{"id":"3bf6f2cc-057c-4ec4-b9f2-092b30f30acb","arxiv_id":"2505.07976","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"Buzdin step widths in a φ0 Josephson junction are shown to equal a product of two Bessel functions, leaving the critical current unchanged.","lead":"This paper studies a special superconducting junction called a φ0 junction when only the magnetic part of a microwave signal drives it. The authors show that the resulting 'Buzdin steps' have widths set by a product of two Bessel functions, which makes them clearly different from ordinary Shapiro steps.","discovery_kind":"first_principles","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The analytic width formula (Eq. 12) depends on discarding the homogeneous LLG solution with a single initial condition; near resonance that mode contributes a J0(h_R/ω_R) term to the step width, so Eq. 12 is not justified unless transients are shown to vanish.","rationale":"The central new result is the analytic Bessel-product formula for the Buzdin step width. The numerical agreement in Fig. 3 is real supporting evidence, and much of the algebra leading to Eq. (12) is internally consistent. But the analytic derivation obtains the clean J_2 J_0 product only after dropping homogeneous solutions of the undamped linearized LLG on the basis of a single initial condition. That step is not innocuous: at the resonance used in the paper, the homogeneous oscillation combines with cos(ω_J t + φ_0) to produce a dc term through the J_0(h_R/ω_R) component of the phase-modulated exponential. Thus, unless damping is explicitly included and shown to eliminate the homogeneous mode in steady state, Eq. (12) is an incomplete first-order result. The reader's weakest assumption points at the same C_n = 0 issue; the β_c inconsistency compounds the problem by making it unclear whether the numerical confirmation was obtained in the regime where the perturbation theory is nominally valid. These issues warrant keeping the conditional verdict: the paper is worth publishing if the authors provide a damped steady-state calculation confirming that the homogeneous contribution vanishes and clarify the β_c discrepancy.","tokens_in":15025,"tokens_out":21705,"duration_ms":228386,"concrete_test":"Re-derive Eq. (F3) from Eq. (B9) without setting C_n = 0: keep a general homogeneous amplitude H(t) = Σ C_n e^{i(ω_F t - h_R/ω_R cos ω_R t)}, fix C_n from the two physical initial conditions m_y(0) = 0 and m_x(0) = 0, and evaluate the time average j^{(1)} = ⟨r m_y cos(ω_J t + φ_0)⟩ at ω_J = ω_R ≈ ω_F. If the resulting H-dependent term (proportional to r Re[Σ C_n e^{-iφ_0}] J_0(h_R/ω_R)) is nonzero, then Eq. (12) omits a contribution of the same order and the Bessel-product formula is not the complete first-order step width; alternatively, the same check can be done numerically by integrating (B2a)-(B2b) with α = 0.01 to steady state.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The load-bearing step is in Appendix B, Eq. (B9): the homogeneous solution of the linearized LLG, C_n e^{i(ω_F t - (h_R/ω_R) cos ω_R t)}, is discarded by saying that the single initial condition m_y = 0 gives C_n = 0. One scalar constraint cannot fix the infinite set C_n, and the discarded mode does not average to zero in the object that defines the step width. Computing j^{(1)} = ⟨r m_y cos(ω_J t + φ_0)⟩ at the resonance ω_J = ω_R ≈ ω_F, the homogeneous piece contributes (r/2) Re[Σ C_n e^{-iφ_0}] J_0(h_R/ω_R), because e^{i(ω_F-ω_J)t} e^{-i(h_R/ω_R) cos ω_R t} has a nonzero time average. This term is of the same order as the retained J_2 J_0 term in Eq. (12) whenever the homogeneous amplitude is comparable to the particular solution, so Eq. (12) is not the complete first-order width. The original LLG has damping α, so in numerical steady state C_n likely decays; but the analytic derivation removes α (Eq. 7) and therefore has no mechanism enforcing C_n = 0. A secondary but related issue is the parameter inconsistency (β_c = 25 in Sec. II vs β_c = 2 in Fig. 2 and the 'same parameters' comparisons), which makes the stated HPT validity condition β_c I ≫ 1 marginal at the simulated currents.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies a current-biased φ0 superconductor-ferromagnet-superconductor Josephson junction driven by the magnetic component of external radiation. Numerically, the authors find that locking of the Josephson oscillations with the magnetic component produces Buzdin steps whose width oscillates with the microwave amplitude in a way qualitatively different from Shapiro steps, and that the critical current remains unchanged. Analytically, using a harmonic perturbation theory in the large-capacitance limit, they derive the step width for the main resonance ω_J=ω_R≈ω_F as a product of two Bessel functions, Eq. (12), and a corresponding expression for the subharmonic ω_J=ω_R≈ω_F/2, Eq. (13). They also report a sudden drop in the magnetization-precession amplitude at locking, attributed to destructive interference, and magnetization reorientation. The analytic formulas are compared with direct numerical integration of the coupled LLG-RCSJ equations.","tokens_in":15468,"tokens_out":6493,"duration_ms":70369,"significance":"If the derivation is correct, the paper provides a useful analytic description of Buzdin steps in φ0 junctions, distinguishing magnetic locking from conventional Shapiro locking and giving a falsifiable prediction for the step-width oscillations. The result is not fitted to numerics; it follows from a perturbative treatment of the model equations, which is a strength. The claim that the critical current is unaffected by the magnetic drive, and the identification of destructive interference in the magnetic subsystem, are also of interest for experiments on SFS junctions. However, the analytic derivation has a load-bearing gap concerning the homogeneous solutions of the linearized LLG equation, and the parameter regimes stated in the text are inconsistent with those used in the numerical comparisons. These issues must be resolved before the central formula can be considered established.","major_comments":[{"comment":"The homogeneous solution of the linearized LLG equation is discarded incorrectly. The text states that substituting the initial condition m_y=0 gives C_n=0 for all n, but a single scalar initial condition cannot determine an infinite set of independent coefficients; more fundamentally, the homogeneous solution of Eq. (B3) is one free constant (or two real constants) before the Jacobi-Anger expansion, so the claim is not mathematically justified. This matters because, at the resonance ω_J=ω_R≈ω_F, the homogeneous term m_y^h(t)=Re[C exp(iω_F t - i(h_R/ω_R)cos ω_R t)] contributes a nonzero time average to r m_y(t) cos(ω_J t+φ_0), proportional to r J_0(h_R/ω_R) Re[C e^{-iφ_0}]. This is of the same order as the retained J_2 J_0 term in Eq. (12) unless a physical mechanism suppresses C. Since Eq. (7) contains no damping, no such mechanism is present in the analytic calculation. The authors should either include α in the linearized equations and take the steady-state limit, or explicitly show that homogeneous terms are transients whose contribution to the time-averaged step width vanishes in that limit. This point is load-bearing for Eq. (12).","section":"Appendix B, Eq. (B9)"},{"comment":"There is a direct inconsistency in the parameter values: Sec. II states that β_c=25 is used, while the caption of Fig. 2 and the subsequent numerical figures use β_c=2. The harmonic perturbation theory is justified by the condition β_c I ≫ 1 stated in Sec. IV and Appendix A. For the bias currents near the steps in the figures, β_c I is of order 1, not much larger than 1, so the nominal validity condition is not met by the parameters actually simulated. The authors should use consistent parameters or explain why the HPT result remains applicable when β_c I is not large.","section":"Sec. II and Fig. 2"},{"comment":"For the subharmonic case ω_J=ω_R≈ω_F/2, the numerical step width deviates substantially from the analytic Bessel-product formula at low h_R, as shown in Fig. 4. The manuscript acknowledges the discrepancy with the word 'anomalies' but does not analyze its origin. Since the paper claims that the Buzdin-step width generally represents a product of two Bessel functions, the unexplained failure at the subharmonic should be addressed, for example by checking whether higher harmonics, finite damping, or the discarded homogeneous modes are responsible. Without this, the generality of the central claim is limited.","section":"Sec. IV, Eq. (13) and Fig. 4"}],"minor_comments":[{"comment":"The paragraph describing the paper's organization says the model is introduced in Sec. III and the peculiarities of Buzdin steps are studied in Sec. III, but the model is in Sec. II and the Buzdin steps are analyzed in Sec. III; the section references should be corrected.","section":"Sec. II (organization paragraph)"},{"comment":"In Eq. (13) the argument of J_3 and J_1 is written as H_r/ω_R, whereas everywhere else the amplitude is denoted h_R; this notation should be made uniform.","section":"Eq. (13) and Appendix G"},{"comment":"The sentence justifying the neglect of r m_y says 'since r<1 then r m_y is much smaller than m_y'; the logic is circular, since r m_y is smaller than m_y precisely by the factor r<1, not by an additional smallness. This should be rephrased.","section":"Sec. IV, text near Eq. (7)"},{"comment":"The phrase 'Gilbert dumping' should be 'Gilbert damping'.","section":"Sec. IV, paragraph on Gilbert damping"},{"comment":"The argument that the zero-voltage state receives no magnetic contribution relies on the real part of i^{-1} times certain complex combinations vanishing; the presentation skips the intermediate algebra and would benefit from showing explicitly why the q=0 terms cancel or have vanishing real part.","section":"Appendix E"}],"recommendation":"major_revision","confidential_remarks":"The central formula is plausible and the numerical comparison is suggestive, but the handling of the homogeneous LLG solution is a genuine flaw in the derivation as written. The parameter inconsistency between β_c=25 in the model section and β_c=2 in the figures also needs to be resolved. I believe the paper can be made publishable with a careful revision, but the current version should not be accepted as is."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Quick take: the core claim, that the width of the first Buzdin step in a φ0 junction driven by the magnetic component is a product of two Bessel functions, is new, analytically derived, and matches numerics in the regime where the approximation should work. The derivation is not fitted, and the destructive-interference picture in Sec. V is a nice addition. I think the result is basically right.\n\nWhat’s genuinely new: the Bessel-product formulas (Eqs. 12 and 13), the explicit identification of the q,n resonance pairs, and the observation that magnetic locking leaves the critical current constant. The authors cite the prior Buzdin-step work fairly, and the comparison with Refs. 8 and 24 is appropriate.\n\nSoft spots, in proportion:\n\n- The βc inconsistency is real and should be fixed: Sec. II states βc = 25, but Fig. 2 (and the “same parameters” comparisons) use βc = 2. The HPT derivation assumes βc I ≫ 1; at the plotted currents βc I ~ 1, so the numerical agreement is better than the approximations warrant. This needs a sentence or two of justification or corrected parameters.\n\n- The stress-test concern about the homogeneous solution in Appendix B does not land. The particular solution constructed with lower limit t0 has zero initial data for m_y (and its derivative, if m_x(0)=0), so the homogeneous coefficient is zero; the paper’s statement is terse but defensible. What would be worth adding is an explicit statement that both m_y(0)=0 and m_x(0)=0, plus a comment on how transients are handled in the numerics.\n\n- Fig. 4: at low h_R, the ω_F/2 case deviates from the analytic curve. The authors mention this but do not explain it. Minor.\n\n- No code or data release. Not fatal for a theory paper, but it would help reproducibility.\n\nWho it’s for: people working on anomalous Josephson junctions, superconducting spintronics, or analytic treatments of driven LLG+RCSJ systems. This is a solid subfield contribution, not a paradigm shift.\n\nRecommendation: send it to peer review. The authors should be asked to fix the βc inconsistency, clarify the initial conditions in Appendix B, and discuss the range of validity of HPT. With those changes, the analytic result is publishable.","headline":"A solid analytic result for magnetically driven Buzdin steps, worth refereeing after the βc inconsistency is fixed; the stress-test worry about the homogeneous solution does not hold up.","tokens_in":15964,"tokens_out":4237,"would_cite":true,"duration_ms":42262,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"In a φ0 Josephson junction, magnetic-component locking produces Buzdin steps whose width is a product of two Bessel functions, unlike electric Shapiro steps.","keywords":["φ0 Josephson junction","Buzdin steps","Shapiro steps","magnetic component","Landau-Lifshitz-Gilbert equations","RCSJ model","Bessel functions","destructive interference"],"falsifier":"Solve the full, unlinearized LLG-RCSJ system at the parameters of the paper's comparison but with $G$ raised by an order of magnitude, and check whether the numerically observed Buzdin step width still vanishes at the zeros of $J_0(h_R/\\omega_R)$ and follows Eq. (12) at small $h_R$; alternatively, keep the homogeneous coefficients $C_n$ in Eq. (B9) instead of setting them to zero and test whether the step width changes. Any non-negligible change in either test would show that the two-Bessel claim depends on the small-oscillation assumption.","tokens_in":14881,"feed_emoji":"🧲","tokens_out":9751,"duration_ms":86934,"temperature":0.7,"pith_summary":"Superconductor-ferromagnet-superconductor Josephson junctions of the $\\varphi_0$ type have a phase shift controlled by the magnetic moment, so microwaves can act on the junction through the magnetic field as well as through the electric field. This paper argues that locking with the magnetic component is a distinct mechanism: it produces Buzdin steps whose width at resonance is a product of two Bessel functions, $J_2(h_R/\\omega_R)J_0(h_R/\\omega_R)$, rather than the single-Bessel, Bessel-like oscillation familiar from Shapiro steps. The same mechanism leaves the critical current unchanged and, at the moment of locking, abruptly suppresses the magnetization-precession amplitude through destructive interference while reorienting the magnetization. These features give an experimentally readable fingerprint that separates magnetic locking from electric Shapiro locking in hybrid superconducting devices.","feed_headline":"Buzdin step width follows a two-Bessel function rule","feed_subtitle":"Magnetic-component locking gives a step width J₂J₀ and leaves the critical current fixed.","key_machinery":"The load-bearing object is the analytic solution for the $y$-component of the magnetization, $m_y(t)$, obtained from the linearized LLG equations under $m_x,m_y\\ll 1$, $m_z\\approx 1$, weak Gilbert damping, and small Josephson-to-magnetic energy ratio $G\\ll 1$. This solution, Eq. (8), is a double Bessel-function series; inserting it into the RCSJ supercurrent term $r m_y\\cos(\\omega_J t+\\phi_0)$ produces a nonvanishing time-averaged current only when the Buzdin condition $q\\omega_R-2\\omega_J=0$ holds, so the steps sit at $\\omega_J=q\\omega_R/2$. Selecting the resonant terms near $\\omega_R\\approx\\omega_F$ yields the $J_2J_0$ product, while the near-half-resonance case yields $J_3J_1$; the same resonance selection also removes the two dominant oscillating terms from $m_y$, which is the mechanism behind the destructive-interference dip.","core_discovery":"Working from the coupled Landau-Lifshitz-Gilbert and resistively-and-capacitively-shunted-junction equations in the weak-coupling limit, the paper derives the width of the first Buzdin step in closed form. For driving near the ferromagnetic resonance, $\\omega_J=\\omega_R\\approx\\omega_F$, the width is $\\Delta j^{(1)} = [r^2 G \\omega_F/(\\omega_R-\\omega_F)]\\,J_2(h_R/\\omega_R)J_0(h_R/\\omega_R)$, and near $\\omega_F/2$ it is $\\Delta j^{(1)} = [r^2 G \\omega_F/(2\\omega_R-\\omega_F)]\\,J_3(h_R/\\omega_R)J_1(h_R/\\omega_R)$. The two-Bessel product is the paper's central explanation for why the amplitude dependence is anomalous: the first maximum is lower than the second, unlike the standard Shapiro Bessel pattern. At zero voltage the averaged magnetic contribution vanishes, so the critical current remains constant as the microwave amplitude is varied; at the lock point, the two dominant resonant terms in $m_y(t)$ become time-independent and drop out of the precession, which the paper reads as destructive interference, accompanied by a magnetization reorientation.","pith_inferences":["The zeros of $J_0(h_R/\\omega_R)$ in the step-width formula provide an amplitude knob at which magnetic locking collapses even though the drive is strong; the paper does not discuss using this as a controlled switch, but the formula implies it.","Because the destructive-interference dip is a magnetic, not a transport, signature, a measurement of ferromagnetic-resonance precession amplitude while sweeping bias current would locate the Buzdin step without contacting the junction's electric leads; this is an experimental extension the paper leaves implicit.","The two-Bessel structure should appear in any junction whose Josephson phase couples linearly to a driven oscillator through a term like $r m_y$, so the fingerprint may be shared by other anomalous Josephson systems; the paper restricts its claim to the $\\varphi_0$ SFS junction in this geometry."],"forward_implications":["If the paper is right, the first Buzdin step at resonance should have two Bessel factors in its amplitude dependence, so the step can be suppressed at amplitudes where $J_2$ alone is still large.","Under magnetic-only drive, the critical current of the $\\varphi_0$ junction should stay flat as the microwave amplitude is swept, in contrast to the correlated zero-step oscillations of electric Shapiro locking.","At the locking condition, the maximum $m_y$ precession amplitude versus bias current should show a sharp dip at $\\omega_J=\\omega_R$, coinciding with the Buzdin step in the current-voltage characteristic.","Near $\\omega_R\\approx\\omega_F/2$, the step width should follow a $J_3J_1$ product at moderate amplitudes, while the low-amplitude regime is where the numerical and analytical results already deviate."],"supporting_citations":[{"why":"Predicts the φ0 phase shift proportional to the magnetic moment and introduces the Buzdin step that this paper explains.","marker":"[8]"},{"why":"Provides the coupled LLG–Josephson equations and the classification of Buzdin, Shapiro, and chimera steps used as the numerical model.","marker":"[24]"},{"why":"Supplies the harmonic perturbation theory in the large-capacitance limit from which the first-order current correction and step width are derived.","marker":"[39–41]"},{"why":"Establishes the Bessel-like amplitude oscillations of Shapiro steps that the Buzdin-step behaviour is contrasted against.","marker":"[28, 31]"},{"why":"Defines Shapiro steps from locking with the electric component, the baseline phenomenon this work distinguishes from magnetic locking.","marker":"[26]"},{"why":"Gives the RCSJ circuit model underlying the junction equation of motion.","marker":"[36]"}],"fun_headline_variants":["Buzdin step width: two Bessel functions, not one","Two-Bessel product sets Buzdin step width","Magnetic locking yields two-Bessel Buzdin steps","Anomalous Buzdin width from two-Bessel product"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The derivation assumes the magnetization stays almost aligned with the easy axis and the perpendicular components $m_x,m_y$ remain tiny, with small coupling $G$, weak damping, and the free homogeneous precession dropped by imposing a single initial condition $m_y=0$; if the homogeneous precession is non-negligible, the Bessel-product width formula and the destructive-interference picture fail.","fun_headline_variants_meta":{"raw":{"variants":["Buzdin step width: two Bessel functions, not one","Two-Bessel product sets Buzdin step width","Magnetic locking yields two-Bessel Buzdin steps","Anomalous Buzdin width from two-Bessel product"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000801,"raw_usage":{"total_tokens":3537,"prompt_tokens":978,"completion_tokens":2559,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":594,"completion_tokens_details":{"reasoning_tokens":2490}},"tokens_in":594,"tokens_out":2559,"duration_ms":18107,"temperature":1.0,"reasoning_tokens":2490,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T22:06:12.598427+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Solve the full, unlinearized LLG-RCSJ system at the parameters of the paper's comparison but with $G$ raised by an order of magnitude, and check whether the numerically observed Buzdin step width still vanishes at the zeros of $J_0(h_R/\\omega_R)$ and follows Eq. (12) at small $h_R$; alternatively, keep the homogeneous coefficients $C_n$ in Eq. (B9) instead of setting them to zero and test whether the step width changes. Any non-negligible change in either test would show that the two-Bessel claim depends on the small-oscillation assumption.","supporting_citations":[{"cited_title":"Cryogenic mem- ory element based on an anomalous josephson junction","cited_arxiv_id":null,"evidence_quote":"Predicts the φ0 phase shift proportional to the magnetic moment and introduces the Buzdin step that this paper explains."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the coupled LLG–Josephson equations and the classification of Buzdin, Shapiro, and chimera steps used as the numerical model."},{"cited_title":"Magnetoelec- tric effects in josephson junctions.Journal of Physics: Condensed Matter, 34(35):353001, 2022","cited_arxiv_id":null,"evidence_quote":"Defines Shapiro steps from locking with the electric component, the baseline phenomenon this work distinguishes from magnetic locking."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Gives the RCSJ circuit model underlying the junction equation of motion."}],"review_version":1}