March 2015
Spotlight Summary by Thomas Udem
Direct 15-GHz mode-spacing optical frequency comb with a Kerr-lens mode-locked Yb:Y2O3 ceramic laser
Mode-locked lasers that emit ultrashort pulses with a very large repetition rate are highly desirable for certain frequency comb applications. The team of Yohei Kobayashi at the University of Tokyo has now been able to push this endeavour further by setting up an Yb:Y2O3 ceramic ring laser that generates pulses of only 152 fs duration with a repetition rate of 15 GHz. The cavity round trip length of their laser is less than 3 cm.
Even though such a laser could be useful for several applications, let me pick one where it seems to fill in a serious gap in laser parameter space. Mode-locked lasers with a repetition rate that exceeds 10 GHz and with a pulse duration around 100 fs have not been possible so far. Such a laser is extremely useful as a frequency comb to calibrate conventional spectrometers. These spectrometers are in use when laser spectroscopy is inapplicable. The most prominent examples are in astronomy. Extra-solar planets are detected by a periodic Doppler shift of the Fraunhofer lines of the recoiling host star. To detect an Earth-like planet in an orbit around a Sun-like star, one needs to calibrate a velocity variation of v=9 cm/sec over the course of one year. This corresponds to a relative calibration reproducibility of Δν⁄ν = v⁄c = 3×10-10 that has not yet been achieved. The large spacing frequency combs could be the solution to this calibration problem. Other applications in precision astronomy also benefit, like the search for possible temporal variation of fundamental constants. The real-time observation of the acceleration of the Hubble expansion due to the predicted prevalence of dark energy would be another one that is even more demanding.
For several other reasons, like background and sensitivity, the optimum resolution of astronomical spectrographs is around 10 GHz. Astronomers are not going to change that. The comb mode spacing must be somewhat larger while at the same time provide enough optical bandwidth. Unfortunately, these are conflicting requirements because a large bandwidth is usually obtained by spectral broadening and thus requires short pulses in combination with large pulse energy. The latter is inversely proportional to the repetition rate. The decisive advantage of an Yb system is that one can readily amplify even pJ optical pulses in optical fibers to provide the pulse energy for spectral broadening. After frequency doubling one reaches the visible region where the most useful Fraunhofer lines are located.
So far the strategy has been to operate the laser at a much lower repetition rate and use several Fabry-Pérot cavities as harmonic mode filters to select modes with a larger spacing. It turns out that subsequent spectral broadening and frequency doubling re-amplifies the unwanted suppressed modes so that the filtering must have a very large suppression ratio. This requires several Fabry-Pérot filter cavities in series and significantly increases system complexity. Alternative approaches have not been successful so far. Very short cavity semiconductor lasers do not reach pulses short enough and suffer from large phase noise. In harmonic mode-locked lasers several pulses are simultaneously stored in the cavity. To give rise to a large and regular mode spacing in the frequency domain, the pulses need to maintain their distances and phases extremely well. Otherwise residual modes, with a spacing given by the fundamental cavity round trip frequency, will emerge that will be re-amplified along with the spectral broadening.
The new laser introduced in this work seem to solve all these issues at once, but is not without competition. Micro resonators as well as newly developed EOM based comb generators allow very large mode spacings. Similar lasers based on titanium sapphire have been demonstrated but do not exceed 10 GHz yet. Unlike those the new Yb:Y2O3 ceramic ring laser is content with a reliable semiconductor pump laser. Therefore one might hope that this system could be engineered to be operated unattended for extended times as required for a reliable calibrator at an astronomical observatory.
You must log in to add comments.
Even though such a laser could be useful for several applications, let me pick one where it seems to fill in a serious gap in laser parameter space. Mode-locked lasers with a repetition rate that exceeds 10 GHz and with a pulse duration around 100 fs have not been possible so far. Such a laser is extremely useful as a frequency comb to calibrate conventional spectrometers. These spectrometers are in use when laser spectroscopy is inapplicable. The most prominent examples are in astronomy. Extra-solar planets are detected by a periodic Doppler shift of the Fraunhofer lines of the recoiling host star. To detect an Earth-like planet in an orbit around a Sun-like star, one needs to calibrate a velocity variation of v=9 cm/sec over the course of one year. This corresponds to a relative calibration reproducibility of Δν⁄ν = v⁄c = 3×10-10 that has not yet been achieved. The large spacing frequency combs could be the solution to this calibration problem. Other applications in precision astronomy also benefit, like the search for possible temporal variation of fundamental constants. The real-time observation of the acceleration of the Hubble expansion due to the predicted prevalence of dark energy would be another one that is even more demanding.
For several other reasons, like background and sensitivity, the optimum resolution of astronomical spectrographs is around 10 GHz. Astronomers are not going to change that. The comb mode spacing must be somewhat larger while at the same time provide enough optical bandwidth. Unfortunately, these are conflicting requirements because a large bandwidth is usually obtained by spectral broadening and thus requires short pulses in combination with large pulse energy. The latter is inversely proportional to the repetition rate. The decisive advantage of an Yb system is that one can readily amplify even pJ optical pulses in optical fibers to provide the pulse energy for spectral broadening. After frequency doubling one reaches the visible region where the most useful Fraunhofer lines are located.
So far the strategy has been to operate the laser at a much lower repetition rate and use several Fabry-Pérot cavities as harmonic mode filters to select modes with a larger spacing. It turns out that subsequent spectral broadening and frequency doubling re-amplifies the unwanted suppressed modes so that the filtering must have a very large suppression ratio. This requires several Fabry-Pérot filter cavities in series and significantly increases system complexity. Alternative approaches have not been successful so far. Very short cavity semiconductor lasers do not reach pulses short enough and suffer from large phase noise. In harmonic mode-locked lasers several pulses are simultaneously stored in the cavity. To give rise to a large and regular mode spacing in the frequency domain, the pulses need to maintain their distances and phases extremely well. Otherwise residual modes, with a spacing given by the fundamental cavity round trip frequency, will emerge that will be re-amplified along with the spectral broadening.
The new laser introduced in this work seem to solve all these issues at once, but is not without competition. Micro resonators as well as newly developed EOM based comb generators allow very large mode spacings. Similar lasers based on titanium sapphire have been demonstrated but do not exceed 10 GHz yet. Unlike those the new Yb:Y2O3 ceramic ring laser is content with a reliable semiconductor pump laser. Therefore one might hope that this system could be engineered to be operated unattended for extended times as required for a reliable calibrator at an astronomical observatory.
Add Comment
You must log in to add comments.
Article Information
Direct 15-GHz mode-spacing optical frequency comb with a Kerr-lens mode-locked Yb:Y2O3 ceramic laser
Mamoru Endo, Isao Ito, and Yohei Kobayashi
Opt. Express 23(2) 1276-1282 (2015) View: Abstract | HTML | PDF