{"id":"d0107b91-bdaf-4994-a6b4-441031ae9b13","arxiv_id":"2501.10650","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":5,"one_line_summary":"Chiral phonons in ferrimagnetic Fe1.75Zn0.25Mo3O8 split by about 20% at zero field and can be non-volatilely switched by reversing the magnetization.","lead":"This paper shows that paired circular vibrations inside a magnetic crystal can be flipped by a moderate magnetic field, and the flip remains after the field is off. The result suggests that vibrations, and the angular momentum they carry, could be controlled by magnetic domains in future devices.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The inference that Raman LR/RL channel reversal proves phonon angular momentum reversal rests on selection-rule arguments without a direct eigenvector measurement; a magnetization-dependent Raman tensor could mimic the swap.","rationale":"The paper is a careful experimental study with multiple consistent observables: the P1 splitting tracks magnetization, the hysteresis is synchronized with RMCD and magnetization, and the Stokes/anti-Stokes rules are reported as opposite. The most load-bearing assumption in the central claim is not that the modes are phonons—persistence to 300 K and E2 symmetry make this highly likely—but that the LR/RL channel exchange upon magnetization reversal is a direct readout of the phonon angular momentum reversal. This assumes the Raman intensity asymmetry is dominated by the phonon eigenvector's circularity; however, magnetic crystals can exhibit morphic Raman tensors with antisymmetric components linear in M that would also swap LR/RL intensities between oppositely magnetized domains even for linearly polarized phonons. The paper does not provide a direct eigenvector measurement or a calculation disentangling these contributions. While the reported opposite Stokes/anti-Stokes selection rules argue in favor of genuine PAM, the data are not shown quantitatively, leaving a residual risk. A spin-polarized DFT+U phonon calculation or a polarization-resolved inelastic x-ray scattering experiment would settle this. The reader's weakest assumption overlaps with this concern, but I would weight the PAM inference more heavily than the phonon-origin identification. Since this concern is the same one that motivated the conditional verdict, I recommend no change to the reader's verdict.","tokens_in":10552,"tokens_out":11148,"duration_ms":118083,"concrete_test":"Perform spin-polarized DFT+U lattice-dynamics calculations for a Zn-substituted supercell of Fe2Mo3O8 in the ferrimagnetic state. Compute the phonon eigenvectors of the P1 modes at the experimental q (2k_inc along c) and evaluate the phonon angular momentum projection (atomic circular motion). Then repeat with the magnetization reversed. If the eigenvectors are not near-circular, or if their handedness does not flip with magnetization, the Raman LR/RL exchange cannot be attributed to PAM reversal. Independently, fit the Raman spectra of Fig. 2(c) with a magneto-optical Raman-tensor model to test whether the intensity asymmetry could arise without eigenvector circularity.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim that reversing the ferrimagnetic magnetization reverses the phonon angular momentum (PAM) of the P1a/P1b modes is inferred from the exchange of the LR and RL Raman channels in Fig. 3(a). This inference requires that the LR/RL intensity asymmetry is governed by the phonon's circular eigenvector, not by a magnetization-dependent complex Raman tensor (morphic/antisymmetric terms). In magnetic crystals, Raman tensors can acquire antisymmetric components proportional to M, which can produce circular intensity asymmetries for even linearly polarized phonons, and these terms change sign with M. The paper's evidence for genuine chiral phonons consists of (i) continuity with the parent compound Fe2Mo3O8, (ii) persistence of the modes at 300 K, and (iii) opposite Stokes/anti-Stokes selection rules; but no direct measurement of the phonon eigenvectors or of the atomic circular motion is reported. If the observed LR/RL swap were dominated by a morphic Raman tensor rather than by the phonon eigenvector, the conclusion of PAM switching would not follow.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript reports polarization-resolved magneto-Raman experiments on the ferrimagnetic polar insulator Fe1.75Zn0.25Mo3O8. The authors identify two low-energy modes (P1a at ~5.21 meV and P1b at ~6.32 meV) as a chiral phonon pair with opposite circular polarization selection rules, a zero-field splitting that reaches ~20% of the phonon frequency, and asymmetric Zeeman shifts in a magnetic field. A peak effective phonon magnetic moment of 2.62 Bohr magnetons is reported near the Neel temperature. The phonon modes follow the ferrimagnetic hysteresis, and the LR/RL Raman channels exchange when the magnetization is reversed, which the authors interpret as switching of phonon pseudo-angular momentum. Raman mapping combined with magneto-optical imaging reveals phononic domains locked to ferrimagnetic domains. A phonon-magnon coupling model with Monte Carlo simulations is presented to account for the splitting, asymmetric shifts, and near-TN enhancement, and the authors speculatively propose topologically protected phononic edge modes at domain boundaries.","tokens_in":10786,"tokens_out":7808,"duration_ms":81476,"significance":"If the interpretation holds, the paper demonstrates a new handle for non-volatilely controlling phonon angular momentum in a magnetic insulator, with potential implications for angular momentum transport and phononic device concepts. The experimental core is strong: the Raman, RMCD, and magnetization data are mutually consistent and cross-checked, and the ~0.2 T switching and the domain correspondence are directly evidenced. The 20% zero-field splitting is unusually large and the near-TN enhancement of the effective phonon magnetic moment is an interesting effect. The paper also benefits from a clear distinction between the measured effects and the more speculative edge-mode proposal. However, two load-bearing interpretive steps -- the conversion of the LR/RL channel exchange into phonon angular momentum reversal and the extraction of the 2.62 Bohr magneton number -- require additional scrutiny, and the theoretical model is not fully documented in the main text.","major_comments":[{"comment":"The conclusion that the LR/RL channel exchange in Fig. 3(a) demonstrates the reversal of phonon angular momentum assumes that the circular intensity asymmetry is controlled by the phonon eigenvector's pseudo-angular momentum. Magnetic crystals can exhibit antisymmetric (morphic) Raman tensor components proportional to magnetization, which produce circular intensity differences for linearly polarized phonons and would also reverse when the magnetization reverses. The opposite Stokes/anti-Stokes selection rules mentioned in the text are a strong indicator of chiral phonon angular momentum, but the paper does not quantify or exclude a morphic contribution. Please provide a Raman tensor analysis, or an independent eigenvector determination (e.g., X-ray measured phonon eigenvectors or ab initio calculation of the Raman tensor), to uniquely assign the LR/RL swap to phonon angular momentum reversal.","section":"Fig. 3(a), 'Magnetic control of chiral phonons'"},{"comment":"The effective phonon magnetic moment of 2.62 Bohr magnetons is extracted by taking a numerical derivative of the phonon frequency with respect to field near zero field. In the near-TN data the field dependence is clearly nonlinear (a strong initial slope that saturates at higher field), so the derivative value depends on the choice of field window and step. No error or fitting details are given. Please report the field increment and the derivative procedure, and provide an uncertainty estimate; if possible, fit the low-field branch to a model function to establish that a well-defined linear regime exists.","section":"Fig. 2(c), 'Phonon magnetic moment extraction'"},{"comment":"The Monte Carlo simulation is stated to reproduce the giant splitting, the asymmetric Zeeman shifts, and the near-TN enhancement, but all model parameters (phonon-magnon coupling gamma, detuning Delta, relative Fe-I versus Fe-II coupling weight, and the magnetic Hamiltonian parameters) are deferred to the Supplementary Materials, which is not part of the reviewed manuscript. If the parameters are tuned to match the measured spectra, then the model's success is not a parameter-free test. Please include the full parameter set and a sensitivity analysis, and state clearly which parameters are fixed by first principles or independent measurements and which are fitted.","section":"Microscopic model and Fig. 2(d)"}],"minor_comments":[{"comment":"The claim of a splitting reaching 20% of the phonon frequency is ambiguous because the splitting of 1.25 meV is 24% of the 5.21 meV P1a frequency and about 20% of the 6.32 meV P1b frequency; please specify the reference frequency used.","section":"Abstract and Section 1"},{"comment":"The phrase 'B is the increment of external' is incomplete; it should read 'external magnetic field'.","section":"Fig. 2(c) caption"},{"comment":"The sentence describing 'opposite selection rules in the Stokes and anti-Stokes processes' would be clearer if it explicitly stated which mode appears in which channel (LR or RL) for Stokes and for anti-Stokes.","section":"Section 1, paragraph 3"},{"comment":"The line-scan spectra show a spectral weight shift across the domain boundary; please state how the boundary width in the Raman mapping compares with the optical resolution, since this affects the interpretation of the two phononic domains as sharply separated.","section":"Fig. 4(c)"},{"comment":"References [34] and [35] are arXiv preprints; if the papers have appeared in peer-reviewed journals, please cite the published versions.","section":"References"}],"recommendation":"major_revision","confidential_remarks":"The experimental core is convincing and well cross-checked, and the manuscript fits the scope of cond-mat.mtrl-sci. The main text relies heavily on the Supplementary Materials for the theoretical model and its parameterization; the editor may wish to ensure that the Supplement is included in the review package and that the parameter choices and fitting procedure are documented. The PAM-switching interpretation and the 2.62 Bohr magneton extraction are the two points that most need strengthening before publication."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Dear Colleague,\n\nThis is a solid experimental paper with a big claim: non-volatile magnetic switching of chiral phonon angular momentum in a ferrimagnetic insulator. The core measurements – temperature-dependent circular Raman, field sweeps, and spatial mapping – are mutually consistent. The hysteresis of the 42/51 cm-1 modes synchronized with magnetization and RMCD is convincing. The zero-field splitting reaching 20% of the phonon frequency is huge, and the correlation with the ferrimagnetic order is clean. The phononic domain mapping locked to ferrimagnetic domains is the most novel part after the switching itself.\n\nThe novelty is real: prior work showed field-induced or magnetic-order-induced chiral phonon splitting, but not non-volatile hysteretic switching. The authors build sensibly on their own 2023 Nature Physics study of the parent antiferromagnet, and the citation pattern is appropriate.\n\nSoft spots, in order. First, the inference that the LR/RL channel swap proves reversal of the phonon angular momentum itself is not airtight. The paper relies on pseudo-angular momentum conservation and continuity with the parent compound, but does not directly measure atomic displacements. A magnetization-dependent antisymmetric (morphic) Raman tensor could, in principle, produce the same channel swap for linearly polarized modes. This is not fatal – the zero-field splitting and the opposite Stokes/anti-Stokes selection rules support the chiral assignment – but the authors should address it explicitly, ideally with a symmetry analysis of the Raman tensor or field-dependent measurements that separate the two mechanisms.\n\nSecond, the 2.62 μB effective phonon magnetic moment near TN comes from a numerical derivative of a nonlinear low-field curve, so the value is window-dependent. It is an effective quantity, so this is minor, but the uncertainty should be quantified.\n\nThird, the Monte Carlo model is described as minimal and the parameters are deferred to the Supplement. The paper honestly admits the simulation misses the non-monotonic field shift of P1b below TN, which limits the model's quantitative weight.\n\nThe topological edge-mode proposal is clearly speculative, and the authors explain why it cannot be observed with far-field Raman. Fine.\n\nOverall, this paper deserves a serious referee. The experimental core is reproducible and the interpretation is plausible. The referee should ask for raw data, model parameters, and a discussion of the morphic tensor alternative. I would cite this with a caveat, and it would make a good reading group discussion about what 'phonon angular momentum switching' really means.\n\nRecommendation: send to peer review.","headline":"Credible demonstration of non-volatile magnetic switching of chiral-phonon Raman response in a ferrimagnet, but the LR/RL channel swap alone does not fully prove phonon angular momentum reversal.","tokens_in":11314,"tokens_out":4936,"would_cite":true,"duration_ms":50508,"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":"In the ferrimagnetic insulator Fe1.75Zn0.25Mo3O8, a pair of chiral phonons splits by about 20% of their frequency at zero field, and a 0.2 T magnetic field non-volatilely switches their angular momentum together with the magnetization.","keywords":["chiral phonons","phonon angular momentum","ferrimagnetic insulator","magneto-Raman spectroscopy","phonon-magnon coupling","phonon magnetic moment","phononic domains","Fe1.75Zn0.25Mo3O8"],"falsifier":"Momentum-resolved inelastic X-ray or neutron scattering on a single ferrimagnetic domain below $T_N$ should resolve the eigenvectors of the 42 and 51 cm$^{-1}$ modes: if their atomic displacements are not circularly polarized in the ab-plane, or if the circular polarization does not reverse when the magnetization is switched by +/-0.2 T, the central claim fails.","tokens_in":2009,"feed_emoji":"🧲","tokens_out":2344,"duration_ms":69734,"temperature":0.7,"pith_summary":"This paper reports that in a polar ferrimagnetic insulator, phonons can carry angular momentum that is controlled by magnetism in a non-volatile way, without changing the material's structure or composition. It identifies a pair of low-energy chiral phonon branches, P1a and P1b, whose zero-field energy splitting reaches about 20% of the phonon frequency, a signature of spontaneously broken time-reversal symmetry in the magnetic ground state. Applying a moderate magnetic field of about 0.2 T reverses the ferrimagnetic magnetization and with it the phonon angular momentum, as read out from circular-polarization-resolved Raman selection rules. Near the magnetic ordering temperature, the effective phonon magnetic moment grows to 2.62 Bohr magnetons, larger than that of a magnon. If correct, this gives a practical, non-volatile handle on phonon angular momentum, useful for angular-momentum transport and phononic domain engineering.","feed_headline":"Chiral phonons in a ferrimagnet flip with a 0.2 T field","feed_subtitle":"Two phonon branches split by 20% at zero field; reversing the magnetization reverses their angular momentum.","key_machinery":"The central object is a pair of $E_2$-symmetry chiral phonons (P1a and P1b) at about 42 and 51 cm$^{-1}$, detected with circular-polarization-resolved Raman scattering in the LR and RL channels. The argument uses pseudo-angular-momentum conservation in the Raman process, with $C_3$ symmetry assigning $\\pm\\hbar$ pseudo-angular momentum to the two branches, and a long-wavelength correspondence between pseudo-angular momentum and real phonon angular momentum from atomic circular motion. The microscopic mechanism is an angular-momentum-selective phonon-magnon coupling: each chiral phonon branch hybridizes with magnons carrying the same angular momentum direction, acquiring a magnetic moment proportional to $(\\gamma/\\Delta)^2 \\mu_{\\mathrm{mag}}$; the ferrimagnetic molecular field produces the giant zero-field splitting, and spin fluctuations near $T_N$ amplify the field-induced shift. Monte Carlo simulation of this minimal model reproduces the observed splitting and asymmetric Zeeman behavior.","core_discovery":"On the paper's own terms, the central discovery is that the P1a and P1b Raman modes in Fe1.75Zn0.25Mo3O8 are a genuinely chiral phonon pair whose angular momentum is locked to the ferrimagnetic order. Spontaneous time-reversal symmetry breaking in the ferrimagnet lifts the degeneracy of the two circularly polarized branches, producing a splitting of 1.25 meV, about 20% of the phonon frequency, at zero applied field. The same ordering allows a moderate field to switch which branch carries which angular momentum, and the switch persists after the field is removed, as shown by a phonon Raman hysteresis synchronized with magnetization and magnetic circular dichroism. The phonon magnetic moment, derived from the field derivative of the phonon frequency, reaches 0.22 Bohr magnetons at low temperature and is enhanced to 2.62 Bohr magnetons near $T_N$ due to spin fluctuations. A phonon-magnon coupling model with Monte Carlo simulations reproduces the giant splitting, the asymmetric Zeeman shifts, and the enhancement near $T_N$. The paper also shows two phononic domain types locked to ferrimagnetic domains and proposes that their boundaries may host topologically protected phononic edge modes.","pith_inferences":["Beyond the paper's data, the same angular-momentum-selective coupling mechanism should make phonon angular momentum switchable in other polar ferrimagnets with $E$-symmetry phonons and a single easy magnetization axis; the essential ingredient is a net magnetization that can be coherently reversed.","The proposed topological edge modes at ferrimagnetic domain walls, if detected with tip-enhanced Raman or a local probe, would turn each domain wall into a switchable one-way channel for phonon angular momentum; this is an inference beyond the paper's domain mapping.","Because the P1 phonons carry nonzero pseudo-angular momentum along the $c$-axis, sweeping the temperature or field across $T_N$ could transiently release this angular momentum into the spin system; measuring a phononic contribution to magnetization dynamics would be a testable extension not reported here."],"forward_implications":["Non-volatile, field-free control of phonon angular momentum is possible in a magnetic insulator: a 0.2 T pulse writes the phonon chirality, and it stays after the field is removed.","The zero-field chiral phonon splitting of 1.25 meV acts as a sensitive optical probe of ferrimagnetic order and spin fluctuations, disappearing above $T_N$ and re-emerging with the magnetic order.","Near $T_N$, the effective phonon magnetic moment of 2.62 $\\mu_B$ exceeds a single magnon moment, implying chiral phonons can serve as strong angular-momentum transducers near the magnetic phase transition.","Ferrimagnetic domains template phononic domains with opposite Zeeman shifts, enabling micrometer-scale engineering of chiral-phonon regions.","Domain walls between phononic domains of opposite gap sign may host topologically protected chiral edge modes, whose propagation direction should reverse when the magnetization is switched."],"supporting_citations":[{"why":"Documents the context of broken-time-reversal-symmetry chiral phonon splitting in magnetic materials; frames the 1.25 meV zero-field splitting as a large effect.","marker":"[32–35]"},{"why":"Identifies the P1 phonons in the parent compound Fe2Mo3O8 and supplies the phonon-magnetic-moment formalism that the paper extends to the Zn-doped ferrimagnet.","marker":"[39]"},{"why":"Provides the magnon-polaron spectrum and spin-wave branches of the parent compound used to assign the magnon modes near the chiral phonons.","marker":"[40]"},{"why":"Gives the Raman selection rules for pseudo-angular-momentum conservation in a crystal with threefold symmetry, used to read out phonon angular momentum from LR and RL channels.","marker":"[41]"},{"why":"Supplies the conservation law of angular momentum in helicity-dependent Raman and Rayleigh scattering that underlies the interpretation of the channel switching.","marker":"[44]"},{"why":"Provides the formula for chiral-phonon magnetic moments induced by coupling with magnons, the model's central mechanism.","marker":"[43]"},{"why":"Gives the typical size of orbital magnetic moments of phonons, the baseline against which the measured 0.22 Bohr magneton moment is three orders of magnitude larger.","marker":"[42]"},{"why":"Establishes the doping-tunable ferrimagnetic phase of Fe2Mo3O8 with Zn substitution, the magnetic ground state that breaks time-reversal symmetry and enables non-volatile switching.","marker":"[37]"}],"fun_headline_variants":["Phonon angular momentum flips with magnetization in a ferrimagnet","Chiral phonons switch like memory in a ferrimagnetic crystal","2.62 Bohr magneton phonons controlled by a magnetic field","Phonons get magnetic memory: angular momentum flips non-volatilely","Chiral phonon pair with 20% splitting toggles with field"],"cache_read_input_tokens":13440,"weakest_assumption_plain":"The load-bearing premise is that the two low-energy modes seen in Raman are phonons with genuinely circular atomic motion, and that the LR/RL Raman channel swap tracks the reversal of that circular motion; if the modes were magnetic or magnetoelastic in character, the phonon-angular-momentum conclusion would not follow.","fun_headline_variants_meta":{"raw":{"variants":["Phonon angular momentum flips with magnetization in a ferrimagnet","Chiral phonons switch like memory in a ferrimagnetic crystal","2.62 Bohr magneton phonons controlled by a magnetic field","Phonons get magnetic memory: angular momentum flips non-volatilely","Chiral phonon pair with 20% splitting toggles with field"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00055,"raw_usage":{"total_tokens":2655,"prompt_tokens":1004,"completion_tokens":1651,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":620,"completion_tokens_details":{"reasoning_tokens":1555}},"tokens_in":620,"tokens_out":1651,"duration_ms":12749,"temperature":1.0,"reasoning_tokens":1555,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-10T19:02:35.525393+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Momentum-resolved inelastic X-ray or neutron scattering on a single ferrimagnetic domain below $T_N$ should resolve the eigenvectors of the 42 and 51 cm$^{-1}$ modes: if their atomic displacements are not circularly polarized in the ab-plane, or if the circular polarization does not reverse when the magnetization is switched by +/-0.2 T, the central claim fails.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the magnon-polaron spectrum and spin-wave branches of the parent compound used to assign the magnon modes near the chiral phonons."},{"cited_title":"Higuchi, N","cited_arxiv_id":null,"evidence_quote":"Gives the Raman selection rules for pseudo-angular-momentum conservation in a crystal with threefold symmetry, used to read out phonon angular momentum from LR and RL channels."},{"cited_title":"Tatsumi, T","cited_arxiv_id":null,"evidence_quote":"Supplies the conservation law of angular momentum in helicity-dependent Raman and Rayleigh scattering that underlies the interpretation of the channel switching."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the formula for chiral-phonon magnetic moments induced by coupling with magnons, the model's central mechanism."},{"cited_title":"Kurumaji, S","cited_arxiv_id":null,"evidence_quote":"Establishes the doping-tunable ferrimagnetic phase of Fe2Mo3O8 with Zn substitution, the magnetic ground state that breaks time-reversal symmetry and enables non-volatile switching."}],"review_version":1}