Objective Ultrashort laser pulses have become an important tool for studying the interaction between lasers and matter and have important application values in the fields of biomedicine, high-energy physics research, and communications. The pulse width is an important parameter of the time characteristics of ultrashort laser pulses. For picosecond and femtosecond laser pulses, the pulse width is often measured by an autocorrelator. The time resolution of the autocorrelator must be accurately calibrated before the measurement. For traditional calibration schemes such as the mobile optical path retarder, although their calibration results are accurate, it cannot be calibrated in a single time. On the contrary, the discriminant rate board method can be calibrated in a single time; however, the accuracy of the calibration result is poor, and the accuracy is not high. This study proposes a new method for calibrating the time resolution of the autocorrelator. A flat crystal that can produce a specific time delay is designed, manufactured, and placed in the optical path during calibration. The time resolution can be obtained through a single calibration. Moreover, the measurement results are accurate and reliable.
Methods The designed and manufactured flat crystal with a specific time delay generates double pulse with a fixed time interval T (ignoring high-order reflections), when the pulse to be measured passes through the flat crystal. When the double pulses met in the autocorrelation crystal, the generated autocorrelation signal was a three-peak structure, i.e., a weaker secondary peak signal appeared at equal distances on the left and right sides of the primary peak signal. The time interval between the main peak and the secondary peak signal was denoted as T. The time resolution of the autocorrelator could be obtained by calculating the number of pixels between the primary and secondary peaks. The influence of the thickness, refractive index, and angle of the flat crystal on the calibration result was then analyzed. The time resolution and the relative expanded uncertainty of the autocorrelator were calculated. Finally, the autocorrelator was calibrated using two other calibration methods. The measurement results of the time resolution and the relative expanded uncertainty were given. Moreover, the advantages and the disadvantages of the three schemes were compared.
Results and Discussions After the flat crystal is placed in the optical path, the placement angle, thickness, and refractive index will affect the calibration accuracy of the time resolution. Figure 4 shows the deviation caused by the placement angle of the flat crystal. The deflection angle deviation only slightly affects the time resolution. A resolution error of 2% requires a deflection angle of 16.6°. The measurement error primarily comes from the deviation in the thickness h of the flat crystal and the reading. Figure 5(a) depicts the autocorrelation signal collected on the CCD. Obvious secondary peak signals can be found at both ends of the main peak signal, which is consistent with the theoretical analysis results. In this experiment, the thickness h=1.02 mm and refractive index n=1.450 of the plane flat crystal were maintained. The time interval between the primary and secondary peaks is 9.86 ps. The time resolution of the autocorrelator is 217.88 fs/pixel. The relatively extended uncertainty is 1.50%. Table 3 shows the results of the time resolution and the expanded uncertainty obtained by the three calibration methods. The results of the time resolution calibration using a flat crystal are accurate and reliable.
Conclusions This study proposes a new method for calibrating the time resolution of an ultrashort pulse measuring device based on a flat crystal. By placing a special flat crystal in front of the ultrashort pulse measuring device, double pulse with a time interval T is generated after the pulse to be measured passes through. The generated autocorrelation signal exhibits weaker sub-peaks on both sides of the primary peak. As shown in Fig. 2, the time interval of the sub-peak 2T can be obtained from the refractive index and the thickness of the flat crystal. The pixel value between the sub-peaks can ascertain the time resolution of a single picosecond autocorrelator and calculate the pulse width at the same time. Subsequently, experiments are performed on a femtosecond laser with a pulse width of 180 fs. Figure 5 illustrates the autocorrelation signal collected by the CCD in the experiment, which is consistent with the theoretical prediction. Using this method to calibrate the time resolution of the autocorrelator yields 217.88 fs/pixel. Compared with the calibration result of 214.27 fs/pixel obtained by the moving optical path retarder method, the relative error is only 1.68%. Compared with the discrimination rate board method, the relative expanded uncertainty of the calibration result using this method is 1.50%, which is far better than the discrimination rate board method of 6.96%. A single calibration of the autocorrelator is realized. In conclusion, the calibration result is accurate and reliable.
