Abstract

Fifty years ago, S.E. Harris [1] proposed a special class of optical parametric oscillators (OPOs), the so-called mirrorless optical parametric oscillators (MOPOs), in which the signal and idler waves are generated in opposite directions. The MOPOs do not require reflected waves or a cavity to establish the positive feedback for oscillation. Instead, the feedback is accomplished by the counter-propagating nonlinear interaction itself. Indeed, it is the counter-propagating phase matching scheme that intrinsically provides the MOPOs with unique spectral properties, namely, generation of intrinsically narrowband counter-propagating idler and a co-propagating signal spectrum of which is a frequency-translated version of the spectrum of the pump. [2-3]. These properties cannot be achieved in conventional OPOs without complex set-ups and components. The large phase mismatch of the counter-propagating waves demands an extraordinarily large birefringence of the nonlinear optical material, which makes quasi-phase matching (QPM) the most feasible route to achieve MOPO. However, MOPOs require the QPM period to be in the sub-micrometer range, which is a challenging task. Thus, it is not surprising that the MOPO had to wait 41 years for its first experimental demonstration [2]. Nevertheless, up-to-now, all the MOPO demonstrations have been done with broadband ps-pumping. This can be ascribed to the fact that the efficiency and threshold of the MOPO depend on the quality and length of the sub-µm QPM grating.

© 2017 IEEE