Limits...
Two-step phase-shifting SPIDER

View Article: PubMed Central - PubMed

ABSTRACT

Comprehensive characterization of ultrafast optical field is critical for ultrashort pulse generation and its application. This paper combines two-step phase-shifting (TSPS) into the spectral phase interferometry for direct electric-field reconstruction (SPIDER) to improve the reconstruction of ultrafast optical-fields. This novel SPIDER can remove experimentally the dc portion occurring in traditional SPIDER method by recording two spectral interferograms with π phase-shifting. As a result, the reconstructed results are much less disturbed by the time delay between the test pulse replicas and the temporal widths of the filter window, thus more reliable. What is more, this SPIDER can work efficiently even the time delay is so small or the measured bandwidth is so narrow that strong overlap happens between the dc and ac portions, which allows it to be able to characterize the test pulses with complicated temporal/spectral structures or narrow bandwidths.

No MeSH data available.


The recorded spectral interferograms I1 and I2 (inset) and the /D(t) − D′(t)/ (black line) and /D(t)/ (blue line) when τ = 0.40 ps (a), and the reconstructed temporal intensity and phase profile by TSPS-SPIDER method (b).
© Copyright Policy - open-access
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC5036035&req=5

f4: The recorded spectral interferograms I1 and I2 (inset) and the /D(t) − D′(t)/ (black line) and /D(t)/ (blue line) when τ = 0.40 ps (a), and the reconstructed temporal intensity and phase profile by TSPS-SPIDER method (b).

Mentions: It is predictable that our TSPS-SPIDER has stronger ability than traditional SPIDER to measure the test pulse with narrower bandwidth. The narrower bandwidth means larger transform-limited pulse duration, so the traditional SPIDER method needs large time delay τ, which may result in the recorded interference fringes to be unresolvable with the available spectrometer, or the significantly increasing effects of uncorrelated noise which diminish as the square root of the number of points per fringe. For example, in Fig. 4(a), we reduce the bandwidth of test pulse down to 7.0 nm by using a bandpass filter, so the corresponding pulse duration shall be not less than 134 fs. When τ = 0.40 ps, there are about 51 sampling points per fringe. Unfortunately, the very strong overlap occurs between the dc and ac portions. In order to avoid the overlap, we estimate that the time delay τ of the test pulse replicas shall be as large as 2.5 ps18. Of course, in this situation, a grating pulse stretcher shall be used instead of the prism pair, and the sampling point will drop to about 8 per fringe. However, by our TSPS-SPIDER, the two ac components centered at τ = ±0.40 ps are well separated temporally, so we can reconstruct the temporal intensity and phase information (see Fig. 4(b)) easily without replacement of the pulse stretcher, meanwhile, the sampling point can maintain about 50 per fringe, which is beneficial to suppress the effect of the uncorrelated noise.


Two-step phase-shifting SPIDER
The recorded spectral interferograms I1 and I2 (inset) and the /D(t) − D′(t)/ (black line) and /D(t)/ (blue line) when τ = 0.40 ps (a), and the reconstructed temporal intensity and phase profile by TSPS-SPIDER method (b).
© Copyright Policy - open-access
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC5036035&req=5

f4: The recorded spectral interferograms I1 and I2 (inset) and the /D(t) − D′(t)/ (black line) and /D(t)/ (blue line) when τ = 0.40 ps (a), and the reconstructed temporal intensity and phase profile by TSPS-SPIDER method (b).
Mentions: It is predictable that our TSPS-SPIDER has stronger ability than traditional SPIDER to measure the test pulse with narrower bandwidth. The narrower bandwidth means larger transform-limited pulse duration, so the traditional SPIDER method needs large time delay τ, which may result in the recorded interference fringes to be unresolvable with the available spectrometer, or the significantly increasing effects of uncorrelated noise which diminish as the square root of the number of points per fringe. For example, in Fig. 4(a), we reduce the bandwidth of test pulse down to 7.0 nm by using a bandpass filter, so the corresponding pulse duration shall be not less than 134 fs. When τ = 0.40 ps, there are about 51 sampling points per fringe. Unfortunately, the very strong overlap occurs between the dc and ac portions. In order to avoid the overlap, we estimate that the time delay τ of the test pulse replicas shall be as large as 2.5 ps18. Of course, in this situation, a grating pulse stretcher shall be used instead of the prism pair, and the sampling point will drop to about 8 per fringe. However, by our TSPS-SPIDER, the two ac components centered at τ = ±0.40 ps are well separated temporally, so we can reconstruct the temporal intensity and phase information (see Fig. 4(b)) easily without replacement of the pulse stretcher, meanwhile, the sampling point can maintain about 50 per fringe, which is beneficial to suppress the effect of the uncorrelated noise.

View Article: PubMed Central - PubMed

ABSTRACT

Comprehensive characterization of ultrafast optical field is critical for ultrashort pulse generation and its application. This paper combines two-step phase-shifting (TSPS) into the spectral phase interferometry for direct electric-field reconstruction (SPIDER) to improve the reconstruction of ultrafast optical-fields. This novel SPIDER can remove experimentally the dc portion occurring in traditional SPIDER method by recording two spectral interferograms with π phase-shifting. As a result, the reconstructed results are much less disturbed by the time delay between the test pulse replicas and the temporal widths of the filter window, thus more reliable. What is more, this SPIDER can work efficiently even the time delay is so small or the measured bandwidth is so narrow that strong overlap happens between the dc and ac portions, which allows it to be able to characterize the test pulses with complicated temporal/spectral structures or narrow bandwidths.

No MeSH data available.