Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Electromagnetic Modes in Spherical Cavities: Complete Theory of Angular Spectra, Dispersion Relations, and Self-Adjoint Extensions

Mustafa Bakr, Tongyu Zhang, Smain Amari

Abstract

We present a complete theory of electromagnetic modes in spherical cavities, resolving fundamental questions about the nature of angular quantization. The standard result that angular indices $(\ell,m)$ must be integers is shown to be a consequence of domain constraints -- regularity at both poles and single-valuedness in the azimuthal coordinate -- rather than a requirement imposed by Maxwell's equations themselves. We prove that, for the sectoral case $ν=m$, the function $\sin^{m}θ$ exactly solves the angular eigenvalue equation for any real $m>0$, giving rise to a continuous dispersion curve. We demonstrate why non-sectoral modes (tesseral and zonal) appear only at isolated integer points on the full sphere, and show how boundary modifications such as cones and wedges convert these isolated points into continuous families of modes. Complete field solutions, wave impedances, and energy integrability conditions are derived. At the limiting point $(ν, m) = (0, 0)$, the electromagnetic field vanishes identically while the underlying Debye potential remains non-trivial -- a distinction with implications for mode counting that connects to longstanding questions in gauge theory and cavity quantization. Full-wave simulations validate the theoretical predictions with sub-percent accuracy. These results raise the possibility of structural analogues in wave equations on curved spacetimes, where conical deficits or horizon excisions similarly modify the angular domain.

Electromagnetic Modes in Spherical Cavities: Complete Theory of Angular Spectra, Dispersion Relations, and Self-Adjoint Extensions

Abstract

We present a complete theory of electromagnetic modes in spherical cavities, resolving fundamental questions about the nature of angular quantization. The standard result that angular indices must be integers is shown to be a consequence of domain constraints -- regularity at both poles and single-valuedness in the azimuthal coordinate -- rather than a requirement imposed by Maxwell's equations themselves. We prove that, for the sectoral case , the function exactly solves the angular eigenvalue equation for any real , giving rise to a continuous dispersion curve. We demonstrate why non-sectoral modes (tesseral and zonal) appear only at isolated integer points on the full sphere, and show how boundary modifications such as cones and wedges convert these isolated points into continuous families of modes. Complete field solutions, wave impedances, and energy integrability conditions are derived. At the limiting point , the electromagnetic field vanishes identically while the underlying Debye potential remains non-trivial -- a distinction with implications for mode counting that connects to longstanding questions in gauge theory and cavity quantization. Full-wave simulations validate the theoretical predictions with sub-percent accuracy. These results raise the possibility of structural analogues in wave equations on curved spacetimes, where conical deficits or horizon excisions similarly modify the angular domain.

Paper Structure

This paper contains 57 sections, 2 theorems, 97 equations, 5 figures, 4 tables.

Table of Contents

  1. Introduction and Motivation
  2. The Spectral Landscape
  3. Azimuthal Propagation and the Quantization of $m$
  4. Polar Regularity and the Constraint on $\nu$
  5. The Three Mode Families
  6. Independence of the Two Constraints $m$ and $\nu$
  7. Comparison with the Cylindrical Case
  8. Summary of Contributions
  9. Organization of This Paper
  10. Mathematical Foundations
  11. The Separated Equations
  12. Frobenius Analysis of the Angular Equation
  13. The Sectoral Case: Exact Solution for $\nu = m$
  14. Non-Sectoral Cases: The Origin of Discreteness
  15. Graphical Representation of the Spectral Landscape
  16. ...and 42 more sections

Key Result

Theorem 2.1

For $\nu = m$ with $m$ any positive real number, the function is an exact solution of the associated Legendre equation eq:angular_ode. This solution vanishes at both poles $\theta = 0$ and $\theta = \pi$ and is everywhere positive on the open interval $(0, \pi)$.

Figures (5)

  • Figure 1: Spectral landscape in the $(\nu, m)$ parameter plane. The sectoral line $\nu = m$ (solid line with filled circles) supports solutions for any real $m > 0$, forming a continuous branch. Zonal modes (open squares, $m = 0$) and tesseral modes (filled triangles, $\nu > m > 0$) exist only at integer points. The region $\nu < m$ (shaded) corresponds to parameters for which no square-integrable solution exists on the full sphere; the physical implications of this regime for restricted domains remain unexplored. The point $(0,0)$ marks the boundary of the sectoral family; here the electromagnetic field vanishes while the underlying potential remains non-trivial (see Section \ref{['sec:null']}).
  • Figure 2: Dispersion curves showing the first TE and TM roots as functions of $\nu$. The TE boundary condition $j_\nu(ka) = 0$ (blue) and TM boundary condition $[xj_\nu(x)]' = 0$ (orange) define continuous curves. Filled circles mark integer values corresponding to standard spherical cavity resonances.
  • Figure 3: Spherical cavity geometries and their spectral characteristics, shown as cross-sections. (a) Full sphere (meridional view): both $\nu$ and $m$ restricted to integers. (b) Azimuthal wedge (equatorial view from above): removing the angular region $\phi > \Phi$ permits non-integer $m = n\pi/\Phi$; hatched region shows removed portion. (c) Single cone at half-angle $\theta_c$: removes the polar cap and allows continuous $\nu(\theta_c) > 0$. (d) Double cone (biconical): symmetric cones at both poles. (e) Cone with wedge: independent control of $\nu$ via cone angle and $m$ via wedge angle; hatched shows removed azimuthal section. (f) Biconical with wedge: full two-parameter access to continuous $(\nu, m)$.
  • Figure 4: Fundamental TM frequency versus blue cavity opening angle $\Phi$. The minimum near $\Phi = 270^\circ$ (corresponding to a $90^\circ$ wedge) arises because this geometry permits $m = 2/3 < 1$, sampling the continuous dispersion curve below the first integer mode. The $\Phi = 180^\circ$ geometry (hemisphere) enforces $m \geq 1$, yielding higher frequencies despite smaller volume. Theory (solid line with circles) computed from Eq. \ref{['eq:m_quantization']} and radial TM quantization; HFSS data (squares) agrees within $1\%$.
  • Figure 5: Fundamental TM frequency (left axis) and angular eigenvalue $\nu$ (right axis) versus cone half-angle $\theta_c$ for a sphere with a single north-pole cone and no wedge ($m = 0$). The cone removes the $\theta = 0$ regularity constraint, allowing access to the continuous branch with $\nu < 1$. Theory curves are computed by solving $P_\nu(\cos\theta_c) = 0$ for $\nu$, followed by TM radial quantization; HFSS data (symbols) agree within $0.7\%$. The full sphere ($\theta_c \to 0$) would require $\nu \geq 1$, yielding $f = 8.73~\mathrm{GHz}$—approximately $40\%$ higher than the cone-modified cavity.

Theorems & Definitions (11)

  • Theorem 2.1: Sectoral Solution
  • proof
  • Theorem 2.2: Spectral Structure on the Full Sphere
  • Remark 3.1: Potential versus field
  • Remark 5.1: Consistency with numerical simulations
  • Remark 5.2: Spherical Neumann function exclusion
  • Remark 5.3: Zonal modes with non-integer $\nu$
  • Remark 6.1: Tesseral modes for non-integer $m$
  • Remark 6.2: Comparison with all-integer case
  • Remark 6.3: Relaxation of the $\nu \geq m$ constraint
  • ...and 1 more