A frequency comb is a light source whose lines are evenly spaced. Their regularity and ability to make precise measurements have revolutionized many fields, and for that reason the Nobel Prize in Physics was awarded for its development. Until recently, frequency combs were primarily based on mode-locked lasers, and could not be used in applications requiring compactness. However, this changed with the introduction of semiconductor frequency combs, which use a technology similar to what you might find in a laser pointer.
Although it is possible to naturally form combs in semiconductor lasers, this mode of operation had not been shown to occur reliably or predictably over a wide dynamic range, since the dispersion could not be controlled by material growth to the necessary precision. We were the first to demonstrate that the concept of dispersion engineering, which had previously been well-established in ultrafast optics, was also valuable for semiconductor quantum cascade lasers. In doing this, we were able to demonstrate the first laser-based terahertz combs, and also the first deliberately-designed combs in semiconductor lasers. We also showed how the temporal profile of these combs could be directly measured, solving a problem that had been standing in the field since 2000.
At long wavelengths, an exciting application of frequency combs is a technique known as dual comb spectroscopy. Spectrometers at mid-infrared and terahertz wavelengths are useful because they can be used to “fingerprint” various chemicals, picking out and isolating the concentrations of many different species. Unfortunately, these spectrometers tend to be large and relegated to the lab. Using the dual comb technique, it is possible to perform spectroscopy without any moving parts, reducing the size of spectroscopic systems from meters to millimeters. We were able to use our combs to perform the first compact dual comb spectroscopy in the terahertz range, and were able to do this using only chip-scale components.
In the course of carrying out this dual comb work, we realized that one of the major limitations of the dual comb technique was that despite its great promise, it had very challenging stability requirements, requirements that would effectively preclude widespread adoption. We decided to take an unconventional approach to this problem, and showed that by using proper signal processing, the dual comb signal alone contained enough information to be corrected entirely computationally. This approach circumvents the usual requirements for temperature stability, optical isolation, and bias stability usually associated with these sorts of systems. (Demo code is available here.)
One of our earliest projects was to use terahertz time-domain spectroscopy to study the dynamics of terahertz quantum cascade lasers and to probe their nanostructures. The fascinating aspect about the terahertz frequency range is that because it lies in the netherworld between electronics and photonics, it is possible to use techniques from each domain to do interesting things. For example, a single device can be made to generate terahertz radiation via a switch—a traditional electronic element—or via stimulated emission—a traditional photonic element.
By combining each of these elements and probing the quantum cascade laser with ultrashort optical pulses, we were able to glean information about the artificial band structure and about the dynamics of these devices. We also used these techniques to perform the first measurements of the dynamics of the group velocity dispersion of quantum cascade lasers, which is important for a large range of nonlinear devices. Quantum cascade lasers are a unique platform for many new types of nonlinear devices, on account of their enormous χ(2) and χ(3) nonlinearities.