Completing the closed-orbit (topological-sum) representation of regular bound spectra: uniform approximations beyond two degrees of freedom, K = 0 coalescences, and convergence of the orbit sum

For quantum systems whose classical motion is regular (multiply periodic), Berry and Tabor (1976) transformed the Einstein–Brillouin–Keller quantization conditions into a 'topological sum' expressing the bound-state density of states entirely in terms of closed classical orbits, including complex orbits below the bottom of the well. Three parts of that construction were left incomplete: the uniform approximation that repairs the stationary-phase evaluation when the contributing torus coalesces with the boundary of action space was derived only for two degrees of freedom and tested only on separable central potentials; the case where the energy contour in action space has inflection points (zero curvature $K=0$), which would produce 'bound-state rainbows' and higher-order catastrophes in the spectrum, was not analyzed at all; and the convergence of the topological sum in the infinite-orbit limit and the resolution of individual levels was not established. Beyond the major progress on adjacent problems — two-dimensional uniformizations, symmetry-reduced family trace formulas, and bifurcation uniform approximations — the higher-dimensional, genuinely non-separable uniform approximation, the existence and spectral signature of $K=0$ coalescences, and a convergence theory for the unsmoothed orbit sum remain open.

Background

In the semiclassical (small $\hbar$) limit, a bound quantum system whose classical motion is regular — that is, multiply periodic, confined to invariant tori in phase space — has energy levels accurately given by the Einstein–Brillouin–Keller (EBK) conditions: each of the independent action integrals $I_1,\dots,I_f$ (half the phase-space integrals of the canonical momenta over the $f$ fundamental loops of the torus) is quantized, $I_i=\hbar(n_i+\mu_i/4)$, with integer quantum numbers $n_i$ and Maslov indices $\mu_i$ counting caustics (surfaces where a classical coordinate description breaks down). In action-angle variables the tori are labeled by the actions $I$, and the energy is a function $E=H(I)$ of them alone; the set of actions at a given energy forms the 'energy contour' in the $f$-dimensional action space.

Berry and Tabor (1976) asked the converse question: can the density of states $n(E)$, the sum of delta-functions at the eigenenergies, be represented purely in terms of classical closed orbits, as Gutzwiller's trace formula does for chaotic systems with isolated periodic orbits (Gutzwiller 1971)? Starting from the EBK conditions rather than from the Green function, they obtained a 'topological sum': $n(E)$ becomes a sum over integer vectors $M$ labeling the topology (winding numbers, plus repetition number) of the families of closed orbits whose frequency ratios are commensurable, each contributing an oscillatory term involving the action $W(M)$ along the orbit family, the curvature $K$ of the energy contour in action space (the curvature of a line for $f=2$, the Gaussian curvature of a surface for $f=3$), and Maslov-type indices. Two refinements were needed for the formula to work in practice. First, stationary phase over the energy contour fails when the stationary torus lies near or outside the boundary of the allowed action region; the missing contributions come from tori with slightly negative action components, corresponding to 'complex closed orbits' — periodic motion below the bottom of the well, with all constants of motion still real. Incorporating these, Berry and Tabor derived a uniform approximation (their equation (24)), a Fresnel-type integral that interpolates smoothly between the stationary-point and endpoint (one-dimensional motion) contributions and also gives the correct harmonic-oscillator limit — but they explicitly restricted its derivation to systems with two degrees of freedom, where the energy contour is a line and the boundary consists of two endpoints. Second, following Balian and Bloch (1972), who studied hard-walled enclosures by giving the energy an imaginary part $\mathrm{i}\gamma$ that smooths each level into a Lorentzian, each orbit is weighted by a decay factor $\exp(-\gamma T/\hbar)$, with $T$ the orbit period; this damping is what makes the sum over infinitely many closed orbits convergent at finite $\gamma$.

The theory is, by construction, a theory of regular spectra only. Percival (1973) distinguished 'regular' from 'irregular' spectra and conjectured that classically ergodic motion produces irregular, effectively random levels; Berry and Tabor noted that for ergodic systems the semiclassical limit of this torus-based construction may not exist at all, and that such systems fall outside the theory (the complementary isolated-orbit trace formula of Gutzwiller and, later, random-matrix theory describe that regime instead).

Three restrictions of the 1976 construction survived the paper. (1) The uniform approximation was derived for two degrees of freedom; the three-dimensional spherical well tested in the paper is, in action space, effectively two-dimensional because of separability, and no genuinely non-separable system was tested. (2) The derivation assumes the energy contour has no inflection points: where the curvature $K$ vanishes, several stationary tori can coalesce, and the stationary-phase topological sum fails; a point of inflection would produce a dense cluster of levels that the authors called a 'bound-state rainbow', with more complicated coalescences giving higher-order catastrophes — in the sense of Thom's catastrophe theory, which classifies the ways stationary points of a phase integral can coalesce and supplies canonical diffraction integrals that uniformize them; the authors cited exactly this literature; they knew of no smooth potential exhibiting an inflection and did not analyze it. (3) The convergence of the topological sum and the resolution of individual levels as the smoothing is removed ($\gamma\to 0$) were illustrated numerically but not analyzed. Neighboring work has since built a large structure around this problem: level-spacing statistics of regular spectra (Berry and Tabor 1977b), EBK quantization of non-separable molecular systems (Martens and Ezra 1985), uniform approximations at resonances and bifurcations of periodic orbits (Ozorio de Almeida and Hannay 1987; Schomerus and Sieber 1997; Sieber 1997; Bartsch, Main and Wunner 1999; Arita and Brack 2008), trace formulas for continuous families of orbits under symmetry (Creagh and Littlejohn 1991, 1992; Robbins, Creagh and Littlejohn 1989), and modern extensions of the Berry–Tabor formula itself to Bethe-ansatz-integrable many-body systems (Urbina, Kelly and Richter 2023).

Problem Statement

Complete the closed-orbit (topological-sum) representation of the regular bound spectrum of multiply periodic systems in the respects left open by Berry and Tabor (1976), and delimit its validity: (1) Derive a uniform approximation for the contributions of closed-orbit families to the density of states of regular systems with $f\ge 3$ degrees of freedom that is genuinely non-separable (not reducible by separability or by a continuous-symmetry reduction), valid both when the stationary torus of the action-space integral approaches and crosses the boundary of the allowed action region — the situation treated in two dimensions by equation (24) of the 1976 paper, including complex closed orbits — and when the stationary points coalesce with one another; in particular, cover the case where the energy contour in action space has regions of vanishing curvature $K=0$ (inflection points for $f=2$), where the present stationary-phase topological sum fails and the spectrum should show 'bound-state rainbows' or higher-order catastrophes. (2) Establish whether smooth potentials with $K=0$ inflection regions on the action-space energy contour exist at all, construct or exclude them, and characterize the resulting spectral signatures (dense level clusters) with the corresponding uniform (catastrophe-type) formula. (3) Determine the convergence properties of the topological sum in the infinite-orbit limit: establish, refute, or delimit the resolution of individual energy levels as the smoothing width $\gamma$ tends to zero, given that finite $\gamma$ damping makes the sum convergent but partial sums over restricted orbit topologies produce false singularities. Deliverables (1)–(3) jointly serve one objective: making the closed-orbit representation of regular spectra a uniformly valid and convergent theory, with the same standard of numerical test against exactly computed spectra (for non-separable as well as separable systems) that the 1976 paper applied to its two-dimensional separable examples. The ergodic/Percival regime is outside the theory by construction and is not part of the requested result.

References

1. M. V. Berry and M. Tabor, 'Closed orbits and the regular bound spectrum', Proc. R. Soc. Lond. A 349, 101–123 (1976). DOI: 10.1098/rspa.1976.0062, https://doi.org/10.1098/rspa.1976.0062
2. M. V. Berry and M. Tabor, 'Calculating the bound spectrum by path summation in action-angle variables', J. Phys. A: Math. Gen. 10, 371–379 (1977). DOI: 10.1088/0305-4470/10/3/009, https://doi.org/10.1088/0305-4470/10/3/009
3. M. V. Berry and M. Tabor, 'Level clustering in the regular spectrum', Proc. R. Soc. Lond. A 356, 375–394 (1977). DOI: 10.1098/rspa.1977.0140, https://doi.org/10.1098/rspa.1977.0140
4. M. C. Gutzwiller, 'Periodic orbits and classical quantization conditions', J. Math. Phys. 12, 343–358 (1971). DOI: 10.1063/1.1665596, https://doi.org/10.1063/1.1665596
5. R. Balian and C. Bloch, 'Distribution of eigenfrequencies for the wave equation in a finite domain: III. Eigenfrequency density oscillations', Ann. Phys. 69, 76–160 (1972). DOI: 10.1016/0003-4916(72)90006-1, https://doi.org/10.1016/0003-4916(72)90006-1
6. M. V. Berry and K. E. Mount, 'Semiclassical approximations in wave mechanics', Rep. Prog. Phys. 35, 315–397 (1972). DOI: 10.1088/0034-4885/35/1/306, https://doi.org/10.1088/0034-4885/35/1/306
7. I. C. Percival, 'Regular and irregular spectra', J. Phys. B: At. Mol. Phys. 6, L229–L232 (1973). DOI: 10.1088/0022-3700/6/9/002, https://doi.org/10.1088/0022-3700/6/9/002
8. C. C. Martens and G. S. Ezra, 'EBK quantization of nonseparable systems: A Fourier transform method', J. Chem. Phys. 83, 2990 (1985). DOI: 10.1063/1.449202, https://doi.org/10.1063/1.449202
9. A. M. Ozorio de Almeida and J. H. Hannay, 'Resonant periodic orbits and the semiclassical energy spectrum', J. Phys. A: Math. Gen. 20, 5873–5883 (1987). DOI: 10.1088/0305-4470/20/17/021, https://doi.org/10.1088/0305-4470/20/17/021
10. J. P. Keating and M. V. Berry, 'False singularities in partial sums over closed orbits', J. Phys. A: Math. Gen. 20, L1139–L1141 (1987). DOI: 10.1088/0305-4470/20/17/002, https://doi.org/10.1088/0305-4470/20/17/002
11. J. M. Robbins, S. C. Creagh and R. G. Littlejohn, 'Complex periodic orbits in the rotational spectrum of molecules: The example of SF6', Phys. Rev. A 39, 2838–2854 (1989). DOI: 10.1103/PhysRevA.39.2838, https://doi.org/10.1103/PhysRevA.39.2838
12. S. C. Creagh and R. G. Littlejohn, 'Semiclassical trace formulas in the presence of continuous symmetries', Phys. Rev. A 44, 836–850 (1991). DOI: 10.1103/PhysRevA.44.836, https://doi.org/10.1103/PhysRevA.44.836
13. S. C. Creagh and R. G. Littlejohn, 'Semiclassical trace formulae for systems with non-Abelian symmetry', J. Phys. A: Math. Gen. 25, 1643–1669 (1992). DOI: 10.1088/0305-4470/25/6/021, https://doi.org/10.1088/0305-4470/25/6/021
14. H. Schomerus and M. Sieber, 'Bifurcations of periodic orbits and uniform approximations', J. Phys. A: Math. Gen. 30, 4537 (1997). DOI: 10.1088/0305-4470/30/13/010, https://doi.org/10.1088/0305-4470/30/13/010
15. M. Sieber, 'Semiclassical transition from an elliptical to an oval billiard', J. Phys. A: Math. Gen. 30, 4563 (1997). DOI: 10.1088/0305-4470/30/13/011, https://doi.org/10.1088/0305-4470/30/13/011
16. M. Sieber, 'Billiard systems in three dimensions: the boundary integral equation and the trace formula', Nonlinearity 11, 1607–1623 (1998). DOI: 10.1088/0951-7715/11/6/010, https://doi.org/10.1088/0951-7715/11/6/010
17. T. Bartsch, J. Main and G. Wunner, 'Uniform approximations for non-generic bifurcation scenarios including bifurcations of ghost orbits', Ann. Phys. 277, 19–73 (1999). DOI: 10.1006/aphy.1999.5961, https://doi.org/10.1006/aphy.1999.5961
18. J. Marklof, 'The Berry–Tabor conjecture', in European Congress of Mathematics (Progress in Mathematics 202, Birkhäuser, 2001). DOI: 10.1007/978-3-0348-8266-8_36, https://doi.org/10.1007/978-3-0348-8266-8_36
19. K. Arita and M. Brack, 'Normal forms and uniform approximations for bridge orbit bifurcations', J. Phys. A: Math. Theor. 41, 385207 (2008). DOI: 10.1088/1751-8113/41/38/385207, https://doi.org/10.1088/1751-8113/41/38/385207
20. J. D. Urbina, M. Kelly and K. Richter, 'Periodic orbit theory of Bethe-integrable quantum systems: an N-particle Berry–Tabor trace formula', J. Phys. A: Math. Theor. 56, 214001 (2023). DOI: 10.1088/1751-8121/accee6, https://doi.org/10.1088/1751-8121/accee6