摘要： In recent years attosecond science motivated the development of laser systems, which provide millijoule energy level few-cycle pulses. These experiments rely on a field with stable amplitude and carrier envelope phase (CEP) to provide reproducible results [1]. As a consequence, diagnostic devices, which are capable of single-shot CEP measurement and pulse tagging, became a must-have equipment in today’s laser systems. The self-referenced interferometry such as f-to-2f [2] are one of the most typical way to characterize the shift of the CEP in a single-shot manner. The highest achievable recording rate is mostly limited by the measurement times of the optical spectroscopes, which can only reach 10 kHz even with fast detector array. Dispersive Fourier Transform (DFT) [3,4] can be used to bypass this limitation, so single-shot recordings of the output signal of an f-to-2f interferometer becomes possible even at high repetition rates. The method depends on an optical element with enough dispersion to stretch to pulse duration up to the nanosecond range. Therefore, the spectral modulation pattern containing the encoded CEP appears in the temporal domain, which allows for tracking with a relatively slow photodetector. To demonstrate the validity and performance of this concept, a CEP drift measurement was performed on the state-of-the-art mid-infrared (MIR) laser system at ELI-ALPS [5], which provides mid-IR laser pulses at 100 kHz. As a comparison, an alternative measurement was performed in parallel with a grating spectrometer (Fringeezz, Fastlite) [6] at 10 kHz sampling rate. The two recorded signals were synchronized and a decimated 10 kHz subset was extracted, where correlation between measurements is the highest. One of the main limitation of the DFT method originates from the time jitter between the output of the laser and the TTL signal used for triggering. This time jitter changes the time delay of the modulation pattern, creating an additional noise source for the measured relative CEP value. In order to perform truly time jitter-free single-shot CEP drift measurement, an additional CEP independent spectral modulation was introduced, which can be used to determine the noise originating from the jitter. The jitter-free decimated dataset against the grating spectrometer measurement is displayed on Fig. 1. The calculated CEP noises are summarized in Table 1. These values agree within the limit of the uncertainty of these measurements, validating that this new CEP drift measurement method is easily scalable to arbitrary repetition rates.