Pith. sign in

REVIEW 3 major objections 5 minor 33 references

Semiclassical approach to Bose-Einstein condensates in a triple well potential

T0 review · 3 major / 5 minor · reviewed 2026-08-28 · deepseek-v4-flash

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

Pith's one-line read For a triple-well condensate of thirty particles, the guiding centers of the classical mean-field motion classify 180 of the 496 eigenstates and assign them geometric quantum numbers.

desk verdict A sound transfer of a molecular semiclassical classification to a three-mode BEC, let down by an unverified identification of geometric quantum numbers with particle numbers. read the letter →

arxiv quant-ph/0604158 v2 pith:CVVMGSH3 submitted 2006-04-20 quant-ph

S. Mossmann , C. Jung This is my paper · ORCID
classification quant-ph PACS 03.75.Kk03.75.Lm03.65.Sq
keywords Bose-EinsteincondensatetriplewellpotentialsemiclassicalwavefunctionsBose-Hubbardmodelmean-fieldGross-Pitaevskiiguidingcentersgeometricquantumnumbersmodeentanglement
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

This paper tries to establish that the quantum eigenstates of a Bose-Einstein condensate in a triple well potential are not opaque numerical objects: when the many-body problem is truncated to three modes and read in action-angle variables, 180 of the 496 eigenstates for thirty particles organize around the same geometric structures that organize the classical mean-field dynamics. The paper shows that along each guiding center, the wave function is close to a plane wave times a transverse oscillator, so each assigned state gets a longitudinal and a transverse quantum number. From those two numbers one can read off how many particles sit in each well and which modes are entangled, without any further calculation. A sympathetic reader would care because this turns diagonalization output into an intuitive picture and transfers the same classification to the Gross-Pitaevskii time evolution, predicting phase locking, self-trapping, and chaotic intermittency from the quantum side.

What carries the argument

The working machinery is the semiclassical wave function on the toroidal configuration space. Starting from a Bose-Hubbard eigenstate expanded in number states |n1,n2,n3>, the paper reinterprets the expansion coefficients as the Fourier series of a wave function in the relative angles psi_1 = phi_1 - phi_2 and psi_2 = phi_3 - phi_2, dropping the global phase. On this two-torus, density crests trace the classical organizing centers, and the gradient of the phase gives the local action, so counting the phase advance around a crest yields the longitudinal quantum number and counting transverse nodal lines yields the transverse quantum number. The companion classical object is the action-angle Hamiltonian reduced by the conserved total action, whose guiding centers - primary tori, locked lines, the diagonal, and the point (0,0) - define the classes. The zero-point-corrected correspondence I_k <-> n_k + 1/2 makes the quantum-classical comparison quantitative at N = 30.

What would settle it

Pick any state assigned to the psi_1 = 0 class (type C) and trace the phase of its semiclassical wave function around the density crest: the method predicts the total phase advance is exactly an integer multiple of 2*pi and that mode 3 is nearly pure with that integer as its particle number. A numerical computation showing a non-integer phase advance, or substantial entanglement of mode 3, would falsify the assignment.

Watch

Extended reading notes

Core claim

The central claim is that the eigenstates of the three-mode Bose-Hubbard Hamiltonian with a conserved particle number can be classified by the guiding centers of the corresponding classical Hamiltonian in action-angle variables, obtained by Heisenberg substitution. After reduction by the conserved total action K = 31.5, the classical motion lives on a two-torus, and its organizing structures are primary tori covering the whole torus (type A), one-dimensional frequency-locked curves psi_2 = 0 (type B) and psi_1 = 0 (type C), the diagonal psi_1 = psi_2 (type D), a point center at (0,0) with oscillations around it (type E1), and large-scale chaotic motion (type E2). For 180 of the 496 eigenstates of the N = 30 system, the density and phase of the semiclassical wave function identify one of these centers. Along a density crest the phase advances by an integer multiple of 2*pi, giving the longitudinal quantum number; the number of nodal lines transverse to the center gives the transverse quantum number. The idealized wave functions are products of a plane wave in the longitudinal direction and an oscillator function transversely, and translating them back to well coordinates shows which modes separate and which remain entangled. The same assignment applies to the time evolution of the mean-field equations: initial conditions built from the quantum states evolve with the phase-locking, independent, or intermittent behavior of the corresponding class.

Load-bearing premise

Everything rests on the truncation to three modes - one localized state per well, with higher excited states and fourth-order coupling terms neglected - so if the real condensate populates additional bands or sites, the 180 classified eigenstates belong to a model that is not the physical system.

Editorial extensions

If this is right

  • For the N = 30 triple well, 180 of 496 eigenstates can be labeled by geometric quantum numbers read directly from density and phase plots, so classification requires no additional diagonalization or extra observables.
  • The quantum numbers translate into physical content: in type C states, the longitudinal number counts particles in well 3 while wells 1 and 2 are entangled; in type A states, two longitudinal numbers count particles in wells 1 and 3 and all modes separate; in type E1 states, all three modes are entangled.
  • The same (A)-(E2) scheme classifies the time evolution of the Gross-Pitaevskii mean-field equations, so diagonalizing the quantum Hamiltonian and plotting eigenfunctions yields a grid of initial conditions whose behavior - phase locking, self-trapping, chaotic intermittency - is known in advance.
  • The semiclassical correspondence with the zero-point shift I_k <-> n_k + 1/2 gives better agreement for N = 30 than the naive identification I_k = n_k, and the method is designed to improve as N grows.
  • Because the method is formulated in general action-angle variables, it extends to symmetric systems with more than three wells, where the reduced configuration torus is higher-dimensional.

Reading between the lines

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

  • If the guiding-center classification persists for larger N or more wells, it offers a cheap classicality diagnostic: the unassignable eigenstates are exactly those sitting over chaotic mean-field regions, suggesting a state-by-state breakdown of the Gross-Pitaevskii description that a simple density plot could reveal.
  • The idealized wave-function forms give direct, testable entanglement predictions - type A factorizes, type E1 entangles all modes - so entanglement entropies computed from the exact eigenstates could validate the visual classification quantitatively.
  • The phase-function 'lift' from configuration space to action space could be turned into a systematic algorithm that extracts local phase gradients to reconstruct the actions for each density crest, potentially assigning quantum numbers to some of the 316 unassigned states that currently fall between classes.
  • Applied to a four-well ring, one would expect new resonance lines and mixed guiding centers; a testable prediction is that the fraction of assignable states changes with the trap geometry, indicating which shapes are best described by mean-field theory.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Request a human review

A listed scientist reviews the paper for a fee and the review publishes here regardless of verdict. See the reviewers or get listed.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 5 minor

Summary. The paper presents a semiclassical method for analyzing the three-mode Bose-Hubbard model describing a Bose-Einstein condensate in a triple-well potential. The authors construct a classical Hamiltonian via Heisenberg substitution (I ↔ n + 1/2), reduce the conserved total action to obtain a two-degree-of-freedom system, and represent quantum eigenstates as Fourier series on the corresponding torus. They then compare these 'semiclassical wave functions' with classical guiding centers (primary tori, periodic orbits, fixed points, chaotic regions) and classify 180 of the 496 eigenstates into types (A) through (E2). For each class they assign geometric quantum numbers (μ_l, μ_t) and claim these encode particle numbers in individual wells and the entanglement structure of the modes, allowing such properties to be read off 'without further calculations.' They also use the classification to interpret solutions of the Gross-Pitaevskii mean-field equations.

Significance. If the central identification is correct, the method would provide an intuitive and transferable tool for few-mode Bose-Einstein condensates, connecting quantum many-body eigenstates to classical phase-space structures without parameter fitting. The manuscript has clear strengths: the canonical transformation in Section II A is explicit and correct, the Fourier representation of eigenstates is well defined, the idealized wave functions of Section IV are internally consistent, and the authors are transparent about the finite-resolution effects for N=30 and about the mode-truncation caveat in the Conclusion. The semiclassical correspondence used is a standard external benchmark rather than a fitted assumption. However, the paper's main advertised capability—reading off particle numbers and entanglement from phase advances—is not quantitatively validated against the actual eigenstates, and this gap is load-bearing for the central claims.

major comments (3)
  1. [Section IV, type C (Eqs. (29)-(30))] The identification 'the number of particles in mode 3 is given by μ_l' is derived for the idealized plane-wave form e^{iμ_l ψ2} and is then applied to actual eigenstates. For a superposition of number states, the phase winding number of the wave function is a topological invariant that does not generally equal the mean occupation number; for example, the state (|0,N,0⟩+|N,0,0⟩)/√2 has ⟨n_1⟩=N/2 but a winding number of 1 in the relative angle. The paper never computes the expectation values ⟨n_1⟩, ⟨n_2⟩, ⟨n_3⟩, nor the leading number-state components, for the 180 assigned states. Without such a check, the abstract's claim that particle exchange can be read off 'without further calculations' is unsupported even within the Bose-Hubbard model.
  2. [Section IV and Conclusion] The entanglement characterizations (e.g., 'mode 3 separates' for type C in Eq. (30), 'the three degrees of freedom are completely disentangled' for type A in Eq. (34)) are inferred from the idealized wave functions, not from the actual eigenstates. No entanglement measure (such as the von Neumann entropy of a reduced density matrix or a Schmidt-number analysis) is computed to verify that the mode separability or entanglement structure of the actual eigenstates matches the idealized forms. Since the stated utility of the method includes characterizing entanglement, this missing validation is a central gap.
  3. [Section IV, final paragraph] The authors acknowledge that 'the above graphical classification of the semiclassical wave functions is not strict and some functions allow ambiguous assignments,' but they do not quantify how many of the 180 assigned states are unambiguous or provide an objective criterion for resolving ambiguities. The headline result is the count '180 of the 496 eigenstates'; without a robustness estimate, this number is not well defined. Adding a stability check (e.g., overlap-based assignment or sensitivity to small parameter changes) would materially strengthen the classification claim.
minor comments (5)
  1. [Title] The title contains an extra space: 't riple well' should read 'triple well.'
  2. [Section IV, type D] In the paragraph describing state Φ420 and Φ359, the text reads 'µl = 6 for state Φ420 and µl = 8 for state Φ420'; the second reference should presumably be to Φ359.
  3. [Section IV, type B] The phrase 'the states of class B loose their characteristics' should use 'lose' rather than 'loose.'
  4. [Section V B, Eq. (41)] The text states that the coefficients c_{n1,n2,n3} in Eq. (41) are real-valued, but if the eigenstate coefficients are complex, the expectation value in Eq. (42) should use |c|² rather than c². Please clarify under which conditions the coefficients are real.
  5. [Section V A, Figs. 12 and 13] The classical time-evolution classification in Fig. 12 shows no type (D) points, while the quantum classification in Fig. 13 does include type (D). This discrepancy is not explicitly discussed; a sentence explaining whether this is a finite-time effect or a genuine difference would be helpful.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the quantum-to-classical comparison is self-contained; self-citations are historical, not load-bearing.

full rationale

The derivation chain is self-contained. The paper diagonalizes the Bose-Hubbard Hamiltonian (3)-(5) exactly for N=30, expands the eigenstates in the angle representation (22), and compares the resulting density and phase patterns with structures computed independently from the classical Hamiltonian (8) obtained by the standard Heisenberg substitution (6) with the usual I <-> n+1/2 correspondence (7). No parameter is fitted to a subset of data and then "predicted"; the longitudinal numbers mu_l are read from phase advances of the semiclassical wave functions, not from occupation expectation values, and the identification of mu_l with particle numbers in a given well (Eqs. (30), (32), (34)) is a semiclassical interpretation of the idealized form, not a reduction of the assignment to the input. Self-citations [14]-[16] describe the molecular origin of the method but are not load-bearing: the classification procedure (phase advance counting, nodal counting) is fully specified in Section IV, and no uniqueness theorem or unverified prior result is invoked to force the assignment. The truncation caveat in Section VI and the unsupported (but non-circular) equation of mu_l with particle numbers are correctness concerns, not circularity.

Assumptions & free parameters 2 free parameters · 4 assumptions · 0 invented entities

The central claim rests on the choice of the Bose-Hubbard three-mode model, the semiclassical action-angle substitution, and the interpretive step that treats the angle representation as a wave function on the torus. These are standard or explicitly flagged assumptions. No new physical entities are introduced, and no parameters are fitted to data beyond the illustrative parameter set.

free parameters (2)
  • Hamiltonian model parameters (omega_i, x_i, k_ij) = omega1=0.1, omega2=0, omega3=-0.1, x1=x2=x3=0.1, k12=k23=0.5
    Chosen by hand for the demonstration and noted as consistent with experimental realizations; not fitted to any data. The classification counts and the dynamics depend on this specific parameter set.
  • Particle number N = 30
    Chosen as a test case where exact diagonalization is easy but the spectrum has structure; not fitted. The finite resolution of the angle representation and the semiclassical correspondence depend on this value.
assumptions (4)
  • domain assumption The classical Hamiltonian is obtained from the quantum Bose-Hubbard Hamiltonian via Heisenberg substitution a_k -> sqrt(I_k) exp(i phi_k), with the semiclassical correspondence I_k <-> n_k + 1/2.
    Section II.A; this correspondence has errors of order hbar^2 and is assumed to be globally good based on experience with molecular systems.
  • domain assumption Only the lowest Wannier state per well (m=1) and only three wells are kept; higher bands and higher-order inter-mode terms are neglected.
    Section II, 'Assuming Bose-Einstein condensates, we can restrict ourselves to the lowest energy state m=1'; Section VI concedes the truncation depends on the basis and potential.
  • domain assumption The angular representation |phi1,phi2,phi3> given by the finite Fourier sum over number states is a valid semiclassical wave function on the toroidal configuration space, and phase gradients can be read as classical actions.
    Sections II.B and IV; the representation is non-orthogonal for finite N (Eq. 19) and the semiclassical interpretation is argued by analogy to molecular systems.
  • standard math The conserved total action K reduces the three-mode system to a two-degree-of-freedom system without loss of relevant dynamics.
    Section II.A, generating function Eq. (11); this is a standard canonical reduction using the conserved quantity.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Semiclassical approach to Bose-Einstein condensates in a triple well potential." pith.science (2026). https://pith.science/paper/CVVMGSH3

@misc{pith2026quant-ph0604158,
  author       = {Pith},
  title        = {Pith review of: Semiclassical approach to Bose-Einstein condensates in a triple well potential},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/CVVMGSH3}},
  note         = {Machine review of arXiv:quant-ph/0604158}
}
read the original abstract

We present a new approach for the analysis of Bose-Einstein condensates in a few mode approximation. This method has already been used to successfully analyze the vibrational modes in various molecular systems and offers a new perspective on the dynamics in many particle bosonic systems. We discuss a system consisting of a Bose-Einstein condensate in a triple well potential. Such systems correspond to classical Hamiltonian systems with three degrees of freedom. The semiclassical approach allows a simple visualization of the eigenstates of the quantum system referring to the underlying classical dynamics. From this classification we can read off the dynamical properties of the eigenstates such as particle exchange between the wells and entanglement without further calculations. In addition, this approach offers new insights into the validity of the mean-field description of the many particle system by the Gross-Pitaevskii equation, since we make use of exactly this correspondence in our semiclassical analysis. We choose a three mode system in order to visualize it easily and, moreover, to have a sufficiently interesting structure, although the method can also be extended to higher dimensional systems.

Figures

Figures reproduced from arXiv: quant-ph/0604158 by the authors.

Figure 1
Figure 1. FIG. 1: The classical reduced system for energy [PITH_FULL_IMAGE:figures/full_fig_p007_1.png] view at source ↗
Figure 2
Figure 2. FIG. 2: (Color online) The classical reduced system for an [PITH_FULL_IMAGE:figures/full_fig_p007_2.png] view at source ↗
Figure 3
Figure 3. FIG. 3: (Color online) The classical reduced system for an en [PITH_FULL_IMAGE:figures/full_fig_p008_3.png] view at source ↗
Figures from the paper (11 more)
Figure 4
Figure 4. Figure 4: FIG. 4: (Color online) Classical reduced system for energy [PITH_FULL_IMAGE:figures/full_fig_p009_4.png]
Figure 5
Figure 5. Figure 5: FIG. 5: Gray scale plot of the squared modulus of the (a) [PITH_FULL_IMAGE:figures/full_fig_p010_5.png]
Figure 6
Figure 6. Figure 6: FIG. 6: Plot of the eigenfunctions Φ [PITH_FULL_IMAGE:figures/full_fig_p011_6.png]
Figure 7
Figure 7. Figure 7: FIG. 7: Plot of the eigenfunctions Φ [PITH_FULL_IMAGE:figures/full_fig_p012_7.png]
Figure 9
Figure 9. Figure 9: FIG. 9: Plot of the eigenfunctions Φ [PITH_FULL_IMAGE:figures/full_fig_p013_9.png]
Figure 11
Figure 11. Figure 11: FIG. 11: (Color online) Time evolution for an initial condit [PITH_FULL_IMAGE:figures/full_fig_p014_11.png]
Figure 10
Figure 10. Figure 10: FIG. 10: (Color online) Time evolution of Eq. (37) for an ini [PITH_FULL_IMAGE:figures/full_fig_p014_10.png]
Figure 13
Figure 13. Figure 13: FIG. 13: (Color online) Characterization of the classical a [PITH_FULL_IMAGE:figures/full_fig_p015_13.png]
Figure 12
Figure 12. Figure 12: FIG. 12: (Color online) Characterization of the classical [PITH_FULL_IMAGE:figures/full_fig_p015_12.png]
Figure 14
Figure 14. Figure 14: FIG. 14: (Color online) Time evolution of the mean-field [PITH_FULL_IMAGE:figures/full_fig_p016_14.png]
Figure 15
Figure 15. Figure 15: FIG. 15: (Color online) Time evolution of Eq. (37) for the [PITH_FULL_IMAGE:figures/full_fig_p016_15.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

33 extracted references · 33 canonical work pages

  1. [1]

    Albiez, R

    M. Albiez, R. Gati, J. F¨ olling, S. Hunsmann, M. Cristiani and M. K. Oberthaler, Phys. Rev. Lett. 95 (2005) 010402

    work page 2005

  2. [2]

    (38) as the number of particles in mode k

    (38) In this way, we can construct initial conditions ck(0) =√Ik, where the action Ik can be interpreted quantum mechanically via Eq. (38) as the number of particles in mode k. Furthermore, we can use this correspondence in order to construct initial conditions resembling the properties of the eigenstates of the system. Before we explain this in more deta...

    work page

  3. [3]

    Schumm, S

    T. Schumm, S. Hofferberth, L. M. Andersson, S. Wilder- muth, S. Groth, I. Bar-Joseph, J. Schmiedmayer and P. Kr¨ uger, Nature Physics1 (2005) 57

    work page 2005

  4. [4]

    Smerzi, S

    A. Smerzi, S. Fantoni, S. Giovanazzi, and S. R. Shenoy, Phys. Rev. Lett. 79 (1997) 4950

    work page 1997

  5. [5]

    Witthaut, E

    D. Witthaut, E. M. Graefe, and H. J. Korsch, Phys. Rev. A 73 (2006) 063609

    work page 2006

  6. [6]

    J. Liu, B. Wu, and Q. Niu, Phys. Rev. Lett. 90 (2003) 170404

    work page 2003

  7. [7]

    Biao Wu and Qian Niu, New J. Phys. 5 (2003) 104

    work page 2003

  8. [8]

    G. J. Milburn, J. Corney, E. M. Wright, and D. F. Walls, Phys. Rev. A 55 (1997) 4318

    work page 1997

Show all 33 references
  1. [9]

    Smerzi and Srikanth Raghavan, Phys

    A. Smerzi and Srikanth Raghavan, Phys. Rev. A 61 (2000) 063601

    2000

  2. [10]

    J. R. Anglin and A.Vardi, Phys. Rev. A 64 (2001) 013605

    2001

  3. [11]

    J. R. Anglin, P. Drummond and A. Smerzi, Phys. Rev. A 64 (2001) 063605

    2001

  4. [12]

    Thommen, J

    Q. Thommen, J. C. Garreau, and V. Zehnl´ e, Phys. Rev. Lett. 91 (2003) 210405

    2003

  5. [13]

    K. W. Mahmud, H. Perry and W. P. Reinhardt, Phys. Rev. A 71 (2005) 023615

    2005

  6. [14]

    E. L. Sibert III and A. B. McCoy, J. Chem. Phys. 105 (1996) 469

    1996

  7. [15]

    M. P. Jacobson, C. Jung, H. S. Taylor, and R. W. Field, J. Chem. Phys. 111 (1999) 600

    1999

  8. [16]

    C. Jung, H. S. Taylor and E. Atilgan, J. Phys. Chem. A 106 (2002) 3092

    2002

  9. [17]

    C. Jung, C. Mejia-Monasterio, and H. S. Taylor, J. Chem. Phys. 120 (2004) 4194

    2004

  10. [18]

    E. M. Graefe, H. J. Korsch, and D. Witthaut, Phys. Rev. A 73 (2006) 013617

    2006

  11. [19]

    Nemoto, C

    K. Nemoto, C. A. Holmes, G. J. Milburn and W. J. Munro, Phys. Rev. A 63 (2000) 013604

    2000

  12. [20]

    Buonsante, R

    P. Buonsante, R. Franzosi and V. Penna, Phys. Rev. Lett. 90 (2003) 050404

    2003

  13. [21]

    Franzosi and V

    R. Franzosi and V. Penna, Phys. Rev. E 67 (2003) 046227

    2003

  14. [22]

    Buonsante, R

    P. Buonsante, R. Franzosi and V. Penna, J. Phys. B 37 (2004) S229

    2004

  15. [23]

    Kohn, Phys

    W. Kohn, Phys. Rev. 115 (1959) 809

    1959

  16. [24]

    M. P. A. Fisher, P. B. Weichman, G. Grinstein, and D. S. Fisher, Phys. Rev. B 40 (1989) 546

    1989

  17. [25]

    Spekkens and J.E

    R.W. Spekkens and J.E. Sipe, Phys. Rev. A 59 (1999) 3868

    1999

  18. [26]

    Heisenberg, Z

    W. Heisenberg, Z. Physik 33 (1925) 879

    1925

  19. [27]

    Brack and R

    M. Brack and R. K. Bhaduri, Semiclassical physics , Ad- dison Wesley, 1997

    1997

  20. [28]

    M. C. Gutzwiller, Chaos in Classical and Quantum Me- chanics, Springer, New York, 1990 18

    1990

  21. [29]

    B. V. Chirikov, Phys. Rep. 52 (1979) 263

    1979

  22. [30]

    R. Jost, M. Joyeux, S. Skokov and J. Bowman, J. Chem. Phys. 111 (1999) 6807

    1999

  23. [31]

    Azzam, R

    T. Azzam, R. Schinke, S. Farantos, M. Joyeux and K. Peterson, J. Chem. Phys. 118 (2003) 9643

    2003

  24. [32]

    Gomez Llorente, J

    J. Gomez Llorente, J. Zakrzewski and H. S. Taylor, J. Chem. Phys. 90 (1989) 1505

    1989

  25. [33]

    Trombettoni and A

    A. Trombettoni and A. Smerzi, Phys. Rev. Lett. 86 (2001) 2353

    2001

Pith tools

Reviewed August 28, 2026 · model on record in the stance chip above.