Feeds:
Posts
Consider for a moment the validity of the classical-quantum link used by people in geometric quantization.  Suppose the space is a riemannian manifold $M$.  The classical Hamiltonian of a free particle is $H(p,q) = 1/2 \| q \|^2$ and the quantum Hamiltonian is $1/2 \Delta_M$ using the Laplace-Beltrami operator.  So the energy spectrum for the quantum case are the eigenfunctions of the Laplacian.  Thus the ‘quantum wavefunctions’ are derived from the eigenfunctions of the Laplacian.  One thing to keep in mind with these considerations is that quasimodes of the Laplacian concentrate near the periodic geodesics and in fact the flow of the classical Hamiltonian in the case above is the geodesic flow.  In the case where $M$ is a sphere, and unlike the case of manifolds with negative curvature, the geodesic flow is completely integrable.  For a sphere more holds true: every geodesic is closed of the same length.
Now on a 4-sphere of radius $1/h$, the eigenfunctions of the Laplacian are the spherical harmonics and it is known also that the natural Dirac operator has spectrum $k h$ for $k \in \mathbf{Z}$.  There is a problem to be resolved however, which is that if the photons are spherical harmonics there is a problem with localization.  The particle-quality of a photon needs to be clear in accordance with the Compton effect experiments of photons.