September 2016
Spotlight Summary by Thomas Udem
Generation of a frequency comb spanning more than 3.6 octaves from ultraviolet to mid infrared
Self-referenced frequency combs have been around for almost two decades and have found numerous applications in the physical sciences. Initially, they were designed to measure the frequency of continuous wave (CW) lasers used to perform spectroscopy of a narrow transition line, either of fundamental interest or as a reference oscillator for an all-optical clock. It was quickly realized that frequency combs may also be used for direct frequency comb spectroscopy (DFCS). This may seem not to have an advantage because the large number of spectator modes gives rise to an elevated background as well as AC Stark shifts, without contributing to the signal. One exception is the two-photon transition, where the modes contribute pairwise to the excitation energy. The sum of all modes contributes to the signal such that the signal-to-AC Stark shift and signal-to-background ratio is the same as for a CW laser of the same average power. In the time domain, though, the frequency comb corresponds to a pulse train with very large peak intensities that can be converted to a much shorter wavelength through nonlinear effects. Using the process of high harmonic generation, even the hard X-ray regime has been reached.
A real-world application where DFCS has a clear advantage arises when it comes to probing broad multi-component spectra, such as those of molecules. The large number of modes that make up the comb allows for massive parallel measurements as in Fourier transform spectroscopy. In contrast to the latter, DFCS has several advantages. A broadband frequency comb represents a “white-light laser”: high power spectral density and spatial coherence at the same time. Therefore interference can be obtained with high contrast. Unlike the usual thermal white-light source used in Fourier transform spectroscopy, the frequency comb can be efficiently coupled into a cavity in order to resonantly enhance the power. This allows for higher sensitivity and opens possibilities for “tricks” such as Vernier spectroscopy. Using two frequency combs with slightly different repetition rates, a spectrogram can be obtained within microseconds without any moving parts. Monitoring rapidly changing broad spectra, such as in chemical reactions, becomes a possibility. For these applications, single-mode resolution can be reached, but is usually not required.
The usefulness of broadband DFCS depends on the optical bandwidth that the frequency comb covers. For trace gas analysis, the so-called fingerprint region (~6-25 µm) is important, because this is where organic compounds have well-distinguishable spectral features. This region is notoriously difficult to reach with lasers, therefore the red-shifting Raman effect may be employed also in conjunction with DFCS. Now Iwakuni and co-workers have taken the next step in extending the possibilities of DFCS by demonstrating a frequency comb that covers 0.35-4.4 µm: 3.6 optical octaves, the broadest frequency comb yet. Even though it just misses the fingerprint region, it will still be useful for certain types of molecular spectroscopy. Spectral broadening often goes along with loss of coherence, which could be avoided here. More research is required to fully understand the spectral broadening that takes place in the waveguide. As of now it seems clear that both χ(2) and χ(3) processes take place in terms of sum-frequency generation, difference-frequency generation and self-phase modulation. This means that different spectral regions of the frequency comb have different carrier-envelope offset (CEO) frequencies. This makes it more difficult to interpret DFCS spectra when single-mode resolution is required, but it also allows to directly detect the CEO frequency, as the authors demonstrate. Maybe the most appealing property of their set-up is its simple and robust design with almost no free-space optics; this should allow for applications outside the laboratory.
You must log in to add comments.
A real-world application where DFCS has a clear advantage arises when it comes to probing broad multi-component spectra, such as those of molecules. The large number of modes that make up the comb allows for massive parallel measurements as in Fourier transform spectroscopy. In contrast to the latter, DFCS has several advantages. A broadband frequency comb represents a “white-light laser”: high power spectral density and spatial coherence at the same time. Therefore interference can be obtained with high contrast. Unlike the usual thermal white-light source used in Fourier transform spectroscopy, the frequency comb can be efficiently coupled into a cavity in order to resonantly enhance the power. This allows for higher sensitivity and opens possibilities for “tricks” such as Vernier spectroscopy. Using two frequency combs with slightly different repetition rates, a spectrogram can be obtained within microseconds without any moving parts. Monitoring rapidly changing broad spectra, such as in chemical reactions, becomes a possibility. For these applications, single-mode resolution can be reached, but is usually not required.
The usefulness of broadband DFCS depends on the optical bandwidth that the frequency comb covers. For trace gas analysis, the so-called fingerprint region (~6-25 µm) is important, because this is where organic compounds have well-distinguishable spectral features. This region is notoriously difficult to reach with lasers, therefore the red-shifting Raman effect may be employed also in conjunction with DFCS. Now Iwakuni and co-workers have taken the next step in extending the possibilities of DFCS by demonstrating a frequency comb that covers 0.35-4.4 µm: 3.6 optical octaves, the broadest frequency comb yet. Even though it just misses the fingerprint region, it will still be useful for certain types of molecular spectroscopy. Spectral broadening often goes along with loss of coherence, which could be avoided here. More research is required to fully understand the spectral broadening that takes place in the waveguide. As of now it seems clear that both χ(2) and χ(3) processes take place in terms of sum-frequency generation, difference-frequency generation and self-phase modulation. This means that different spectral regions of the frequency comb have different carrier-envelope offset (CEO) frequencies. This makes it more difficult to interpret DFCS spectra when single-mode resolution is required, but it also allows to directly detect the CEO frequency, as the authors demonstrate. Maybe the most appealing property of their set-up is its simple and robust design with almost no free-space optics; this should allow for applications outside the laboratory.
Add Comment
You must log in to add comments.
Article Information
Generation of a frequency comb spanning more than 3.6 octaves from ultraviolet to mid infrared
Kana Iwakuni, Sho Okubo, Osamu Tadanaga, Hajime Inaba, Atsushi Onae, Feng-Lei Hong, and Hiroyuki Sasada
Opt. Lett. 41(17) 3980-3983 (2016) View: Abstract | HTML | PDF