{"id":"5a08b416-11fe-4ab1-b707-bcacd36dd2dd","arxiv_id":"2507.19665","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":8,"one_line_summary":"Current-induced magnetization switching in V-doped topological insulator films is thermally activated by Joule heating, with stretched-exponential reversal of independent magnetic domains.","lead":"Researchers measured how fast a quantum anomalous Hall device flips its magnetization when hit by short voltage pulses, and found the flip is driven by heat from the pulse rather than by spin torques. The result points to a thermal knob for controlling chiral edge-state direction in topological materials.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The thermal-activation claim rests on an unmeasured electron temperature: α is extrapolated from DC sub-Kelvin thermometry to Te>10 K pulses, and the assumed Pdiss∝V^2 ignores the strong Rxx(Te) variation; direct in-pulse Te measurement would settle it.","rationale":"The paper is a careful experimental study with strong supporting controls: the collapse of Rxy(t) over four decades of pulse width, the stretched-exponential form, the polarity independence, and the absence of in-plane-field switching are all consistent with a Joule-heating-driven reversal and argue against spin-transfer torques as the dominant mechanism in these samples. The reader's conditional verdict is appropriate. My concern sharpens the reader's weakest assumption rather than replacing it: the missing piece is not merely uniformity of Te or the low-to-high-temperature extrapolation of α, but the fact that the conversion from applied voltage to Te assumes a constant resistance, while the paper's own Rxx thermometry shows a strongly temperature/bias-dependent resistance. This does not invalidate the qualitative conclusion, but it means the quantitative Arrhenius parameters and the claimed Te>10 K are model-dependent. A direct in-pulse measurement of Te—or at least a measurement of the pulse current waveform—would discriminate between a genuine thermal-activation mechanism and any other mechanism yielding an exponential dependence on an inverse power of V. Since the reviewer already flagged the temperature inference as the main weakness and placed the verdict at CONDITIONAL, my read does not move the verdict; it does, however, identify a specific experimental check that should be reported before the quantitative thermal-activation law can be considered fully established.","tokens_in":12283,"tokens_out":10583,"duration_ms":145611,"concrete_test":"Perform a synchronized pump-probe measurement: apply the switching pulse and, with a variable sub-pulse delay, probe Rxx with a weak fast pulse (or fast sampling of the current/voltage) to directly measure Te(t) during the 25 ns–100 µs excitation. Then compare the measured Te(t) with the model Te=(Tp^α+P/Σ)^(1/α) using the DC-calibrated α and the simultaneously recorded pulse current and sample voltage. If the measured Te deviates from the model by more than the stated uncertainties, the thermal-activation scaling is unsupported. If direct Te(t) is impractical, record the current waveform during the pulse and test whether tsw is a single-valued function of the time-integrated Joule energy rather than of the generator voltage V alone.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The load-bearing step is the inference of a high electron temperature Te from the pulse voltage. The model sets Pdiss∝V^2 with an effectively constant resistance and then uses α obtained from DC self-heating thermometry (Table S3: α=2.7±0.2 for Sample A, 3.8±0.2 for Sample B) at Te≲0.8 K to claim Te∝V^(2/α) and Te>10 K under pulses. Two things are unverified. First, the DC calibration measures only Rxx(Te) and Rxx(I_DC) at sub-Kelvin temperatures; it does not measure the electron-phonon cooling exponent at Te>10 K, where hot-phonon effects, high-field transport, and activated bulk conduction can all change the effective α. Second, the paper's own data show Rxx changing by orders of magnitude with bias and temperature, so the dissipated power is V^2/R(Te), not simply ∝V^2. If R(Te) is power-law or activated, the true scaling Te∝V^(2/(α+m)) (or more complex) alters the fitted exponent and the extracted A(H) values. Because t0 and A(H) are free parameters in each tsw(V) fit, the Arrhenius collapse in Fig. 4b is partly tautological: rescaling by the fitted A(H)/A(H0) forces the different H-sets onto one curve and does not independently verify the H-dependence of the barrier. The data are consistent with thermal activation, but the central quantitative link—that tsw is controlled by a known Te above 10 K—is not uniquely established; an exponential dependence on an inverse power of V would fit many non-thermal mechanisms as well.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper reports time-resolved measurements of current-induced magnetization switching in 8 nm V-doped (Bi,Sb)2Te3 quantum anomalous Hall devices. Repeated voltage pulses progressively reverse the Hall resistance with a stretched-exponential time dependence, and the extracted switching time tsw varies by about nine orders of magnitude with pulse amplitude and by about four orders with magnetic field. The authors propose that Joule heating raises the electron temperature Te, with Te proportional to V^(2/alpha) from an assumed electron-phonon cooling law, and that switching is thermally activated with tsw = t0 exp[A(H)/V^(2/alpha)]. The field dependence is modeled as E(H) = E0(1 - H/Ha)^n. They further argue that polarity independence and the absence of in-plane-field switching rule out spin-transfer torque as the dominant mechanism. The central claim is that reversal is thermally activated by Joule heating in a disordered magnetic landscape.","tokens_in":12740,"tokens_out":3880,"duration_ms":41004,"significance":"If the thermal-activation interpretation is correct, the paper gives an important dynamical fingerprint of the disordered magnetic landscape in QAH systems and suggests a route to heat-mediated control of chiral edge states. The strengths include the wide dynamic range of the measurements, the stroboscopic protocol validated by pulse-width and waiting-time independence, the polarity and field controls that disfavor spin-transfer torque, and the availability of supporting data and code from Zenodo. The principal weakness is that the quantitative temperature-voltage relation is inferred from a model rather than measured: the electron temperature during pulses is not directly probed, and the manuscript's own supplementary text acknowledges strong approximations. The claim is therefore plausible and well-supported phenomenologically, but the central quantitative link still needs strengthening.","major_comments":[{"comment":"The quantitative link between pulse voltage and electron temperature is assumed rather than measured. The model sets Pdiss proportional to V^2 and Pout = Sigma(Te^alpha - Tp^alpha), with alpha calibrated from DC sub-Kelvin thermometry (Table S3), and then extrapolates to pulsed conditions where Te exceeds 10 K. Since the paper's own data (Fig. S7) show Rxx varying by orders of magnitude with applied power, the dissipated power should be V^2/R(Te), and any temperature dependence of R or a change in the cooling exponent at high Te modifies the predicted scaling Te proportional to V^(2/alpha). This is load-bearing because the extracted A(H) values and the thermal-activation interpretation rely on that scaling. Direct in-pulse determination of Te, or at least a demonstration that the DC-calibrated alpha and the constant-resistance approximation hold for Te > 10 K, is needed.","section":"Model for heat-induced reversal; Eq. tsw = t0 exp(A/V^(2/alpha))"},{"comment":"The collapse in Fig. 4b rescales tsw/t0 by the fitted ratio A(H)/A(H0), where t0 and A(H) are free parameters of the same Arrhenius fits to tsw(V). Therefore the collapse is a consistency check of the assumed functional form rather than an independent verification of the field dependence of the barrier. The field-dependence analysis similarly fixes n = 3 and fits H0 from the same tsw(H) data, with n = 2-4 also acceptable. As a result, the data are consistent with thermal activation but do not uniquely establish it; an exponential dependence on an inverse power of V would also fit the limited voltage range. Please state explicitly which predictions are falsifiable and compare the thermal-activation model quantitatively with non-thermal alternatives.","section":"Experimental observations; Fig. 4b"},{"comment":"The attempt time t0 is fitted separately in the voltage-dependence and field-dependence analyses, yielding values that differ by an order of magnitude for the same sample and comparable conditions: for Sample B at 0.6 T, Table S1 gives t0 = 1.2 ns, while Table S2 at 6 V gives t0 = 11.5 ns. Since t0 is a physical attempt time, it should be common across both dependencies for a given sample and field. The discrepancy weakens the quantitative consistency of the Arrhenius model and should either be resolved by a global fit or justified explicitly.","section":"Tables S1 and S2"}],"minor_comments":[{"comment":"Page 2 states that transport measurements were 'performed 20 mK'; the word 'at' is missing before the temperature.","section":"Experimental setup"},{"comment":"Page 2 contains a duplicated phrase: 'we focus on this article in two main devices' should read 'we focus on two main devices'.","section":"Experimental setup"},{"comment":"The horizontal axis label in Fig. 3d appears as '0H' rather than 'mu0 H'; please correct the typographical rendering.","section":"Fig. 3d"},{"comment":"The stretched-exponential fits for Sample B deviate systematically from the data on short timescales. Please provide fit-quality measures or a more detailed discussion of why tsw remains correctly extracted despite these deviations.","section":"Supplementary, Fig. S4"},{"comment":"The statement that Rxy is proportional to M 'following the law of the classical anomalous Hall effect' is a simplifying assumption; please add a reference or caveat for its validity in a strongly disordered QAH system.","section":"Magnetization dynamics"},{"comment":"Please clarify whether V in the expressions Pdiss proportional to V^2 and A/V^(2/alpha) is the generator output amplitude, the voltage across the sample, or the voltage after line attenuation, since the paper notes that absolute pulse amplitude is not relevant to the analysis.","section":"Model for heat-induced reversal"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is a carefully executed transport study with a plausible thermal-activation scenario, but the central quantitative link between pulse voltage and electron temperature is model-dependent and not directly verified. A revision that provides direct evidence of Te under pulses, or an independent test of the Te(V) scaling, would substantially raise the manuscript's impact. The paper fits the journal's scope; I would be willing to review a revised version."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"First, know that this is a solid experimental paper. The new thing is the time-resolved stroboscopic measurement of current-induced switching in QAH devices: they show the magnetization only evolves under voltage drive, not between pulses, and extract switching times covering nine orders of magnitude in voltage and four in field. The collapse of Rxy(t) for different pulse widths onto a single stretched-exponential curve is clean and convincing. The data and code are on Zenodo, which helps.\n\nThe paper makes two claims. One, that spin-transfer torque is not the dominant mechanism. That is well supported by the polarity-independence and the absence of switching with in-plane fields. Two, that reversal is thermally activated, with Te extracted from a Joule-heating model. This is plausible but not uniquely established. The main soft spot, also the load-bearing one, is the electron temperature. They calibrate the electron-phonon cooling exponent alpha from DC resistance thermometry at sub-Kelvin temperatures, then extrapolate the same power law to Te > 10 K during pulses, where hot-phonon effects and activated bulk conduction could change alpha. They also take Pdiss ∝ V^2 with an effectively constant resistance, while their own data show Rxx changing by orders of magnitude with bias and temperature. If R(Te) varies as a power law, the voltage exponent in the Arrhenius law changes, and the fitted A(H) values shift. Because t0 and A(H) are free parameters, the collapse in Fig. 4b is a consistency check rather than an independent prediction.\n\nNone of this kills the paper. The qualitative conclusion—heating drives the switching, not torques—rests on controls that don't depend on the alpha extrapolation. The Arrhenius-like voltage dependence over 9 decades is a strong empirical observation, and the stretched-exponential dynamics is a reasonable interpretation of a disordered magnetic landscape. The field dependence fit fixes n=3 and has large uncertainties, but they are transparent about that. The paper would benefit from direct in-pulse Te measurement or a measurement of alpha at higher temperatures, but that is a request for future work, not a fatal flaw.\n\nWho is this for? Experimentalists working on QAH, chiral edge states, or current-induced switching in topological materials. It deserves a serious referee; I'd send it to review with a recommendation for revision, mainly to strengthen the Te calibration discussion and tone down the quantitative claims.","headline":"Careful time-resolved study with a plausible thermal-activation mechanism; the quantitative Te model is underdetermined, but the qualitative case against spin-transfer torque is solid.","tokens_in":13258,"tokens_out":3101,"would_cite":true,"duration_ms":34949,"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":"Current-induced magnetization reversal in V-doped (Bi,Sb)2Te3 quantum anomalous Hall devices is thermally activated by Joule heating, follows an Arrhenius law over nine orders of magnitude, and is not caused by spin-transfer torques.","keywords":["quantum anomalous Hall effect","magnetization switching","Joule heating","thermal activation","Arrhenius law","stretched exponential","V-doped (Bi,Sb)2Te3","chiral edge states"],"falsifier":"Measure the electron temperature directly during the same 25 ns pulses used for switching (e.g., by noise thermometry or a second calibrated mesa) and check whether $T_e$ scales as $V^{2/\\alpha}$ and reaches the value required by the Arrhenius fit; if switching proceeds at a $T_e$ far below that value, or if $t_{\\rm sw}$ changes when the substrate thermal path is altered, the thermal-activation claim is disproven.","tokens_in":12079,"feed_emoji":"🔥","tokens_out":9861,"duration_ms":97541,"temperature":0.7,"pith_summary":"This paper reports time-resolved measurements of current-induced magnetization switching in 8 nm films of V-doped (Bi,Sb)2Te3 in the quantum anomalous Hall regime, a topological phase whose conduction is carried by a single chiral edge state tied to the magnetization direction. It argues that the reversal is thermally activated: voltage pulses dissipate Joule heat that raises the electron temperature, exponentially shortening the switching time, and the reversal proceeds by flipping many independent magnetic domains. The signature is an Arrhenius law $t_{\\rm sw} = t_0 \\exp(A/V^{2/\\alpha})$ that holds over nine orders of magnitude in switching time, with $\\alpha$ determined from separate DC thermometry. If correct, this identifies a thermal pathway for flipping the chirality of quantum anomalous Hall edge states and rules out spin-transfer torques in these devices.","feed_headline":"Joule heating, not spin torque, drives QAH switching","feed_subtitle":"Voltage pulses heat electrons and cut switching time by nine orders of magnitude. Thermal control of chiral edge states follows.","key_machinery":"The load-bearing object is the Arrhenius activation law chained to a Joule-heating/electron-phonon cooling balance. A pulse dissipates power $P_{\\rm diss} \\propto V^2$; the electrons equilibrate at temperature $T_e$; cooling to the lattice follows $P_{\\rm out} = \\Sigma(T_e^\\alpha - T_p^\\alpha)$; equating the two gives $T_e \\propto V^{2/\\alpha}$ at high drive. Inserting this into $t_{\\rm sw} = t_0 \\exp(E(H)/(k_B T_e))$ yields the fitted form $t_{\\rm sw} = t_0 \\exp(A/V^{2/\\alpha})$. The stretched-exponential response models the reversal as the sum of many independent domain flips with a broad distribution of switching times.","core_discovery":"The core claim is that in these V-BST quantum anomalous Hall devices, current-induced magnetization reversal is driven by Joule heating, not by spin-transfer torques. Under a small opposing magnetic field, a voltage pulse heats the electron system above 10 K, exponentially reducing the switching time of individual magnetic domains according to $t_{\\rm sw} = t_0 \\exp(E(H)/(k_B T_e))$ with $T_e \\propto V^{2/\\alpha}$. The reversal is not a single coherent flip: the Hall resistance relaxes as a stretched exponential $R_{xy}(t) = R_0(1-2\\exp(-(t/t_{\\rm sw})^\\beta))$ with $\\beta$ as low as 0.4, indicating many independent switching regions. The same Arrhenius form reproduces the magnetic-field dependence of $t_{\\rm sw}$, and the absence of switching under in-plane fields plus the independence of pulse polarity are presented as ruling out spin-transfer torques as the dominant mechanism.","pith_inferences":["If the thermal picture is right, devices on better heat-sinking substrates should switch more slowly at the same pulse amplitude; measuring $t_{\\rm sw}(V)$ across substrates with different thermal conductivity would test this directly.","The same model may explain why some earlier QAH switching experiments saw a sharp voltage threshold: the threshold could be where Joule heating pushes $T_e$ over $E(H)/k_B$, rather than where spin torque overcomes damping.","Tracking the stretching exponent $\\beta$ as a function of $t_{\\rm sw}$ could map the distribution of domain sizes, connecting the fitted exponent to the actual magnetic puddle landscape."],"forward_implications":["Switching times can be predicted from the pulse amplitude alone once $\\alpha$ and the barrier scale $A(H)$ are known, without invoking spin-torque physics.","The stroboscopic resistance technique can probe the onset of percolative transport through magnetic puddles during partial reversal.","Thermal, rather than electrical, control of the magnetization implies that local heating—by a focused laser or a heated gate—should reverse the edge-state direction in a confined region.","The extracted anisotropy parameters ($n \\simeq 3$, $H_a \\simeq 1\\text{–}1.6$ T) provide quantitative inputs for designing QAH-based reconfigurable circuits."],"supporting_citations":[{"why":"Reports spin-orbit-torque switching in Cr-doped BST; the polarity and in-plane-field results here are incompatible with that mechanism.","marker":"[16]"},{"why":"Earlier observation of current-induced switching under conditions not explained by spin torque; the current range used here matches this study.","marker":"[17]"},{"why":"Provides the stretched-exponential relaxation form used to fit the switching dynamics.","marker":"[20]"},{"why":"Supports interpreting stretched-exponential dynamics as independent disordered magnetic domains.","marker":"[21]"},{"why":"Documents strong magnetic disorder in the parent compound Cr-doped BST, grounding the multi-domain assumption.","marker":"[23]"},{"why":"Establishes the breakdown threshold of the QAH effect, the regime in which Joule heating becomes significant.","marker":"[24]"},{"why":"Supplies the electron-phonon cooling power law and the calibration of the exponent alpha used in the Arrhenius fit.","marker":"[25]"},{"why":"Benchmark electron-phonon cooling power law in low-temperature nanodevices, supporting the power-law form used in the model.","marker":"[26]"}],"fun_headline_variants":["QAH switching: Joule heating, not spin torque","Heat flips QAH edge states, not spin torque","Joule heating controls QAH magnetization switching","Thermal reversal of chiral edge states in QAH","Switching QAH devices by Joule heating only"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The model assumes that the whole electron system heats uniformly to a single temperature $T_e$ and that the electron-phonon cooling exponent $\\alpha$, calibrated from DC resistance thermometry below about 0.6 K, still holds when pulses push $T_e$ above 10 K; if temperature is non-uniform or the cooling law changes under strong bias, the Arrhenius reading of $t_{\\rm sw}(V)$ would lose its quantitative foundation.","fun_headline_variants_meta":{"raw":{"variants":["QAH switching: Joule heating, not spin torque","Heat flips QAH edge states, not spin torque","Joule heating controls QAH magnetization switching","Thermal reversal of chiral edge states in QAH","Switching QAH devices by Joule heating only"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000208,"raw_usage":{"total_tokens":1353,"prompt_tokens":842,"completion_tokens":511,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":458,"completion_tokens_details":{"reasoning_tokens":437}},"tokens_in":458,"tokens_out":511,"duration_ms":5436,"temperature":1.0,"reasoning_tokens":437,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T14:10:05.138384+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the electron temperature directly during the same 25 ns pulses used for switching (e.g., by noise thermometry or a second calibrated mesa) and check whether $T_e$ scales as $V^{2/\\alpha}$ and reaches the value required by the Arrhenius fit; if switching proceeds at a $T_e$ far below that value, or if $t_{\\rm sw}$ changes when the substrate thermal path is altered, the thermal-activation claim is disproven.","supporting_citations":[{"cited_title":"Yuan , author L","cited_arxiv_id":null,"evidence_quote":"Reports spin-orbit-torque switching in Cr-doped BST; the polarity and in-plane-field results here are incompatible with that mechanism."},{"cited_title":"Zimmermann , author J","cited_arxiv_id":null,"evidence_quote":"Earlier observation of current-induced switching under conditions not explained by spin torque; the current range used here matches this study."},{"cited_title":"Xi , author K.-Z","cited_arxiv_id":null,"evidence_quote":"Supports interpreting stretched-exponential dynamics as independent disordered magnetic domains."},{"cited_title":"Lee , author C","cited_arxiv_id":null,"evidence_quote":"Documents strong magnetic disorder in the parent compound Cr-doped BST, grounding the multi-domain assumption."},{"cited_title":"Breakdown of the quantum anomalous Hall effect under microwave drives","cited_arxiv_id":"2505.23156","evidence_quote":"Supplies the electron-phonon cooling power law and the calibration of the exponent alpha used in the Arrhenius fit."},{"cited_title":"Jezouin , author F","cited_arxiv_id":null,"evidence_quote":"Benchmark electron-phonon cooling power law in low-temperature nanodevices, supporting the power-law form used in the model."}],"review_version":1}