Phonon-polaritons propagating in crystal volume offer the possibility of transferring information throughout matter (via phonons) at high (photon-like) velocity and tunable frequency/wavelength in the far-infrared. However, from the phonon-polariton Raman cross section, the phonon and photon advantages seem mutually exclusive. Either the phonon-polaritons are fast (photon-like) but hardly supported by the lattice (not phonon-like), or they are well supported by the lattice (testified by a high Raman efficiency) but slow (not photon-like). An optimal phonon–photon coupling is currently being searched for in hexagonal GaN by near-forward Raman scattering across parallel crystal faces with an in-plane singular c→-axis. Two accessible phonons, i.e., the ordinary A1 and E1 ones, plus two refractive indices, i.e., the ordinary and extraordinary ones, generate various phonon-polariton candidates. Notably, in perfect forward Raman scattering and in crossed polarizations of the incident (∥c→) and scattered lights, which, altogether, maximizes the transferred wavevector to the crystal at minimum scattering angle, a fast phonon-polariton, stemming from the bottleneck of the E1 dispersion on entry to the deep photon-like regime, is activated. It is well supported by the lattice since its Raman signal is strong and sharp, enhanced by multi-reflection of the laser beam between crystal faces at near-normal incidence. This fast Raman-enhanced phonon-polariton is interesting for infrared photonics in that it cumulates the advantages of a photon (speed) and of a phonon (Raman intensity). Besides, it commutates from a phonon-polariton to a phonon by deviating from normal incidence or by permuting the incident and scattered polarizations, with potential applications as a vibrational switch.
Malekzadeh et al. (Mon,) studied this question.