Photonic local oscillators and modulators that include whispering-gallery- mode (WGM) optical resonators have been proposed as power-efficient devices for generating and detecting radiation at frequencies of the order of a terahertz. These devices are intended especially to satisfy anticipated needs for receivers capable of detecting low-power, narrow-band terahertz signals to be used for sensing substances of interest in scientific and military applications. At present, available terahertz- signal detectors are power-inefficient and do not afford the spectral and amplitude resolution needed for detecting such signals.
The proposed devices would not be designed according to the conventional approach of direct detection of terahertz radiation. Instead, terahertz radiation would first be up-converted into the optical domain, wherein signals could be processed efficiently by photonic means and detected by optical photodetectors, which are more efficient than are photodetectors used in conventional direct detection of terahertz radiation. The photonic devices used to effect the upconversion would include a tunable optical local oscillator and a novel electrooptical modulator.
A local oscillator according to the proposal would be a WGM-based modelocked laser operating at a desired pulserepetition rate of the order of a terahertz. The oscillator would include a terahertz optical filter based on a WGM microresonator, a fiber-optic delay line, an optical amplifier (which could be either a semiconductor optical amplifier or an erbiumdoped optical fiberamplifier), and a WGM Ka-band modulator (see figure). The terahertz repetition rate would be obtained through harmonic mode locking: for example, by modulating the light at a frequency of 33 GHz and locking each 33d optical mode, one would create a 1.089-THz pulse train. The high resonance quality factors (Q values) of WGM optical resonators should make it possible to decrease signal-generation threshold power levels significantly below those of other optical-signal-generation devices.
An electro-optical modulator as proposed would be a triply resonant compound WGM device. The modulator would comprise a periodically poled LiNbO3 WGM optical resonator stacked almost in contact with an Si WGM terahertz resonator. It would be necessary to use these two resonators because LiNbO3 would afford the needed combination of high Q for the optical modes and enough nonlinearity for efficient interaction between light and terahertz radiation, while Si would afford the needed high Q for terahertz radiation.
Because Si absorbs light, it would be necessary to minimize penetration of light into the Si resonator. Because LiNbO3 absorbs terahertz radiation more than Si does, the portion of the LiNbO3 volume wherein the light and the terahertz radiation interact should be less than the volume of the terahertz mode. These requirements would be satisfied by, among other things, positioning the two resonators with a gap of ≈1 μm between them and utilizing evanescent-field coupling between the light and the terahertz radiation. The periodicity of the poling of the LiNbO3 would be chosen to ensure the required matching of phases between the light and the terahertz radiation.