{"id":"f0a609fe-b828-441b-8179-12c73c229d80","arxiv_id":"2501.01150","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":5,"one_line_summary":"Numerical LLG simulations of Mn3Ge show thickness-dependent terahertz standing magnon modes and a spin-current switching phase diagram with picosecond switching regions.","lead":"This paper uses computer simulations to show that a short pulse of spin-polarized current can excite terahertz-frequency standing spin waves and flip the magnetic ground state of the non-collinear antiferromagnet Mn3Ge. It suggests antiferromagnets could serve as fast, stray-field-free building blocks for terahertz magnonics and ultrafast memory.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"STT normalization in Eq. (2) is inconsistent with the exponential current decay in Eq. (3): using 1/d instead of 1/λ_STT absorbs only λ_STT/d of the injected spin current, biasing the switching phase diagram.","rationale":"The paper's central claims are computational: a thickness-dependent THz magnon spectrum and a spin-current switching phase diagram. The magnon spectrum is a linear response property that should survive a rescaling of the drive strength, so the standing-wave frequencies are not the main casualty of the normalization error. The switching phase diagram, however, depends directly on the magnitude and spatial distribution of the spin torque. If the torque is too weak by a factor of λ_STT/d, then the reported critical currents, the switching/non-switching regions in Fig. 4, and the statement that thicker films require larger j0 are all quantitatively wrong. This is an internal inconsistency because the exponential decay in Eq. (3) already specifies that the torque is concentrated near the interface; the prefactor 1/d in Eq. (2) contradicts that concentration by spreading the torque over the entire film thickness. The reader's weakest assumption focused on the values of λ_STT, θ, and the pulse profile in Eq. (3); that is a related but distinct concern. Our objection is stronger: independent of what the true parameters are, the combination of Eqs. (2) and (3) does not conserve spin angular momentum. No amount of parameter tuning can fix that. The appropriate response is to correct the prefactor, rerun the simulations, and re-report the phase diagram and critical current values. The magnon spectrum claim may then be verified as unaffected. Therefore the verdict remains conditional, but with a more specific and more load-bearing condition than the reader's parameter-uncertainty concern.","tokens_in":9805,"tokens_out":15842,"duration_ms":180895,"concrete_test":"Re-run the LLG simulations for d=4.18 nm and d=10.78 nm with the STT prefactor in Eq. (2) changed from γℏθ/(2edM_s) to γℏθ/(2eλ_STT M_s), keeping all other parameters (including j0, τ1, τ2, α) fixed. Then compare the switching phase diagram of Fig. 4 and the critical j0 as a function of d against the published results. If the switching boundaries shift by more than the grid spacing, or if the required j0 no longer increases with d for d≫λ_STT, the phase diagram and the thickness dependence of switching are artifacts of the 1/d normalization. As a secondary check, verify that the standing-wave peak frequencies remain at the nπ/d positions; if they shift, the standing-wave assignment also needs revision.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The most load-bearing flaw is an internal normalization inconsistency between Eq. (2) and Eq. (3), not simply uncertain parameters. Equation (2) writes the spin-transfer torque as (γℏθ/2edM_s) j_s S×(S×ẑ), with d the total Mn3Ge thickness. Equation (3) then sets j_s(z,t)=j0 e^{-z/λ_STT} times the temporal factor. For a spin current that is absorbed as it propagates, the local torque density must be proportional to -∂j_s/∂z = j_s/λ_STT, so the prefactor should be γℏθ/(2eλ_STT M_s), not γℏθ/(2ed M_s). The present form integrates to ∫0^d (j0 e^{-z/λ}/d) dz = j0 λ/d, so for d=4.18 nm and λ_STT=1 nm the model transfers only about 24% of the injected spin angular momentum, and for d=10.78 nm about 9%. Because the prefactor contains the total thickness d, the reported need for larger j0 in thicker films is partly a model artifact rather than a genuine penetration-length effect. The switching phase diagram in Fig. 4, including the critical j0 and τ2 boundaries, is therefore quantitatively unreliable as stated. The magnon eigenmode frequencies should be insensitive to the overall torque scale, so the standing-wave spectrum claim is less affected, but mode amplitudes and the thickness trend of the switching threshold can change.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript studies a Fe|Au|Mn3Ge spin valve driven by a femtosecond-laser-generated spin current. The authors solve an atomistic Landau-Lifshitz-Gilbert (LLG) equation with a spin-transfer torque term and report two main results: (i) FFT magnon spectra showing standing spin-wave modes whose number increases with Mn3Ge thickness, with a thickness-independent fundamental antiferromagnetic resonance around 0.13 THz, and (ii) a switching phase diagram in the (j0, tau2) plane for the Neel vector and octupole moment, containing switching and non-switching regions. They conclude that non-collinear antiferromagnets can support thickness-tunable THz magnon modes and picosecond-scale switching.","tokens_in":10019,"tokens_out":8965,"duration_ms":89676,"significance":"If the results are correct, this is a useful numerical demonstration for the emerging fields of THz magnonics and ultrafast antiferromagnetic spintronics. The simulation approach is standard atomistic LLG dynamics with parameters taken from the literature, and no parameter is fitted to the target spectra or phase boundaries, so the qualitative observations (thickness-dependent mode counting, existence of switching and non-switching regions) are transparent and reproducible in principle. The main value is the proposal that Mn3Ge in particular combines standing-wave THz magnon modes with current-controlled switching. The paper does not provide an analytical theory or experimental validation, but for a simulation study that is acceptable if the numerical implementation is clean and the reported claims are supported by the data.","major_comments":[{"comment":"The spin-transfer torque normalization is internally inconsistent. Equation (2) uses the prefactor gamma hbar theta/(2 e d M_s) with d the total film thickness, while Eq. (3) makes j_s decay spatially as j0 e^{-z/lambda_STT}. For a spin current that is absorbed as it propagates, the local torque density should be proportional to -partial j_s/partial z = j_s/lambda_STT, so the prefactor should be gamma hbar theta/(2 e lambda_STT M_s), not gamma hbar theta/(2 e d M_s). With the present form, the integrated torque is proportional to integral_0^d (j0 e^{-z/lambda}/d) dz = j0 (lambda/d)(1-e^{-d/lambda}), which is about 24% of the injected spin angular momentum for d=4.18 nm and about 9% for d=10.78 nm. Because the prefactor contains the total thickness d, the statement in Sec. III that larger j_s is needed for larger d is partly an artifact of this normalization, and the quantitative switching phase diagram in Fig. 4, including the critical j0 and tau2 boundaries, is not reliable as stated. The magnon eigenfrequencies in Figs. 2 and 3 are less affected because the torque scale multiplies all driving terms, but the mode amplitudes and any thickness trend of the switching threshold should be recomputed with a conserved normalization.","section":"Sec. II, Eqs. (2)-(3)"},{"comment":"The Discussion claims that the use of non-collinear antiferromagnets can boost resonant frequencies 'higher than 10 THz', but no computed spectrum with a numerical frequency axis is reported and the maximum frequency reached in the simulations is never stated in the text. The only quantitative frequency given is the fundamental mode at 0.13 THz, and the text mentions up to n=10 visible modes for d=10.78 nm without stating their frequencies. This unsupported quantitative claim should either be removed or substantiated by reporting the actual maximum peak frequency and the frequency range shown in Figs. 2 and 3. Without this, the quantitative reach of the central THz claim cannot be assessed.","section":"Sec. IV, Discussion"}],"minor_comments":[{"comment":"The FFT amplitude in Eq. (4) is written without an absolute value or a normalization by Nsteps; please state that the plotted quantity is |A(z,f)| and specify the normalization used, so that the amplitude comparisons in Figs. 2 and 3 are well-defined.","section":"Sec. II, Eq. (4)"},{"comment":"The frequency axes of Figs. 2 and 3 are not discussed in the text and no numerical frequency values are quoted beyond the 0.13 THz fundamental mode; please add explicit frequency axes and units and state the maximum frequency shown.","section":"Sec. III, Figs. 2-3"},{"comment":"The boundary conditions along the z direction are not stated; since the standing-wave picture relies on reflection at the two ends of the Mn3Ge layer, please specify whether the top and bottom surfaces are open, absorbing, or periodic.","section":"Sec. II"},{"comment":"The phase diagram would be easier to assess if the authors stated the grid of (j0, tau2) points used and the criterion used to classify a point as switching (for example, the sign of l1 after a fixed integration time).","section":"Sec. III, Fig. 4"},{"comment":"The choices theta=1 and lambda_STT=1 nm are taken from ferromagnetic-material literature; because these are strong assumptions for Fe|Au|Mn3Ge, a brief sensitivity check varying lambda_STT and theta would strengthen the switching conclusions.","section":"Sec. II"},{"comment":"There are several grammatical slips that should be corrected, including 'absorbtion' in Sec. II, 'switching process occurs place for all' after Fig. 4, and 'Similar to region-III, spins return to their initial states in region V indicates' in Sec. III.","section":"Secs. II-III"}],"recommendation":"major_revision","confidential_remarks":"The normalization inconsistency between Eqs. (2) and (3) is correctable by replacing the prefactor 1/d with 1/lambda_STT and rerunning the switching simulations; I do not see a reason to reject, because the standing-wave mode count and the existence of switching and non-switching regions are plausible and the fix is local. The >10 THz claim should be either substantiated with numerical frequencies or removed. I recommend major revision rather than rejection."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Short version: there is a real result here — the thickness-dependent standing spin-wave spectrum in Mn3Ge — but the switching half of the paper is built on an STT term whose normalization does not match its own z-decaying current, and the Discussion's >10 THz claim has no spectrum behind it.\n\nWhat is new: a systematic atomistic LLG study of Mn3Ge under a femtosecond spin-current pulse, showing that the number of magnon modes grows with thickness and that the switching behavior forms a phase diagram in the j0–τ2 plane. The methods are standard, the parameters come from the literature, and the FFT spectra qualitatively support the standing-wave picture. The fundamental AFM mode at 0.13 THz and the higher confined modes are plausible. That part deserves to be taken seriously.\n\nSoft spots:\n\n1. The stress-test note holds up. Eq. (2) uses 1/d in the STT prefactor while Eq. (3) makes j_s decay as e^{-z/λ_STT}. For a current absorbed as it propagates, the local torque should be proportional to -∂j_s/∂z = j_s/λ_STT. With the printed form the integrated torque is j0 λ_STT/d, so for d = 4.18 nm and λ_STT = 1 nm only about 24% of the injected angular momentum is transferred, and less for thicker films. The switching thresholds and the reported trend that thicker films need larger j0 are therefore quantitatively contaminated by the prefactor choice, not just by material parameters. This is fixable, but it needs fixing before the switching claims are accepted.\n\n2. The Discussion says non-collinear antiferromagnets can boost resonant frequencies above 10 THz, but no spectrum in the paper shows that. Either show the relevant spectrum or cut the sentence.\n\n3. No code, data, or parameter files are provided, and there is no sensitivity analysis for λ_STT = 1 nm and θ = 1. For a purely computational paper, that is now standard practice and should be requested.\n\nThe magnon spectrum claim is probably robust to these issues because eigenfrequencies do not depend on the torque amplitude; the switching thresholds do. This is a conditional accept-to-revise situation, not a desk reject. A serious referee should engage with it and force the normalization fix and the 10 THz cleanup. After that, it becomes a solid Mn3Ge simulation paper.","headline":"A legitimate Mn3Ge LLG study with a credible thickness-dependent standing-wave spectrum, but the switching phase diagram rests on an inconsistent STT normalization and the 10 THz claim is unshown.","tokens_in":10684,"tokens_out":4346,"would_cite":false,"duration_ms":46133,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["82D40"],"pacs":[],"model":"deepseek-v4-flash","headline":"A femtosecond spin-current pulse generates thickness-tunable terahertz standing spin waves in the non-collinear antiferromagnet Mn3Ge, and can reverse its Neel vector within picoseconds.","keywords":["terahertz magnonics","non-collinear antiferromagnet","Mn3Ge","spin transfer torque","standing spin waves","Neel vector switching","femtosecond spin current","atomistic spin simulations"],"falsifier":"A thickness-series experiment on Mn3Ge thin films (e.g., 4, 6, 8, 10 nm) under femtosecond laser excitation should resolve standing spin-wave peaks whose number increases and whose frequencies decrease with thickness; its absence, or a spectrum fixed by the bulk magnon dispersion rather than by $k_n = n\\pi/d$, would falsify the standing-wave claim. The switching claim is testable by measuring the Neel vector (via X-ray magnetic linear dichroism or the anomalous Hall signal) after single femtosecond pulses and checking that switching occurs only in the predicted $(j_0, \\tau_2)$ regions.","tokens_in":1758,"feed_emoji":"🧲","tokens_out":2613,"duration_ms":53109,"temperature":0.7,"pith_summary":"This paper claims that a femtosecond-laser-generated spin current pulse can excite terahertz standing spin waves in the non-collinear antiferromagnet Mn3Ge, with a magnon spectrum whose mode count and frequencies are controlled by the film thickness. In the same heterostructure (Fe|Au|Mn3Ge), the spin transfer torque can also reverse the Neel vector and the associated magnetic octupole moment within a few picoseconds, provided the current amplitude and pulse duration lie in the appropriate region of a switching phase diagram. If true, this makes non-collinear antiferromagnets a practical platform for terahertz magnonics and ultrafast memory, and gives thickness as a tuning knob for terahertz magnon frequencies.","feed_headline":"THz spin waves in Mn3Ge switch with a spin-current pulse","feed_subtitle":"The magnon spectrum depends on film thickness, and a switching phase diagram maps where the Neel vector flips in picoseconds.","key_machinery":"The central object is the spin Hamiltonian of Mn3Ge (intralayer antiferromagnetic exchange, interlayer ferromagnetic and antiferromagnetic exchange, intralayer Dzyaloshinskii-Moriya interaction, and easy-axis in-plane anisotropy) combined with the Landau-Lifshitz-Gilbert equation augmented by an anti-damping spin transfer torque. The torque is driven by the assumed spin-current pulse $j_s = j_0 e^{-z/\\lambda_{\\mathrm{STT}}} e^{-t/\\tau_2}/(1+e^{-(t-t_0)/\\tau_1})$, whose finite penetration depth and femtosecond time profile create a spatially and temporally localized excitation. Reflections of the excited spin waves at the open film surfaces produce standing modes at wave vectors quantized by the thickness, which is the mechanism that ties the magnon spectrum to $d$.","core_discovery":"The paper establishes, through numerical solution of the Landau-Lifshitz-Gilbert equation with spin transfer torque, that a spatiotemporal spin current pulse in Mn3Ge generates standing spin waves at quantized wave vectors $k_n = n\\pi/d$ imposed by the finite film thickness. The fundamental antiferromagnetic resonance sits at 0.13 THz regardless of thickness, while the higher modes shift downward and become more numerous as the film thickens; for $d = 10.78$ nm the spectrum shows $n = 0$ through $n = 10$ modes. The same torque, with $j_0$ and $\\tau_2$ in the switching phase-diagram regions labeled II and IV, rotates every sublattice spin by $180^\\circ$, reversing both Neel vectors and the octupole moment. Too weak or too strong a pulse leaves the ground state unchanged, and an intermediate case returns via a full $360^\\circ$ rotation. The claim is that thickness acts as a control parameter for terahertz magnon modes and that spin-current pulses provide deterministic picosecond switching in a non-collinear antiferromagnet.","pith_inferences":["If the standing-wave condition holds, the same quantization argument should apply to other non-collinear antiferromagnets with kagome order, making the mode spectrum a fingerprint of thickness rather than of material-specific dispersion.","A natural extension is to test whether the 360-degree return rotation leaves any transient topological or handedness signature that could be read out electrically even when no net switching occurs.","The phase diagram suggests that modest heating or strain, which alters the anisotropy parameter $K$, will shift the switching boundaries; those boundaries are therefore a sensitive probe of the energy barrier between degenerate ground states."],"forward_implications":["Thickness becomes a practical control knob for terahertz magnon frequencies in Mn3Ge, with mode spacing set by the confinement condition $k_n = n\\pi/d$ shrinking as the film grows.","A single spin-current pulse can reverse the Mn3Ge ground state (Neel vector and octupole moment) in picoseconds without an applied field.","Switching is non-monotonic in pulse strength: a 360-degree rotation returns the system to its initial state at high $j_0$, defining non-switching islands in the phase diagram.","Two pulses separated by a few picoseconds can double-switch the system back to its original configuration, suggesting a way to write and erase with the same device.","Standing spin waves at several terahertz raise the accessible magnon frequencies beyond those reported for Fe and collinear Mn2Au."],"supporting_citations":[{"why":"Supplies the superdiffusive spin transport mechanism that converts the femtosecond laser pulse into a spin current.","marker":"[10]"},{"why":"Provides the experimental and theoretical basis for nanoscale interface confinement of ultrafast spin transfer torque driving non-uniform spin dynamics.","marker":"[13]"},{"why":"Establishes the micromagnetic approach for ultrafast magnon generation by femtosecond spin current pulses, which the standing spin wave picture extends.","marker":"[14]"},{"why":"Demonstrates Neel vector switching and terahertz spin-wave excitation in Mn2Au, the collinear baseline this paper compares against.","marker":"[24]"},{"why":"Shows deterministic electrical switching of a non-collinear antiferromagnet (Mn3Sn), the switching precedent this work adapts to Mn3Ge with a laser-driven spin current.","marker":"[33]"},{"why":"Provides the antichiral spin order and soft-mode characterization of Mn3Ge used for the Hamiltonian.","marker":"[37]"},{"why":"Supplies the magnetic interaction parameters for Mn3Ge used in the atomistic simulations.","marker":"[38]"},{"why":"Gives the spin current penetration depth value $\\lambda_{\\mathrm{STT}} = 1$ nm used in the pulse profile.","marker":"[40]"}],"fun_headline_variants":["Mn3Ge THz magnons: thickness tunes modes, spin current switches","Picosecond switching and THz magnons in Mn3Ge via spin current","Thickness-dependent THz spin waves and spin-current switching in Mn3Ge","Mn3Ge: THz magnons switch via spin current, thickness tunes spectrum","Spin-current pulse flips Neel vectors and excites THz magnons in Mn3Ge"],"cache_read_input_tokens":12672,"weakest_assumption_plain":"The results rest on the assumed spatiotemporal shape of the spin-current pulse that hits Mn3Ge: a superdiffusive profile with 1 nm penetration depth, full spin polarization (spin Hall angle 1), the stated femtosecond rise and decay times, and a Fe|Au|Mn3Ge stack at zero temperature.","fun_headline_variants_meta":{"raw":{"variants":["Mn3Ge THz magnons: thickness tunes modes, spin current switches","Picosecond switching and THz magnons in Mn3Ge via spin current","Thickness-dependent THz spin waves and spin-current switching in Mn3Ge","Mn3Ge: THz magnons switch via spin current, thickness tunes spectrum","Spin-current pulse flips Neel vectors and excites THz magnons in Mn3Ge"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000992,"raw_usage":{"total_tokens":4186,"prompt_tokens":909,"completion_tokens":3277,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":525,"completion_tokens_details":{"reasoning_tokens":3171}},"tokens_in":525,"tokens_out":3277,"duration_ms":20690,"temperature":1.0,"reasoning_tokens":3171,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-10T22:33:54.848430+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A thickness-series experiment on Mn3Ge thin films (e.g., 4, 6, 8, 10 nm) under femtosecond laser excitation should resolve standing spin-wave peaks whose number increases and whose frequencies decrease with thickness; its absence, or a spectrum fixed by the bulk magnon dispersion rather than by $k_n = n\\pi/d$, would falsify the standing-wave claim. The switching claim is testable by measuring the Neel vector (via X-ray magnetic linear dichroism or the anomalous Hall signal) after single femtosecond pulses and checking that switching occurs only in the predicted $(j_0, \\tau_2)$ regions.","supporting_citations":[{"cited_title":"Bergeard, M","cited_arxiv_id":null,"evidence_quote":"Supplies the superdiffusive spin transport mechanism that converts the femtosecond laser pulse into a spin current."},{"cited_title":"Bal´ aˇ z, M.ˇZonda, K","cited_arxiv_id":null,"evidence_quote":"Provides the experimental and theoretical basis for nanoscale interface confinement of ultrafast spin transfer torque driving non-uniform spin dynamics."},{"cited_title":"Razdolski, A","cited_arxiv_id":null,"evidence_quote":"Establishes the micromagnetic approach for ultrafast magnon generation by femtosecond spin current pulses, which the standing spin wave picture extends."},{"cited_title":"Chirac, J.-Y","cited_arxiv_id":null,"evidence_quote":"Demonstrates Neel vector switching and terahertz spin-wave excitation in Mn2Au, the collinear baseline this paper compares against."},{"cited_title":"Shukla and S","cited_arxiv_id":null,"evidence_quote":"Shows deterministic electrical switching of a non-collinear antiferromagnet (Mn3Sn), the switching precedent this work adapts to Mn3Ge with a laser-driven spin current."},{"cited_title":"Dasgupta, Tuning the transport properties of Mn 3Ge through the effect of strain on its magnetism, Phys","cited_arxiv_id":null,"evidence_quote":"Provides the antichiral spin order and soft-mode characterization of Mn3Ge used for the Hamiltonian."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the magnetic interaction parameters for Mn3Ge used in the atomistic simulations."},{"cited_title":"Alekhin, I","cited_arxiv_id":null,"evidence_quote":"Gives the spin current penetration depth value $\\lambda_{\\mathrm{STT}} = 1$ nm used in the pulse profile."}],"review_version":1}