Two-color X-ray operation
A magnetic chicane is located in the SASE3 undulator section, with 11 undulators before and 10 undulators after the chicane. This setup allows imposing a variable delay on the electron bunches (dark blue) and to set up two independent X-ray photon energies in the two sections, such that the second color (red) is delayed with respect to the first color (yellow). The photon energies can be freely chosen (within the boundary conditions of the given electron beam energy).
Only half of the undulators are available for each of the colors in this case. In addition, the electron beam quality is partially diminished in the first half of the undulators, making lasing more difficult for the second color. Combined, both effects result in a significant loss of pulse energy compared to standard single-color operation. Maximum achieved performances so far (until September 2022):
- 2200 eV: 150 µJ per color
- 1600 eV: 200 µJ per color
- 1000 eV: 700 µJ per color
- 600 eV: 900 µJ per color
If the pulse energy of the pump pulse (first color) is significantly lower than the probe pulse (second color), the probe pulse intensity can be much higher. For example, for a pump of 1 µJ at 930 eV, a probe of 3 mJ at 945 eV has been demonstrated. The average spectral distribution of the two colors can be characterized simultaneously to the experiment with an appropriate setup of the photoelectron spectrometer (PES) located behind the last undulator. The shot-to-shot pulse energy variations can also be measured using the PES, but an absolute pulse energy calibration requires dedicated measurements. The X-ray gas monitor detectors (XGMDs) which are typically used in standard operation are not able to distinguish between different photon energies and thus measure the integral of both pulses.
The maximum possible delay between the colors depends on the electron beam energy. At the typical setpoints of 11.5 GeV, 14.0 GeV, and 16.0 GeV, these are ~900 fs, 600 fs, and 400 fs, respectively (solid blue line in the left figure). At the moment, significant negative delays (second color arrives before the first color) are not yet available, but the installation of an additional optical delay line (ODL) will happen in the near future, which, if used, will impose a fixed delay of ~150 fs to the first color and thus allow for delay scans across time overlap, but consequently reduces the maximum delay by the same amount if inserted (dashed orange line in the left figure). Very small negative delays (~10 fs) can also be achieved through the so-called fresh-slice technique, which is only available on a best-effort basis within an R&D collaboration and at the cost of a reduced number of pulses per train (no interleaved operation with SASE1 possible).
The two colors are generated by separate undulator sections, which also means that their source positions are substantially different. In the SQS instrument, this results in the fact that the two colors are not focused at the same distance from the horizontal and vertical KB focusing optics (HKB and VKB). This is illustrated in the right figure. The experiment can thus either be carried out in an intermediate plane (I, black), where the foci of both colors have similar size but are larger than in standard operation, or a deliberate choice can be made to have one focus larger than the other one (I1 or I2 planes, blue or red). The spatial overlap of both colors at the focal plane is very stable based on current experience and there is essentially no arrival time jitter between the pulses.
Instead of using the two undulator sections for the two colors, alternatively, two different photon energies can also be generated with alternating undulators. In this case, the source positions of both colors are identical, but no delay can be introduced between them, i.e., they always arrive at the same time at the experiment.
The two-color mode can be combined with one of the short pulse modes (see below), which are only available on a best-effort basis within an R&D collaboration and at the cost of a reduced number of pulses per train (no interleaved operation with SASE1 possible).
Some unwanted side effects have been observed in two-color operation the past, which are to a large extend solved, but still require special attention:
-
The intensity of the probe pulse can decrease for larger delays. In addition to the hard limit on the maximum possible delay (see above), this can constrain the delay to smaller numbers, depending on the experimental requirements.
-
Blocking the individual colors for "pump-only" and "probe-only" measurements is often required by the experiments. Removing the second color can easily be done by opening the second half of the undulators, however, special tricks need to be applied to remove the first color without influencing the pulse parameters of the second color.
-
Close to time overlap, "cross-talk" between the pulses has been observed, meaning that the pulse energy of the probe pulse becomes significantly more intense. This effect is particularly pronounced if the two colors are very close in photon energy or are harmoics of each other. It can be circumvented with the fresh-slice technique, where different parts of the electron bunch lase for the different colors.
References for two-color mode:
S. Serkez et al., Appl. Sci. 10, 2728 (2020)
G. Geloni et al., arXiv:1706.00423 (2017)