August 2011
Spotlight Summary by Francesco Morichetti
Silicon-based monolithic optical frequency comb source
Optical frequency combs made of a large number of precisely spaced and sharp spectral lines distributed over a broad spectral range are becoming more and more attractive tools in many different fields. Optical clocks and precision spectroscopy were among the first applications, but more recently the use of optical combs has been proposed in many other emerging areas, including ultra-high-speed communications systems, optical waveform and microwave signal synthesis, and astronomical spectrographs calibration.
There is a traditional and well-established way to generate optical frequency combs that makes use of ultrafast mode-locked lasers. While being accurate, efficient, and reliable for commercial deployment, the main drawback of these systems is the need for bulky and expensive optical equipment. Moreover, the maximum achievable line separation of optical combs generated by mode-locked lasers hardly exceeds the range of 10 GHz.
The “comb-generator dream” is now a compact, low-cost and high repetition rate device. In recent years a radically new concept has been proposed, which represents a fundamental step toward this goal. The basic principle is parametric frequency conversion in an integrated optical microresonator with a high resonance quality factor (Q) and more specifically a powerful combination of optical parametric amplification and cavity-enhanced four-wave mixing (FWM). Impressive results have been achieved with silica microtoroidal resonators, demonstrating the generation of spectral bandwidths of up to 150 THz (octave-spanning) with a comb spacing of 850 GHz, equidistant to less than 10-17 relative to the optical frequency.
So, has the comb-generator dream already come to reality? Unfortunately, not yet, because microtoroidal resonators do not lack weaknesses in themselves. Although the fabrication technology is well assessed, the optical coupling with the external world is seriously tricky. Tapered optical fibers need to be placed and held at a precise distance from the resonator to achieve the desired Q-factor of the loaded cavity, as the mechanical stability and reliability of this coupling system is a critical issue.
A powerful solution is the monolithic integration of the resonator and the optical waveguide injecting and extracting the light from the same. This is the strategy used in the work by M. A. Foster and co-workers, who realized a broadband optical frequency comb generator exploiting optical parametric oscillation in a CMOS-compatible integrated microring resonator. The ring resonator and the bus waveguide were fabricated monolithically in a single silicon nitride layer using electron-beam lithography and subsequently covered with a silica film. The direct integration of the entire device on the same chip-scale platform allows fabrication of an extremely robust, compact, and sealed device, with coupling and operation insensitive to the surrounding environment.
Broadband operation is achieved by fully exploiting a further advantage of photonic integrated technologies, that is, the possibility of optimizing the design of the optical waveguide to produce a suitable (anomalous) dispersion. In this way the generation of more than 350 comb lines spanning over 75 THz (1375 to 2100 nm) and spaced by 204 GHz was demonstrated, with a mode equidistance, measured over a 14.5 THz span, better than 3×10-15 relative to the optical frequency.
A skeptical reader could argue that the optical Q factor of microring resonators is generally much lower than that of state-of-the-art silica microtoroids. That is correct; indeed the SiN resonator employed by M. A. Turner and coworkers has a Q factor of only 3×105. However, the Q-factor reduction is largely compensated by the tight confinement of the field inside the microresonators and the higher nonlinearity of SiN, which is almost one order of magnitude higher than in silica. As a result, no significant reduction in the efficiency of the comb generator is observed with respect to microtoroid-based solutions.
It is difficult to predict if and when the technology of comb generators based on microresonators will be mature enough to replace current laser-based frequency comb generators. The perception is now that microresonators can provide a much easier access to high repetition rates, in the range of 10 to 1000 GHz, through compact, low-cost, chip-scale integrated devices. This is what many applications are waiting for, especially in astronomy, microwave photonics, or telecommunications fields. And maybe, this could be the key to let the comb-generator dream become reality.
You must log in to add comments.
There is a traditional and well-established way to generate optical frequency combs that makes use of ultrafast mode-locked lasers. While being accurate, efficient, and reliable for commercial deployment, the main drawback of these systems is the need for bulky and expensive optical equipment. Moreover, the maximum achievable line separation of optical combs generated by mode-locked lasers hardly exceeds the range of 10 GHz.
The “comb-generator dream” is now a compact, low-cost and high repetition rate device. In recent years a radically new concept has been proposed, which represents a fundamental step toward this goal. The basic principle is parametric frequency conversion in an integrated optical microresonator with a high resonance quality factor (Q) and more specifically a powerful combination of optical parametric amplification and cavity-enhanced four-wave mixing (FWM). Impressive results have been achieved with silica microtoroidal resonators, demonstrating the generation of spectral bandwidths of up to 150 THz (octave-spanning) with a comb spacing of 850 GHz, equidistant to less than 10-17 relative to the optical frequency.
So, has the comb-generator dream already come to reality? Unfortunately, not yet, because microtoroidal resonators do not lack weaknesses in themselves. Although the fabrication technology is well assessed, the optical coupling with the external world is seriously tricky. Tapered optical fibers need to be placed and held at a precise distance from the resonator to achieve the desired Q-factor of the loaded cavity, as the mechanical stability and reliability of this coupling system is a critical issue.
A powerful solution is the monolithic integration of the resonator and the optical waveguide injecting and extracting the light from the same. This is the strategy used in the work by M. A. Foster and co-workers, who realized a broadband optical frequency comb generator exploiting optical parametric oscillation in a CMOS-compatible integrated microring resonator. The ring resonator and the bus waveguide were fabricated monolithically in a single silicon nitride layer using electron-beam lithography and subsequently covered with a silica film. The direct integration of the entire device on the same chip-scale platform allows fabrication of an extremely robust, compact, and sealed device, with coupling and operation insensitive to the surrounding environment.
Broadband operation is achieved by fully exploiting a further advantage of photonic integrated technologies, that is, the possibility of optimizing the design of the optical waveguide to produce a suitable (anomalous) dispersion. In this way the generation of more than 350 comb lines spanning over 75 THz (1375 to 2100 nm) and spaced by 204 GHz was demonstrated, with a mode equidistance, measured over a 14.5 THz span, better than 3×10-15 relative to the optical frequency.
A skeptical reader could argue that the optical Q factor of microring resonators is generally much lower than that of state-of-the-art silica microtoroids. That is correct; indeed the SiN resonator employed by M. A. Turner and coworkers has a Q factor of only 3×105. However, the Q-factor reduction is largely compensated by the tight confinement of the field inside the microresonators and the higher nonlinearity of SiN, which is almost one order of magnitude higher than in silica. As a result, no significant reduction in the efficiency of the comb generator is observed with respect to microtoroid-based solutions.
It is difficult to predict if and when the technology of comb generators based on microresonators will be mature enough to replace current laser-based frequency comb generators. The perception is now that microresonators can provide a much easier access to high repetition rates, in the range of 10 to 1000 GHz, through compact, low-cost, chip-scale integrated devices. This is what many applications are waiting for, especially in astronomy, microwave photonics, or telecommunications fields. And maybe, this could be the key to let the comb-generator dream become reality.
Add Comment
You must log in to add comments.
Article Information
Silicon-based monolithic optical frequency comb source
Mark A. Foster, Jacob S. Levy, Onur Kuzucu, Kasturi Saha, Michal Lipson, and Alexander L. Gaeta
Opt. Express 19(15) 14233-14239 (2011) View: Abstract | HTML | PDF