Electronic control of the multiscale time delay between two femtosecond amplifiers seeded by free-running oscillators

Abstract

Many systems, e.g. complex biomolecules, exhibit a multiscale distribution of time constants, from hundreds of femtoseconds up to milliseconds. As previously demonstrated [1], such a huge time range can be covered using pump-probe spectroscopy when pump and probe pulses are delivered using two different femtosecond amplifiers, electronically triggered at desired time delays. However, this method requires the two amplifiers to be seeded by two synchronized oscillators of identical repetition rates, so that fine-tuning of the time delay can be achieved by changing the relative phase between the two oscillators [1]. Recently, we have demonstrated that Arbitrary-Detuning Asynchronous Optical Sampling (AD-ASOPS) [2] could be readily applied to femtosecond amplifiers [3]. As shown in Fig. 1(a), this method consists of determining the actual time delay between each oscillator pulse pair using an opto-electronic system (AD-ASOPS device in Fig. 1(a)) counting the exact number of pulses from each oscillator between coincidences [2]. The knowledge of the pulse numbers actually amplified at a 1-kHz repetition rate enables the a posteriori determination of the time delay between amplified pulses with sub-picosecond accuracy [3]. The main advantage is that application to pre-existing commercial femtosecond amplifiers is now straightforward, as it requires no change on the oscillators. The drawback is that, for a specific target delay, time delays actually achieved span an interval corresponding to the oscillator period (here 12.5 ns). While delays can be determined a posteriori with sub-picosecond accuracy, the resulting delay distribution is uniform whereas a logarithmic distribution (including below 12.5 ns) would be more desirable.