purely from non-parabolic diagonal band g(k) weighted by these envelopes, with no magnetic hybridization. The paper does not show that omitting the off-diagonal blocks of g^epsilon_SS' and g^epsilon_alpha alpha' k flattens the series. Thus the 'necessary' component of the central claim is not established by the presented calculations.","agreement_with_reader":"disagree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This manuscript develops a generalized first-principles framework for many-body exciton g-factors in arbitrarily oriented magnetic fields, extending earlier GW-BSE treatments by including off-diagonal matrix elements of the spin and orbital angular momentum operators in both the single-particle and exciton bases. The method is applied to monolayer WSe2, where it reproduces the known low-energy exciton Zeeman splittings and the magnetic-field-induced brightening of dark and grey excitons, and it is used to address the long-standing experimental puzzle of the non-monotonic 1s-4s A-exciton g-factor Rydberg series. The authors also report magnetic-field-induced hybridization of higher-energy s- and p-like excitons and support their calculations with a symmetry analysis based on the D3h point group.","tokens_in":32255,"tokens_out":6170,"duration_ms":52325,"significance":"If the central claim is fully established, this work would provide a valuable and general ab initio tool for predicting magneto-optical responses of excitons in van der Waals materials, going beyond existing diagonal-only treatments. The low-energy validation is convincing: the GW-BSE results satisfy the symmetry constraints, reproduce dark/grey exciton brightening mechanisms, and are cross-checked against WIEN2k all-electron calculations. The explicit demonstration that off-diagonal elements appear in the computed g-factor matrices for higher excitons is also an important step. However, the key Rydberg-series conclusion currently rests on a comparison with a simplified two-band model rather than on a controlled diagonal-only limit of the same GW-BSE calculation, so the necessity part of the claim is not yet established. With the additional control calculation requested below, the paper would be a strong contribution to the field.","major_comments":[{"comment":"The central claim that off-diagonal spin/orbital matrix elements are responsible for the non-monotonic 1s-4s A-exciton g-factors is not established, because the only comparison made is between the full generalized GW-BSE result and a parabolic two-band model. When all off-diagonal single-particle and excitonic elements are set to zero, Eq. (1) reduces exactly to SM Eq. (S12), ⟨g_z⟩_S = Σ_{vck} |A^S_{vck}|^2 (g^z_{ck} − g^z_{vk}), which still contains the full k-dependent diagonal band g-factors. The two-band model uses g_X(k) = g_X(0) + g_X(2) k^2 with g_X(2) > 0, so its Rydberg series is monotonic by construction rather than as a consequence of neglecting off-diagonal terms. Since the GW-BSE 1s-4s envelopes in Fig. S6 sample progressively larger k, the non-monotonic pattern in Fig. 4(b) could arise from the non-parabolic k-dependence of the diagonal band g-factors alone. Please compute the diagonal-only limit of Eq. (S12) using the same GW-BSE envelopes and full k-dependent band g-factors; if it is still non-monotonic, the necessity of the off-diagonal terms for the Rydberg claim is not supported, and if it is monotonic, showing this explicitly would strengthen the paper.","section":"Fig. 4(b); SM Eq. (S12)"},{"comment":"The quantitative Rydberg-series prediction relies on orbital angular momentum matrix elements evaluated from DFT wavefunctions with a rigid GW scissor shift applied to the conduction bands, and the authors explicitly neglect non-local contributions to the band g-factors (footnote 61, Refs. [107,108]). The higher 2s-4s exciton envelopes weight larger |k| (Fig. S6), where the scissor approximation and the omitted non-local terms could modify the k-dependence of g^z_{ck} and g^z_{vk} in a state-dependent way. Please quantify this sensitivity, for example by comparing the k-dependent L^z matrix elements against the WIEN2k all-electron results over the k range sampled by the Rydberg envelopes, or by estimating the size of the non-local corrections from Refs. [107,108]; if those corrections are sizable, the predicted non-monotonic pattern could change.","section":"Footnote 61; SM Sec. I"},{"comment":"The experimental data in Fig. 4(a) come from hBN-encapsulated samples, while the GW-BSE calculation in Fig. 4(b) is performed for a vacuum environment; the two-band model is evaluated for both vacuum and hBN, but no GW-BSE calculation including the hBN dielectric environment is presented. Because the spatial extent and envelope k-distribution of Rydberg excitons depend on the dielectric environment, the comparison in support of the central claim is only qualitative. Please either provide a quantitative agreement metric or show explicitly that the dielectric environment does not change the predicted non-monotonic ordering.","section":"Fig. 4(a,b)"}],"minor_comments":[{"comment":"The caption labels both panels as (a); the second panel should be labeled (b).","section":"Fig. 4 caption"},{"comment":"There are several typographical errors, including 'perturbativelly' in the Introduction, 'non-m onotonic' in the title, and the panel labels in the Fig. 3 caption, where '(c) s+, (e) s−, and (e) z' should presumably read '(c) s+, (d) s−, and (e) z'.","section":"Introduction and abstract"},{"comment":"The sentence 'This can only be achieved with in-plane (B ⊥ z) fields' could be misread as applying to all G-D couplings; the D-G brightening under B ∥ z is correctly described elsewhere, so a small clarifying phrase about A-G/A-D mixing would improve readability.","section":"SM Sec. III.B"}],"recommendation":"major_revision","confidential_remarks":"The manuscript makes a valuable methodological advance and the low-energy validation is solid, but the central Rydberg-series claim requires a diagonal-only control calculation before the necessity of off-diagonal terms can be accepted. I would be willing to review a revised version. The authors should also be encouraged to make the diagonal-only dataset available as part of the Supplemental Material."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nThe useful thing to know about this paper is that it adds something the GW-BSE g-factor literature was missing: full off-diagonal spin and orbital angular momentum matrices, in both the single-particle and exciton bases. That is a real generalization of Deilmann et al. and the earlier diagonal-only work. The implementation is careful, and the low-energy validation is solid—Zeeman splitting and dark/grey brightening match known results, and the symmetry analysis is used honestly, including flagging the numerical mixing of degenerate A states. The WIEN2k cross-check of the orbital elements is a good reproducibility practice. I do not see circularity: the g-factors are not fitted to experiment, and the earlier calculations they cite are independent.\n\nThe soft spot is the central claim about the non-monotonic Rydberg series. The paper says the off-diagonal terms 'naturally emerge' and cannot be captured by simplified models. That is true against the two-band parabolic model, but that model is monotonic by construction (positive g2). The stress-test point is fair: setting off-diagonal blocks to zero in Eq. (1) leaves a diagonal sum over the full GW-BSE envelopes with the full k-dependent band g-factors. The 2s–4s envelopes spread far from K, and non-parabolic diagonal g(k) alone could produce the wiggle. The paper does not show that omitting off-diagonal terms flattens the series. So 'necessary' is not established; 'sufficient to reproduce' is. That is a meaningful difference, and the abstract's language overstates it.\n\nTwo lesser issues. The footnote about neglecting non-local band g-factor corrections is honest but unquantified; if those terms are sizable at the k-points that dominate 2s–4s, the pattern could shift. And no code or data is released, which makes it harder to run the missing control. The experimental comparison is qualitative, but the four data sets disagree anyway, so I would not hold that against the mechanism.\n\nBottom line: this is a solid paper with a genuine new formalism and a plausible resolution of a known puzzle. It deserves a serious referee. The referee should ask for the diagonal-only control and a discussion of what happens if non-local g-factor terms are included or estimated. I would send it out, and I would expect revision rather than rejection.\n\nFor my own work, I would cite it—the generalized formalism is useful regardless of the necessity claim.","headline":"Generalized exciton g-factors are a real step forward, but the paper overstates that off-diagonal terms are necessary for the non-monotonic Rydberg series—the diagonal-only control is missing.","tokens_in":32876,"tokens_out":2863,"would_cite":true,"duration_ms":26152,"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":"A generalized GW-BSE formalism including off-diagonal spin and orbital angular momentum matrix elements reproduces the non-monotonic Rydberg series of exciton g-factors in monolayer WSe2.","keywords":["exciton g-factors","GW-Bethe-Salpeter equation","monolayer WSe2","off-diagonal angular momentum","non-monotonic Rydberg series","magnetic brightening","spin-valley physics","magneto-optics"],"falsifier":"Compute the 1s-4s A-exciton g-factors with the same off-diagonal formalism but with a fully self-consistent treatment of the quasiparticle wavefunctions and with non-local band g-factor corrections included, and check whether the non-monotonic dips in the Rydberg series survive with the same positions and depths.","tokens_in":31780,"feed_emoji":"🧲","tokens_out":5451,"duration_ms":47626,"temperature":0.7,"pith_summary":"This paper argues that previous ab initio computations of exciton g-factors missed key physics because they kept only diagonal spin and orbital angular momentum matrix elements. By including off-diagonal elements in both the single-particle and exciton bases, the authors' generalized many-body GW-BSE framework reproduces known magneto-optical results for monolayer WSe2 and yields, without free parameters, the experimentally observed non-monotonic sequence of A-exciton g-factors from 1s to 4s. The same off-diagonal terms are claimed to drive the magnetic brightening of dark and grey excitons and the magnetic hybridization of higher-energy s- and p-like excitons. A sympathetic reader would care because this turns a long-standing experimental puzzle into a calculable many-body effect rather than an unexplained sample-dependent anomaly.","feed_headline":"One missing term explains WSe2's strange exciton g-factors","feed_subtitle":"A many-body calculation reproduces the non-monotonic 1s-4s Rydberg series and magnetic brightening experiments.","key_machinery":"The central object is the generalized exciton g-factor matrix element $g^\\epsilon_{SS'} = \\sum_{vck} (A^S_{vck})^* \\big[\\sum_{c'} A^{S'}_{vc'k} g^\\epsilon_{cc'k} - \\sum_{v'} A^{S'}_{v'ck} g^\\epsilon_{vv'k}\\big]$, where $A^S_{vck}$ is the GW-BSE exciton amplitude and $g^\\epsilon_{\\alpha\\alpha' k} = \\langle \\alpha k | \\hat{L}^\\epsilon + \\hat{\\Sigma}^\\epsilon | \\alpha' k\\rangle$ are single-particle g-factor matrix elements built from orbital and spin angular momentum operators in the Bloch basis. The off-diagonal components of these matrices, absent in previous diagonal-only treatments, allow an external magnetic field to couple different exciton states, including degenerate A-exciton partners and distinct s- and p-like Rydberg states. The orbital part is evaluated through momentum matrix elements and a finite band summation, while symmetry analysis based on the $D_{3h}$ point group supplies the allowed coupling patterns between the bright, grey, and dark exciton subspaces for arbitrary field orientation.","core_discovery":"The central claim is that the non-monotonic Rydberg series of A-exciton g-factors in monolayer WSe2 arises from magnetic-field-induced hybridization captured by off-diagonal matrix elements of the spin and orbital angular momentum operators, both between single-particle bands and between many-body exciton states. Earlier formulations, which retained only diagonal expectation values of these operators, produce a monotonic decreasing trend for the 1s-4s g-factors and cannot reproduce the experimental measurements. In the generalized formalism, the term $\\hat{g}\\cdot B$ with a non-diagonal exciton g-factor matrix naturally mixes exciton subspaces, and the resulting orbital and spin mixing yields non-monotonic g-factors for the A exciton series, with similar non-monotonic structure predicted for the dark/grey Rydberg excitons. The same mechanism explains the brightening of optically inactive dark and grey excitons under out-of-plane, tilted, and in-plane magnetic fields, as well as the field-induced s-p mixing of high-energy excitons.","pith_inferences":["Inference: the quantitative positions and depths of the non-monotonic dips likely depend on the neglected non-local contributions to single-particle band g-factors, which the paper explicitly sets aside; including those corrections could shift the pattern even if the off-diagonal hybridization mechanism is right.","Inference: the same calculation should yield a similarly non-monotonic Rydberg g-factor sequence in closely related monolayers such as MoSe2 and WS2, with material-specific details set by the band structure; this is a testable transfer prediction.","Inference: field-angle-resolved photoluminescence of the 2s and 3s A-exciton states, which is not reported in the experimental papers cited here, would distinguish the hybridization mechanism from environment-induced renormalization of the exciton wavefunctions.","Inference: the magnetic s-p mixing described here and the electric-field s-p mixing observed in monolayer WSe2 likely share the same orbital angular momentum matrix elements, so combined magneto-electro-optical measurements could isolate the orbital part of the response."],"forward_implications":["If the formalism is correct, the non-monotonic 1s-4s g-factor sequence in WSe2 is a genuine magnetic-hybridization effect, so simplified single-band or Wannier models will continue to fail for high-lying Rydberg excitons.","Magnetic brightening of dark and grey excitons under out-of-plane, tilted, and in-plane fields follows from the same off-diagonal angular momentum elements, giving one unified mechanism for several previously separate experimental observations.","High-energy s- and p-like excitons should hybridize under magnetic fields, with nominally dark states brightening as the field direction is tilted; this is shown in the paper for absorption near 1.95 eV in WSe2 and can be tested directly.","Because the framework is built for arbitrary magnetic field directions and arbitrary van der Waals materials, the same machinery can predict g-factors, brightening, and hybridization in heterostructures, chiral excitons, and other two-dimensional systems.","The predicted non-monotonic g-factors for the dark and grey Rydberg exciton series provide a concrete target for future magneto-optical measurements beyond the bright A series."],"supporting_citations":[{"why":"Provides the prior ab initio diagonal-only exciton g-factor approach that this work generalizes by adding off-diagonal matrix elements and degenerate subspaces.","marker":"[56]"},{"why":"Gives the earlier exciton g-factor derivation in the excitonic basis that Eq. (1) extends to off-diagonal terms.","marker":"[57]"},{"why":"Supplies the first-principles framework for exciton g-factors in van der Waals heterostructures used as the starting point for the orbital angular momentum evaluation.","marker":"[96]"},{"why":"Provides the experimental fine structure and dark exciton lifetime data used to benchmark the calculated A-G and D-G splittings and out-of-plane brightening.","marker":"[33]"},{"why":"Reports magnetic brightening of dark excitons in monolayers under in-plane fields, a key validation target for the off-diagonal coupling mechanism.","marker":"[50]"},{"why":"Documents valley coherence and in-plane-field brightening of dark excitons, another experimental benchmark for the generalized g-factor formalism.","marker":"[51]"},{"why":"One of the experimental measurements of the non-monotonic 1s-4s A-exciton g-factor Rydberg series that the paper aims to reproduce.","marker":"[36]"},{"why":"Additional experimental magnetophotoluminescence data for the 1s-4s Rydberg g-factor series used as a comparison target.","marker":"[37]"},{"why":"Provides magneto-photocurrent measurements of Rydberg exciton g-factors, another independent experimental dataset for the non-monotonic series.","marker":"[39]"}],"fun_headline_variants":["Off-diagonal terms crack WSe2 exciton g-factor puzzle","Key term fixes WSe2's non-monotonic exciton g-factors","Magnetic hybridization explains WSe2 Rydberg g-factors","Why WSe2 exciton g-factors bend: missing term found","New theory nails WSe2 dark exciton brightening and g-factors"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The calculation assumes that the orbital contribution to each electron's magnetic response is fully captured by wavefunction-based matrix elements evaluated with a rigid energy-gap correction and a finite band sum, leaving out non-local contributions to the single-particle g-factors that the paper explicitly does not include.","fun_headline_variants_meta":{"raw":{"variants":["Off-diagonal terms crack WSe2 exciton g-factor puzzle","Key term fixes WSe2's non-monotonic exciton g-factors","Magnetic hybridization explains WSe2 Rydberg g-factors","Why WSe2 exciton g-factors bend: missing term found","New theory nails WSe2 dark exciton brightening and g-factors"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000557,"raw_usage":{"total_tokens":2695,"prompt_tokens":1035,"completion_tokens":1660,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":651,"completion_tokens_details":{"reasoning_tokens":1563}},"tokens_in":651,"tokens_out":1660,"duration_ms":11023,"temperature":1.0,"reasoning_tokens":1563,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T14:30:00.339765+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute the 1s-4s A-exciton g-factors with the same off-diagonal formalism but with a fully self-consistent treatment of the quasiparticle wavefunctions and with non-local band g-factor corrections included, and check whether the non-monotonic dips in the Rydberg series survive with the same positions and depths.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the first-principles framework for exciton g-factors in van der Waals heterostructures used as the starting point for the orbital angular momentum evaluation."}],"review_version":1}