{"id":"e75939db-22a1-408d-a219-6fd1e32bde78","arxiv_id":"2502.10181","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"In collinear antiferromagnet Cr2O3, terahertz electric fields can excite the 0.165 THz spin resonance via the magnetoelectric effect, with an efficiency comparable to magnetic-field excitation.","lead":"A team showed that terahertz electric fields can make the spins in an antiferromagnet called Cr2O3 start to oscillate, without needing a magnetic field. The effect is nearly as strong as magnetic excitation, which could lead to faster, lower-power spintronic devices controlled by electric fields.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Residual H_THz leakage from unquantified polarization purity and crystal alignment could bias the magnetoelectric-geometry amplitude; the α=-45° null and domain-phase flip are strong internal controls but lack error bars, leaving the quantitative 'comparable torques' claim conditional.","rationale":"The reader identifies the same weakest assumption: the need for H_THz to be exactly parallel to the Néel vector in the magnetoelectric geometry. I agree that this is the linchpin for attributing the signal to the electric field. However, the paper contains internal controls that partially mitigate the concern: the π phase flip under domain reversal in the ME geometry (Fig. 2c) is inconsistent with a dominant residual Zeeman drive (which would not flip), and the sharp null at α=-45° (Fig. 1c) provides a direct, model-independent measure of the relative electric and magnetic torque amplitudes. These controls make a large H_x leakage very unlikely. What remains is a quantitative gap: the precision of the null angle, the phase-flip quality, and the polarization extinction are not reported, so the uncertainty on the 'comparable effects' claim is unknown. This does not invalidate the existence of electric-field-driven spin dynamics, but it does justify the CONDITIONAL verdict. I see no internal inconsistency in the derivation of Eq. (17) from Eq. (14), and the estimated λ_THz close to the static value is a reasonable consistency check. Thus I would keep the reader's CONDITIONAL verdict unchanged, with the same request for error bars, alignment characterization, and ideally a control material or an independent THz magnetoelectric measurement.","tokens_in":17772,"tokens_out":20607,"duration_ms":237516,"concrete_test":"Perform an angular scan of the THz polarization with fine steps (e.g., 5° or better) across the null at α=-45°, fit the signed oscillation amplitude to A(α) = A_E cos α + A_H sin α, and report the fitted null angle and the ratio A_H/A_E with 95% confidence intervals. Repeat in a known single domain. Independently measure the THz polarization extinction ratio of the two wire-grid polarizers with a crossed-polarizer setup, and determine the c-axis orientation of the Cr2O3 slab by X-ray diffraction. If the null angle is -45° within the uncertainty and the residual Zeeman contribution in the ME geometry is below, say, 5% of the magnetoelectric amplitude, the concern is settled.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim that the 0.165 THz spin resonance is excited by the THz electric field in the H‖L geometry depends on the Zeeman torque γω_AH_x being negligible. The paper does not report the THz polarization extinction ratio, the accuracy of the crystal c-axis alignment with the y-axis, or the residual H_x component in the magnetoelectric geometry. A small misalignment (c-axis tilt or polarizer leakage) would introduce a Zeeman drive that scales linearly with H and does not flip phase under L reversal, whereas the magnetoelectric drive does flip phase. The observed π phase shift in the ME geometry and the near-zero amplitude at α=-45° are strong internal controls: they imply the ME torque dominates and that the two torque amplitudes are approximately equal. However, the phase-flip measurement is shown only qualitatively (Fig. 2c, 'clear difference'), and the α-scan in Fig. 1c has no reported error bars or fitted null-angle uncertainty. A contamination of, say, 10-20% of the ME amplitude from residual H_x would still produce an apparent phase shift close to π and would shift the null by several degrees, yet would change the inferred ratio of torques by 10-20%, undermining the precise claim of 'comparable effects'. Since the THz magnetoelectric parameter λ_THz is extracted from this equality, the quantitative comparison is not independently verified. Without a quantitative control (e.g., a non-magnetoelectric antiferromagnet in the same geometry, or a measured extinction ratio and X-ray alignment), the uncertainty in the central quantitative claim remains unknown.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript reports THz pump–infrared probe experiments on a single crystal of the collinear antiferromagnet Cr2O3 and shows that the 0.165 THz antiferromagnetic resonance can be excited not only in the conventional Zeeman geometry (H_THz perpendicular to the Néel vector L) but also in the geometry H_THz parallel to L, i.e. E_THz perpendicular to L. The authors attribute the second excitation channel to the linear magnetoelectric torque and support this with an angle-dependent amplitude maximum at α = 45° and a near-null at α = −45°, linear scaling with THz field strength, and a π phase shift upon reversal of the Néel vector in the magnetoelectric geometry. A two-sublattice Lagrangian model is developed and reduced to Eq. (17), m_ẍ + ω_M² m_x = ±ω_A ω_ME⊥ E_x + γω_A H_x, which accounts for the observed domain-dependent phase behavior. From the assumed near-equality of the two torques in Eq. (17), the authors estimate THz magnetoelectric parameters λ_THz⊥κ_THz⊥ ≈ −1 and α_THz⊥ ≈ −1.2 × 10⁻⁴.","tokens_in":18040,"tokens_out":12245,"duration_ms":117979,"significance":"If the central claim holds, the paper is significant: electric-field excitation of spin resonance in a collinear antiferromagnet with efficiency comparable to the THz Zeeman torque would extend ultrafast electric-field control beyond electromagnon-based multiferroics and is directly relevant to antiferromagnetic spintronics and THz magnonics. The qualitative claim rests on strong internal controls that are independent of parameter fitting: the α-dependence with a null at −45°, the linear field scaling, and the domain-reversal phase flip are all predicted by the symmetry-based model and are observed. The main reservations concern the quantitative branch: the amplitudes and phase data are presented without error bars, the residual Zeeman contamination in the nominally zero-torque geometry is not bounded, and the THz magnetoelectric parameters are extracted from the same 'comparable torques' assumption that the paper aims to establish.","major_comments":[{"comment":"The magnetoelectric geometry is defined by H_THz parallel to L, so the Zeeman torque γω_A H_x in Eq. (17) is assumed to vanish. The manuscript does not report the accuracy of the crystal c-axis alignment relative to the y axis, the extinction ratio of the two wire-grid polarizers, or any direct bound on the residual H_x in this geometry; a 2° misalignment already gives about 3.5% and a 5° misalignment about 9% H_x contamination. Such contamination shifts the α-null in Fig. 1c by several degrees and changes the inferred magnetoelectric/Zeeman amplitude ratio by the same percentage without destroying the near-π phase flip seen in Fig. 2c. Please state the alignment accuracy and polarization purity quantitatively and provide an explicit upper bound on the residual Zeeman contribution in the magnetoelectric geometry.","section":"Fig. 1a / Experimental setup"},{"comment":"The values λ_THz⊥κ_THz⊥ ≈ −1 and α_THz⊥ ≈ −1.2 × 10⁻⁴ are obtained by inserting the assumption |ω_Aω_ME⊥E_x| ≈ |γω_A H_x| into Eq. (17), i.e., by assuming the very 'comparable torques' result that the paper claims to demonstrate. This makes the quantitative branch of the parameter estimate circular, although the qualitative claim of electric-field excitation is not circular because the α-scan null and the domain-reversal phase shift are independent of parameter fitting. Please provide an independent determination of the THz magnetoelectric parameter from the static α⊥ together with measured dielectric and optical responses, or explicitly label the THz parameters as order-of-magnitude estimates made under the equality assumption.","section":"Methods, after Eq. (20)"},{"comment":"The amplitudes, the α-null angle, the field-scaling data, and the domain-phase comparison are presented without error bars or statistical uncertainties. Since the central claim is that the electric- and magnetic-field torques produce 'comparable' spin dynamics, the paper needs to report measurement statistics, fit uncertainties (especially for the null angle in Fig. 1c), and residuals for the linear fits in Fig. 1d, so that the allowed residual-H contamination can be assessed against the claimed precision.","section":"Figs. 1c–1e and 2c"}],"minor_comments":[{"comment":"Please define the L↑/L↓ sign convention for the ± and ∓ signs in these equations and check the sign chain between Eqs. (16) and (17), since the text currently requires the reader to infer the domain assignment.","section":"Eqs. (1), (14), (16)–(18)"},{"comment":"The statement that 'the combination l_y l_z ... behaves like mx' is compressed; an explicit two-line derivation would help the reader see why the domain-phase test is insensitive to whether the detected signal is m_x or l_y l_z.","section":"After Eq. (18)"},{"comment":"The number of averaged laser shots or measurement repeats used for each transient is not stated; please add this information to the Methods section.","section":"Experimental setup"},{"comment":"The main text refers to 'Supplementary Figure' and 'Supplementary Note' without numbering more than once; please itemize the supplementary material. Also, in the Experimental setup paragraph, 'polarizes' should be 'polarizers'.","section":"Supplementary material / text"}],"recommendation":"major_revision","confidential_remarks":"The paper is likely to be of interest to the ultrafast magnetism and antiferromagnetic spintronics communities, and the qualitative claim is supported by controls that are not circular. The main risk is that the quantitative 'comparable torques' statement and the extracted THz magnetoelectric parameters are not yet supported by the reported statistics and alignment/polarization-purity information. I would ask the editor to require the authors to supply the missing error bars, a quantitative bound on residual Zeeman contamination, and an independent or explicitly labeled parameter estimate before publication."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Short version: the paper shows that the 0.165 THz AFM resonance in Cr2O3 can be excited by the THz electric field via the linear magnetoelectric torque, with a domain-reversal phase flip and an α-scan null that are proper internal controls. I think the central result is real and is new: prior THz work on collinear AFMs used the Zeeman torque, and electromagnon work was in non-collinear magnets. The two-torque model, Eq. (17), fits the angular, linear-field, and domain-phase data without free fitting of the qualitative features. Credit where due: the experiment is clean, the domain visualization is a nice touch, and the α = −45° null is exactly the test that distinguishes a magnetoelectric drive from a residual Zeeman leak.\n\nThe soft spots are where the reader and stress-test put them. First, the quantitative claim that the two torques are 'comparable' rests on the null angle and on the assumption that E and H fields are accurately known. A small c-axis tilt or wire-grid polarizer leakage introduces a residual H_x in the 'ME geometry'; that shifts the null and spoils the phase flip by a small amount, and the paper does not report the extinction ratio or the crystal alignment precision. So the 10–20% error is real and unquantified. Second, Fig. 1c,d have no error bars; the null at −45° looks sharp, but 'looks' is not a number. Third, the extracted THz magnetoelectric parameters are circular: they set λ_THz κ_THz from the observed near-equality and then use it to compute α_THz. That is fine for a consistency check, not a measurement. The data/code availability line is also just 'upon request.' None of this kills the qualitative result, but it keeps the comparison from being quantitative.\n\nThe conclusions overreach a bit: they say 'control' and 'game-changer' when the data show resonant excitation, not switching or writing. The paper would be stronger if they toned that down.\n\nWho this is for: AFM spintronics and THz magnonics people; anyone working on magnetoelectric control of spins. It deserves a serious referee: the claim is important, the internal controls are mostly right, and the flaws are conventional fixable ones—error bars, alignment characterization, and ideally an independent THz measurement of α⊥. I would send it out with a request for those additions.","headline":"A credible demonstration of THz electric-field spin resonance in Cr2O3, with a solid qualitative core and a conditional quantitative branch that needs error bars and alignment characterization.","tokens_in":18656,"tokens_out":5375,"would_cite":true,"duration_ms":55172,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["75.50.Ee","75.85.+t","76.50.+g"],"model":"deepseek-v4-flash","headline":"In collinear antiferromagnet Cr2O3, the 0.165 THz spin resonance is excited by the electric field of a THz pulse via the linear magnetoelectric torque, with efficiency comparable to the Zeeman torque from the pulse's magnetic field.","keywords":["antiferromagnet","magnetoelectric effect","terahertz spin dynamics","Cr2O3","magnon","THz excitation","spin resonance","electric-field control"],"falsifier":"Take a single-domain Cr2O3 sample, measure the oscillation amplitude versus the angle between H_THz and the c-axis through the nominal H_THz ∥ L orientation with the polarization and crystal alignment characterized to about 0.1°, and check whether the signal goes through a flat, non-zero minimum as the model predicts for the magnetoelectric torque, or falls to zero as it would if only a residual perpendicular Zeeman component were responsible.","tokens_in":17547,"feed_emoji":"⚡","tokens_out":4875,"duration_ms":44894,"temperature":0.7,"pith_summary":"This paper shows that in the collinear antiferromagnet Cr2O3, a THz pulse can set the spins precessing using its electric field alone, without an electromagnon. The 0.165 THz antiferromagnetic resonance is excited by the linear magnetoelectric torque, which converts the applied electric field into an effective magnetic action on the spins. In a freely propagating THz wave, this electric torque is as strong as the conventional Zeeman torque from the pulse's magnetic field. The two mechanisms are distinguished experimentally by their different dependence on the Néel vector direction and field orientation: rotating the THz field by 45° versus −45° maximizes one and cancels the other, and reversing the antiferromagnetic domain flips the phase of the electric-field-driven signal but not the magnetic-driven one. The implication is that electric-field pulses, which are easier to localize on a chip than magnetic fields, can be a viable ultrafast control knob for antiferromagnetic spins.","feed_headline":"THz electric field drives spins in antiferromagnet Cr2O3","feed_subtitle":"A 0.165 THz resonance is excited by the magnetoelectric torque as strongly as by the Zeeman torque.","key_machinery":"The load-bearing object is the linear magnetoelectric torque, derived from a two-sublattice Lagrangian for Cr2O3 with energy U = −λ⊥ mx ly Ex − λ‖ my ly Ey − (γS/ħ)(mx Hx + my Hy). This interaction converts an applied electric field into a term in the antiferromagnetic-resonance equation of motion of exactly the same form as the Zeeman torque, with strength set by the magnetoelectric coefficient α⊥. The model reduces the full four-variable dynamics (canting angles ǫ, β and deviations ϑ1, ϕ1) to a driven harmonic oscillator for mx, whose phase and amplitude depend on the domain orientation, providing the signatures that separate electric from magnetic excitation.","core_discovery":"Using THz pump–infrared probe experiments on a single crystal of Cr2O3 below TN = 307 K, the authors observe coherent spin oscillations at 0.165 THz in two geometries: the Zeeman geometry where H_THz is perpendicular to L and the magnetoelectric geometry where H_THz is parallel to L (equivalently E_THz is perpendicular to L). The second geometry is striking because the Zeeman torque must vanish there; nevertheless the oscillation amplitude is comparable to the first geometry and scales linearly with the THz field. The authors attribute this to a linear magnetoelectric torque, captured in their two-sublattice Lagrangian by a term −λ⊥ mx ly Ex. Solving the equations of motion yields mẍ + ωM² mx = ±ωA ωME⊥ E_x + γωA H_x, and the model accounts for the observed linear scaling, the π phase shift upon rotating the fields by π, and the domain-dependent phase flip in the magnetoelectric geometry. The conclusion is that in a collinear magnetoelectric antiferromagnet, the THz electric field excites the same antiferromagnetic resonance as the THz magnetic field, with comparable efficiency.","pith_inferences":["If the effect generalizes, other collinear magnetoelectric antiferromagnets should show similar THz electric-field-excited resonances; a quick survey could be done by repeating the same two-geometry test on candidate crystals.","The paper neglects the internal parametric torques (Ey and Hy terms) at the field strengths used; at higher THz intensities these could drive parametric instabilities, turning the linear effect into a nonlinear amplification channel.","Because electric fields are easier to confine than magnetic fields, the same mechanism might enable writing of nanoscale antiferromagnetic bits with picosecond electric pulses, though the paper demonstrates coherent precession rather than switching."],"forward_implications":["Electric-field pulses applied via on-chip electrodes could excite and manipulate antiferromagnetic spins at THz rates without needing bulky magnetic antennas.","THz spectroscopy and magnonics experiments on magnetoelectric antiferromagnets must now consider electric-field coupling alongside magnetic-field coupling when interpreting signals.","The comparable magnitude of the two torques suggests that engineering the magnetoelectric coefficient could make electric-field control of antiferromagnetic order practical.","The mechanism works in a collinear antiferromagnet without electromagnons, broadening the class of materials for electric-field-driven spin dynamics."],"supporting_citations":[{"why":"Astrov's discovery of the magnetoelectric effect in antiferromagnetics establishes the fundamental phenomenon the paper exploits.","marker":"[27]"},{"why":"Astrov's study of magnetoelectric effect in chromium oxide supplies the value of the magnetoelectric coefficient α⊥.","marker":"[28]"},{"why":"Foner's high-field antiferromagnetic resonance work provides the resonance frequency and exchange/anisotropy parameters used in the model.","marker":"[29]"},{"why":"Kampfrath et al. demonstrated coherent THz control of antiferromagnetic spin waves via the Zeeman torque, the baseline mechanism the electric torque is compared against.","marker":"[22]"},{"why":"Belov et al. give the symmetry-allowed magnetoelectric energy terms in Cr2O3, used to write the interaction Hamiltonian of the model.","marker":"[35]"},{"why":"Bousquet et al. address the sign of the linear magnetoelectric coefficient in Cr2O3, used to interpret the domain-dependent phase flip.","marker":"[40]"},{"why":"Mukhin et al. provide the magnon linewidth and damping time used in the numerical simulations.","marker":"[46]"},{"why":"Fiebig et al. developed the SHG domain visualization technique that allows the authors to measure single antiferromagnetic domains separately.","marker":"[42]"}],"fun_headline_variants":["THz electric field rivals magnetic field for Cr2O3 spins","Magnetoelectric torque matches Zeeman in Cr2O3 at THz","THz E-field drives same spin mode as H-field in Cr2O3","THz E-field matches H-field in driving Cr2O3 spins"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The identification of the electric-field excitation rests on the geometry where the THz magnetic field is exactly parallel to the Néel vector; any residual perpendicular magnetic component from crystal misalignment or polarization leakage would produce the same signal through the Zeeman torque.","fun_headline_variants_meta":{"raw":{"variants":["THz electric field rivals magnetic field for Cr2O3 spins","Magnetoelectric torque matches Zeeman in Cr2O3 at THz","THz E-field drives same spin mode as H-field in Cr2O3","THz E-field matches H-field in driving Cr2O3 spins"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001644,"raw_usage":{"total_tokens":6576,"prompt_tokens":1031,"completion_tokens":5545,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":647,"completion_tokens_details":{"reasoning_tokens":5464}},"tokens_in":647,"tokens_out":5545,"duration_ms":37091,"temperature":1.0,"reasoning_tokens":5464,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T19:05:28.255029+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take a single-domain Cr2O3 sample, measure the oscillation amplitude versus the angle between H_THz and the c-axis through the nominal H_THz ∥ L orientation with the polarization and crystal alignment characterized to about 0.1°, and check whether the signal goes through a flat, non-zero minimum as the model predicts for the magnetoelectric torque, or falls to zero as it would if only a residual perpendicular Zeeman component were responsible.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Astrov's discovery of the magnetoelectric effect in antiferromagnetics establishes the fundamental phenomenon the paper exploits."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Astrov's study of magnetoelectric effect in chromium oxide supplies the value of the magnetoelectric coefficient α⊥."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Foner's high-field antiferromagnetic resonance work provides the resonance frequency and exchange/anisotropy parameters used in the model."},{"cited_title":"V ., Vorob’ev, G","cited_arxiv_id":null,"evidence_quote":"Belov et al. give the symmetry-allowed magnetoelectric energy terms in Cr2O3, used to write the interaction Hamiltonian of the model."},{"cited_title":"A., T ravkin, V","cited_arxiv_id":null,"evidence_quote":"Mukhin et al. provide the magnon linewidth and damping time used in the numerical simulations."},{"cited_title":"L, G., and Pisarev, R","cited_arxiv_id":null,"evidence_quote":"Fiebig et al. developed the SHG domain visualization technique that allows the authors to measure single antiferromagnetic domains separately."}],"review_version":1}