Pith. sign in

REVIEW 4 major objections 5 minor 21 references

The robustness of composite pulses elucidated by classical mechanics: Stability around the globe

T0 review · 4 major / 5 minor · reviewed 2026-08-06 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read This paper proposes that the robustness of composite pulses is a classical caustic: in canonical coordinates on the Bloch sphere, the ensemble focus at the final time appears as the collapse of averaged stability-matrix elements.

desk verdict A useful directional stability diagnostic for composite pulses, but the central caustic claim is asserted rather than demonstrated. read the letter →

arxiv 2507.01364 v1 pith:IB63CBAN submitted 2025-07-02 quant-ph physics.atom-phphysics.class-ph

Jonathan Berkheim , David J. Tannor This is my paper · ORCID
classification quant-phphysics.atom-phphysics.class-ph
keywords compositepulsesBlochspherestabilitymatrixcausticscanonicalcoordinatesresonanceoffsetfieldinhomogeneityquantumcontrol
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper proposes a classical-mechanics explanation for why composite pulses—short sequences of resonant pulses that invert spin populations despite field imperfections—are robust. It maps two-level evolution on the Bloch sphere into the canonical coordinates $\phi$ and $\eta=\cos\theta$, where an ensemble of slightly detuned or mis-calibrated spins behaves like a classical phase-space swarm. In this picture, focusing of the ensemble at the final time is a caustic: the averaged stability-matrix elements collapse into a narrow band. For the $90(x)180(y)90(x)$ pulse, the collapse is direction-selective—field inhomogeneity refocuses the polar coordinate while spreading the azimuth, whereas resonance offset refocuses both. If the conjecture linking stability-element collapse to refocusing holds, the framework gives a visual, quantitative way to understand and design robust pulses.

What carries the argument

The machinery is the stability matrix in canonical spherical coordinates on the Bloch sphere. The change of variables from the Cartesian Bloch vector to $(\phi,\eta)$ with $\eta=\cos\theta$ turns the non-canonical Lie–Poisson dynamics into Hamilton's equations with Poisson bracket $\{\phi,\eta\}=1$. The stability matrix $M_s$ has elements $\partial\phi_f/\partial\phi_i$, $\partial\phi_f/\partial\eta_i$, $\partial\eta_f/\partial\phi_i$, and $\partial\eta_f/\partial\eta_i$; a caustic in the two-dimensional phase space is the vanishing of an appropriate element such as $\partial\phi_f/\partial\eta_i$, signalling that many initial conditions reach the same final coordinate. Because each trajectory in a composite-pulse ensemble runs under a slightly different Hamiltonian, the paper averages the elements over the ensemble and tracks the width of their histograms via the range parameter $h_\zeta(t_f)$. The Jacobian relation $M_c = J_f M_s J_i^{-1}$ connects the spherical and Cartesian stability matrices, and inserting this into the traditional expression $W\equiv\partial\mathbf{r}_f/\partial w$ identifies the pulse's usual error measure with a stability-matrix element whenever the initial configuration depends linearly on the imperfection $w$.

What would settle it

Plot the range parameter $h_\eta(t_f)$ together with the actual spread of the polar angle $\theta$ around the south pole for the $90(x)180(y)90(x)$ pulse under field inhomogeneity: if the histogram minimum occurs at a time where the angular spread is not near its minimum, then histogram collapse is not a reliable refocusing diagnostic.

Watch

Extended reading notes

Core claim

The central claim is that the robustness of composite pulses can be understood as a caustic in a classical canonical description, not as a purely quantum phenomenon. Using $\phi$ (azimuth) and $\eta=\cos\theta$ (polar height) as canonical position and momentum, the paper defines a $2\times2$ stability matrix whose elements measure how the final $\phi$ and $\eta$ respond to small changes in the initial values. Since every member of the ensemble evolves under a slightly different Hamiltonian, the relevant objects are ensemble-averaged stability elements; the paper conjectures that when the histogram of an averaged element collapses to a narrow band at the final time, the ensemble has refocused in that direction. Applied to the $90(x)180(y)90(x)$ pulse, this gives a directional picture: under field inhomogeneity the $\eta$-element refocuses while the $\phi$-element anti-refocuses, whereas under resonance offset both elements refocus at $t_f=T$. The paper also shows that the pulse's traditional perturbative error measure is, under a linearity assumption, one element of this stability matrix, and it explains why the ensemble width is conserved during the second and third segments for field inhomogeneity but not for resonance offset.

Load-bearing premise

The account rests on the unproven conjecture that collapse of the averaged stability histograms indicates refocusing, together with a linearity assumption connecting that collapse to the pulse's usual error measure, an assumption the paper itself notes is poorly satisfied for resonance offsets.

Editorial extensions

If this is right

  • For field inhomogeneity, the $90(x)180(y)90(x)$ pulse refocuses the polar direction $\eta$ while actively spreading the azimuth $\phi$, so its robustness is directional rather than isotropic.
  • For resonance offset, both stability elements collapse at $t_f=T$, matching refocusing in both coordinates and explaining the pulse's compensation of offset imperfections.
  • The traditional first-order perturbative error measure for the pulse is recovered as one stability-matrix element, so minimizing that measure and seeking the caustic are the same operation under linearity.
  • The ensemble width is approximately conserved during the second and third segments under field inhomogeneity because the relevant initial coordinates vanish at the nominal times; under resonance offset no such cancellation occurs and the width changes throughout.
  • Because the analysis needs only the pulse sequence and its imperfections, the caustic-stability picture can be applied to other composite pulse sequences, not just the three-segment case.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • One extension the paper does not draw: the times at which averaged stability elements collapse could serve as an optimization target for designing new pulse timings, since the framework identifies when and in which direction refocusing occurs.
  • The paper's own numerical results show better offset refocusing at times such as $0.4T$ and $0.8T$ than at $t_f=T$; a natural follow-up would be to retune pulse durations to make the caustic coincide with the desired endpoint.
  • The caustic criterion could be used to compare other established composite pulses, for example longer or phase-cycled sequences, by checking which ones show histogram collapse at the target time; the paper does not perform such a comparison.
  • If the linearity assumption between the stability matrix and the error measure fails for offsets, as the paper flags, the caustic picture may still describe refocusing but its equivalence to the usual error measure would need a different derivation.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 5 minor

Summary. The paper proposes a classical-mechanics interpretation of the robustness of composite pulses, using Levitt's 90(x)180(y)90(x) pulse as a case study. The authors map the Bloch-vector dynamics to canonical coordinates (φ, η), define a stability matrix in those coordinates, and argue that ensemble refocusing corresponds to a caustic, i.e., the vanishing of an appropriate stability-matrix element. They report numerical histograms of averaged diagonal stability elements for field-inhomogeneity and resonance-offset ensembles, and use the collapse of these histograms to explain the directional refocusing properties of Levitt's pulse. They further derive an expression for the time variation of the ensemble width and connect Levitt's perturbative error measure to a stability-matrix element under a linearity assumption.

Significance. If the central claim is established, the paper would provide a new, visual, and quantitative framework for understanding composite-pulse robustness, with potential applications to pulse design in NMR, optical spectroscopy, and quantum control. The numerical calculations are carried out with exact Cartesian Bloch dynamics, and the figures clearly convey the qualitative directional effects; the authors are explicit about several limitations of their analysis. However, the core identification between histogram band-collapse and a caustic is currently conjectural, and the paper does not compute the off-diagonal stability element that its own caustic condition requires. The work is original in applying classical stability and caustic concepts to this problem, but the explanatory claim is stronger than the evidence presented.

major comments (4)
  1. [II.C, Eq. (28)] Using the paper's own definition of the Poisson bracket, Eq. (17), and the Lie-Poisson relation Eq. (18), the calculation in Eq. (28) gives {φ,η} = -1, not +1. In detail, {x,z}=y and {y,z}=-x from Eq. (18), so {φ,η} = ∂φ/∂x {x,z} + ∂φ/∂y {y,z} = (-y/(x^2+y^2)) y + (x/(x^2+y^2)) (-x) = -1. Hamilton's equations in Eq. (29) are consistent with {φ,η}=+1, not with Eq. (18). This internal inconsistency affects the claimed canonical structure on which the stability-matrix analysis and the caustic condition rest; the sign convention needs to be corrected and justified.
  2. [II.D and III.A-B (Figs. 3-6)] The caustic condition stated in Section II.D is the vanishing of the off-diagonal element ∂φ_f/∂η_i. However, the numerical evidence in Section III reports only the diagonal elements ∂η_f/∂η_i and ∂φ_f/∂φ_i, through their histograms and the range parameters h_η and h_φ. A collapse of the histogram of a diagonal element means that this element has a narrow spread across the ensemble; it does not imply that any stability-matrix element vanishes, and in particular it does not imply ∂φ_f/∂η_i → 0. Since the abstract's central assertion identifies robust refocusing with the vanishing of an appropriate stability-matrix element, the authors need to compute ∂φ_f/∂η_i (and, if it does not vanish, revise the claim) before the caustic mechanism can be accepted as demonstrated rather than conjectural.
  3. [II.D and III.B] The evaluation times are chosen after the fact: the initial manifold is defined at t_i = T/4 because the swarm 'features maximal spreading', and the success time is t_f = T, which is the pulse's known design target. The stability analysis does not single out these times independently: in Section III.B the authors note that for resonance offsets in the φ-direction there are times around 0.4T and 0.8T where refocusing is better than at T. As a result, the band-collapse at T is partly a re-description of the pulse's known performance rather than a falsifiable prediction. To strengthen the explanatory claim, the authors should show that the caustic-related quantities (e.g., the off-diagonal element or a suitable ensemble-averaged measure) attain a special value at T without using the known success time as input.
  4. [II.D, Eqs. (38)-(39)] The link between Levitt's imperfection measure W = ∂r_f/∂w and the stability matrix M_s requires the assumption that r_i depends approximately linearly on w. The authors concede that this linearity is 'less well satisfied' for resonance offsets, and they are 'cautious about using it to justify Levitt's pulse sequence for resonance offsets'. Since the abstract states that Levitt's perturbative error measure corresponds to one element of the stability matrix, this correspondence should be established quantitatively (e.g., by comparing W with the appropriate matrix element over the actual ensemble range) rather than assumed. Without that check, the claimed connection is heuristic for the resonance-offset case.
minor comments (5)
  1. [II.D, Eq. (36)] The notation defines ⟨M_s⟩ as an integral of a derivative over w, but the text and figures refer to histograms of the elements. Please clarify whether the histograms display the unaveraged ensemble values or the average defined in Eq. (36).
  2. [References [6] and [7]] References [6] and [7] are the same publication; if two distinct articles are intended, correct the citation list.
  3. [II.B, Eq. (20)] The notation ω(r)·∇H(r) with ω(r) a matrix is nonstandard; define the multiplication explicitly.
  4. [III.C, Eq. (57)] The approximation z1^(i) ≈ z1^(0) = 0 is used, but for Ω_i in [0.8,0.9], z1^(i) = cos(Ω_i T/4) ranges up to about 0.156, and the subsequent neglect of sin((Ω_i-Ω_k)t) as '≈ sin(0)' is not justified for t ~ T/2 and differences ~0.1 in Ω. Please provide a quantitative error estimate for the claimed conservation of width along the second and third segments.
  5. [Throughout] The abstract states that this is the first work to introduce a canonical version of the Bloch Equations; given existing literature on Hamiltonian formulations of classical spin dynamics, this novelty claim should be substantiated or softened.

Circularity Check

1 steps flagged · score 2.0 of 10

Mild re-description of Levitt's known refocusing as a caustic; no fitted-parameter or self-citation circularity.

  1. renaming known result [Section II.D (after Eq. 37) and Section IV.D Conclusion]
    "We extend the ordinary framework of caustics, and conjecture that the average stability matrix elements shall indicate refocusing. If the refocusing happens close to the desired endpoint, this signifies that the ensemble has undergone a successful population inversion. In particular, if the histogram of the stability matrix element⟨∂ζf/∂ζi⟩ collapses into a relatively narrow band, this indicates refocusing in theζ-direction. ... In other words, when we look for refocusing of a quantum ensemble, we are actually looking for a classical caustic."

    The paper's central 'why it works' claim is that Levitt's refocusing is a caustic. But the standard caustic condition stated earlier in the same section is ∂ϕ_f/∂η_i = 0, an off-diagonal stability element that is never computed. Instead, the numerical evidence is band-collapse of the diagonal elements ∂η_f/∂η_i and ∂ϕ_f/∂ϕ_i, and the paper stipulates that such collapse 'indicates refocusing' and then identifies refocusing with a caustic. A diagonal element can be constant across the ensemble while remaining nonzero, so the collapse diagnostic is not the stated caustic condition.

full rationale

The paper is largely self-contained: the canonical coordinates are derived from explicit Poisson brackets, and the stability elements are direct finite-difference derivatives of the same exact Bloch dynamics that produce the refocusing. No parameter is fitted to a subset of data, and the benchmark (Levitt's known-effective 90(x)180(y)90(x) pulse) is external. The only co-author citation is a standard textbook used for a textbook formula, so self-citation is not load-bearing. The mild re-descriptive step is concentrated in Section II.D, where 'histogram collapse' is conjectured to indicate refocusing and is then equated with a classical caustic; the actual off-diagonal caustic element is never reported. Other limitations, such as better offset refocusing at t_f≈0.4T and 0.8T, are interpretive/correctness concerns rather than circularity. Overall the circularity is minor and does not undermine the exact dynamical computations.

Assumptions & free parameters 4 free parameters · 5 assumptions · 0 invented entities

The centerpiece (mapping two-level dynamics to Hamiltonian flow on the sphere) rests on standard exact results (Lie-Poisson bracket, RWA frame). The paper's own added structure consists of one conjecture (band-collapse indicates refocusing), one linearity assumption (r_i linear in w) that the authors flag as fragile for offsets, and several hand-chosen analysis parameters (ensemble ranges, pole-avoidance shifts, initial manifold time). No constants are fitted to produce the central claims; the histograms and stability elements are direct computations from exact Bloch dynamics, so the circularity burden is low. The sign inconsistency between eqs. 18 and 28 means the foundational canonical-pair claim is currently mis-stated even though the equations of motion are correct.

free parameters (4)
  • Field-inhomogeneity ensemble range = Ω_1^(j) in [0.8, 0.9]·Ω_1^(0)
    Hand-chosen simulation range (Section I.C) used to demonstrate the directionality claims; the central qualitative results are shown only within this range and are not proven to persist outside it.
  • Resonance-offset ensemble range = Δ^(j) in [0.4, 0.6]
    Hand-chosen simulation range (Section I.C); the authors themselves caution that the link to Levitt's measure is weaker for this ensemble.
  • Pole-avoidance shifts = Δη_0 = 2e-6, Δφ_0 = 1e-6, η_0 = 1-1e-6
    Numerical regularization parameters chosen to bypass the η = ±1 singularity (Section II.D); convergence checks are reported.
  • Initial manifold time = t_i = T/4
    The stability analysis launches from t_i = T/4, where the ensemble 'features maximal spreading' (Section II.D); the stability elements and the directionality conclusions depend on this choice.
assumptions (5)
  • domain assumption Rotating Wave Approximation validity: |ω - ω0| << ω + ω0 (eq. 9)
    Imported from prior literature (Section I.B) to render the dynamics piecewise time-independent; the whole rotation-matrix picture of eq. 12 rests on it.
  • domain assumption Ensemble idealization: no correlations or couplings between spins; pure states only; the two imperfections (field magnitude and resonance offset) are treated separately
    Stated in Sections I.A and I.C ('we will treat imperfections (1) and (2) separately'); real ensembles contain correlations, mixed states, and simultaneous imperfections, and the paper does not extend the analysis to those.
  • ad hoc to paper Caustic criterion: collapse of the histogram of average stability elements indicates refocusing, and divergence indicates anti-refocusing
    Introduced in Section II.D as 'we conjecture that the average stability matrix elements shall indicate refocusing'; it is the load-bearing bridge between the stability computations and the robustness explanation, and it is never proven.
  • ad hoc to paper Linearity of the initial manifold r_i in the imperfection parameter w = Ω_1 (approximately true for Δ)
    Assumed in Section II.D (eq. 38) to identify Levitt's measure W with the spherical stability matrix; the paper admits the assumption is weakly satisfied for resonance offsets.
  • standard math The pair (φ, η = cos θ) is canonical with {φ, η} = 1, reducing the Lie-Poisson flow to a two-dimensional canonical Hamiltonian system
    Derived in Section II.C (eq. 28); the calculation is standard, but as written there is a sign inconsistency with the paper's own eq. 18, and the claimed first-ness overstates the novelty.

how reviews work

0 comments
Cite this review

Pith. "Pith review of The robustness of composite pulses elucidated by classical mechanics: Stability around the globe." pith.science (2026). https://pith.science/paper/IB63CBAN

@misc{pith2026250701364,
  author       = {Pith},
  title        = {Pith review of: The robustness of composite pulses elucidated by classical mechanics: Stability around the globe},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/IB63CBAN}},
  note         = {Machine review of arXiv:2507.01364}
}
abstract

Composite Pulses (CPs) are widely used in Nuclear Magnetic Resonance (NMR), optical spectroscopy, optimal control experiments and quantum computing to manipulate systems that are well-described by a two-level Hamiltonian. A careful design of these pulses can allow the refocusing of an ensemble at a desired state, even if the ensemble experiences imperfections in the magnitude of the external field or resonance offsets. Since the introduction of CPs, several theoretical justifications for their robustness have been suggested. In this work, we suggest another justification based on the classical mechanical concept of a stability matrix. The motion on the Bloch Sphere is mapped to a canonical system of coordinates and the focusing of an ensemble corresponds to caustics, or the vanishing of an appropriate stability matrix element in the canonical coordinates. Our approach highlights the directionality of the refocusing of the ensemble on the Bloch Sphere, revealing how different ensembles refocus along different directions. The approach also clarifies when CPs can induce a change in the width of the ensemble as opposed to simply a rotation of the axes. As a case study, we investigate the $90(x)180(y)90(x)$ CP introduced by Levitt, where the approach provides a new perspective into why this CP is effective.

Figures

Figures reproduced from arXiv: 2507.01364 by the authors.

Figure 1
Figure 1. A set of trajectories on the Bloch Sphere, representing the time-evolution of an ensemble under the Levitt pulse [PITH_FULL_IMAGE:figures/full_fig_p005_1.png] view at source ↗
Figure 2
Figure 2. Left: phase space representation of a trajectory starting from [PITH_FULL_IMAGE:figures/full_fig_p008_2.png] view at source ↗
Figure 3
Figure 3. The time-evolution of a histogram representing the average stability element (in absolute value) [PITH_FULL_IMAGE:figures/full_fig_p012_3.png] view at source ↗
Figures from the paper (4 more)
Figure 4
Figure 4. Figure 4: The time-evolution of a histogram representing the average stability element (in absolute value) [PITH_FULL_IMAGE:figures/full_fig_p012_4.png]
Figure 5
Figure 5. Figure 5: The time-evolution of a histogram presenting the average stability element (in absolute value) [PITH_FULL_IMAGE:figures/full_fig_p013_5.png]
Figure 6
Figure 6. Figure 6: The time-evolution of a histogram presenting the average stability element (in absolute value) [PITH_FULL_IMAGE:figures/full_fig_p014_6.png]
Figure 7
Figure 7. Figure 7: demonstrates the numerical results for the Levitt’s sequence. The left subfigure corresponds to field inhomogeneity and the right subfigure corresponds to resonance offsets. Note that the width of the ensemble is preserved in second and the third segment of the former …

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

21 extracted references · 20 canonical work pages

  1. [1]

    M. H. Levitt and R. Freeman, Journal of Magnetic Resonance (1969)33, 473 (1979)

    work page 1969

  2. [2]

    E. L. Hahn, Physical Review80, 580 (1950)

    work page 1950

  3. [3]

    H. Y. Carr and E. M. Purcell, Physical Review94, 630 (1954)

    1954

  4. [4]

    Meiboom and D

    S. Meiboom and D. Gill, Review of Scientific Instruments29, 688 (1958)

    work page 1958

  5. [5]

    S. J. Glaser, U. Boscain, T. Calarco, C. P. Koch, W. Köckenberger, R. Kosloff, I. Kuprov, B. Luy, S. Schirmer, T. Schulte- Herbrüggen, D. Sugny, and F. K. Wilhelm, The European Physical Journal D69, 279 (2015)

    work page 2015

  6. [7]

    M. H. Levitt, Journal of Magnetic Resonance (1969)50, 95 (1982)

    work page 1969

  7. [8]

    Counsell, M

    C. Counsell, M. Levitt, and R. Ernst, Journal of Magnetic Resonance (1969)63, 133 (1985)

    work page 1969

  8. [9]

    Blümich and H

    B. Blümich and H. Spiess, Journal of Magnetic Resonance (1969)61, 356 (1985)

    work page 1969

Show all 21 references
  1. [10]

    Dridi, M

    G. Dridi, M. Mejatty, S. J. Glaser, and D. Sugny, Physical Review A101, 012321 (2020)

    2020

  2. [11]

    D. J. Tannor,Introduction to Quantum Mechanics: A Time-Dependent Perspective(University Science Books, 2007) pp. 227–272

    2007

  3. [12]

    Kuprov,Spin(Springer International Publishing, Cham, 2023)

    I. Kuprov,Spin(Springer International Publishing, Cham, 2023)

    2023

  4. [13]

    R. P. Feynman, F. L. Vernon, and R. W. Hellwarth, Journal of Applied Physics28, 49 (1957)

    1957

  5. [14]

    M. H. Levitt, Progress in Nuclear Magnetic Resonance Spectroscopy18, 61 (1986)

    1986

  6. [15]

    Note that this notation shall be read as one piece

  7. [16]

    J. E. Marsden and T. S. Ratiu,Introduction to Mechanics and Symmetry, Vol. 17 (Springer New York, New York, NY, 1999)

    1999

  8. [17]

    Ross, Poisson Brackets, Non-Canonical Hamiltonian Systems and Euler’s Rigid Body Equations | Lecture 12 (2020)

    S. Ross, Poisson Brackets, Non-Canonical Hamiltonian Systems and Euler’s Rigid Body Equations | Lecture 12 (2020)

    2020

  9. [18]

    C. P. Slichter,Principles of Magnetic Resonance, Vol. 1 (Springer Berlin Heidelberg, Berlin, Heidelberg, 1990)

    1990

  10. [19]

    Eidson and R

    J. Eidson and R. F. Fox, Physical Review A34, 3288 (1986)

    1986

  11. [20]

    Pomeau, B

    Y. Pomeau, B. Dorizzi, and B. Grammaticos, Physical Review Letters56, 681 (1986)

    1986

  12. [21]

    R. G. Littlejohn, Journal of Statistical Physics68, 7 (1992)

    1992

  13. [22]

    E. J. Heller, The Journal of Chemical Physics65, 4979 (1976)

    1976

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.