Researchers have developed a time-stretch spectroscopy method that resolves non-repetitive spectral events separated by 3 picoseconds, improving temporal resolution by almost three orders of magnitude over previous dispersive Fourier transformation systems.
Dispersive Fourier transformation converts the spectrum of an optical pulse into a time-domain waveform. Chromatic dispersion stretches the pulse so its different wavelengths arrive at different times, allowing the spectrum to be recorded with a single high-speed photodetector rather than a detector array.
This approach can continuously record rapid, unpredictable changes that conventional scanning instruments may miss. However, sufficient pulse stretching is required to maintain spectral resolution. When consecutive pulses arrive closer together than the duration of the stretched waveform, they overlap and create temporal aliasing, concealing the spectrum of each event.
Researchers addressed this problem by storing a sequence of ultrashort pulses in an active optical cavity. The cavity created replicas of the original pulse sequence, separated by its round-trip time. An asynchronous pulse picker then selected a different stored pulse during each circulation.
This process converted an ultrafast pulse sequence into a slower series that could be analyzed one pulse at a time while preserving the original order. Each selected pulse was subsequently measured using a conventional dispersive Fourier transformation system.
The researchers evaluated two pulse-selection methods. The first used a 40-GHz electro-optic intensity modulator to generate 30-picosecond temporal gates. Combined with a fiber storage cavity with a 9.3-nanosecond round-trip time, the system preserved spectral information over hundreds of cavity circulations.
Using this configuration, the team reconstructed spectral changes in mode-locked laser pulses and analyzed modulation instability in a 25-GHz electro-optic frequency comb. It recovered the spectra of neighboring pulses separated by 40 picoseconds that could not be resolved by conventional time-stretch spectroscopy.
The second configuration replaced the electro-optic modulator with an all-optical pulse picker based on optical parametric amplification. This enabled the system to distinguish spectra separated by 3 picoseconds, compared with a stretched waveform lasting approximately 1.8 nanoseconds. Effective spectral resolution decreased to 1.8 nm because the shorter optical gate required broader bandwidth.
The system maintained several established advantages of time-stretch spectroscopy, including single-photodetector detection, extended recording depth, and continuous measurement of events that do not repeat predictably.
However, the active storage cavity requires careful control of gain, loss, dispersion, and nonlinear effects. These factors currently limit the number of pulse replicas and the available recording window. The demonstrated system supported more than 1,200 duplications, but very unstable or longer-lasting events may remain difficult to measure.
Although the experiments focused on optical pulses, the authors identify potential applications in chemical and materials analysis. Resolving events that do not repeat predictably could complement pump–probe measurements, they write, with applications including “chemical reaction dynamics, laser-induced plasmas and phase transitions in materials.”
The team also proposes combining the method with frequency up-conversion techniques to enable measurements in the mid-infrared – a spectral region important for identifying molecules through their characteristic absorption features. This could extend the approach to “ultrafast molecular fingerprinting,” potentially allowing analytical scientists to follow rapid changes in molecular signatures.
