An Experimental Protocol for Femtosecond NIR/UV - XUV Pump-Probe Experiments with Free-Electron Lasers

This method can help answer key questions in physics and chemistry such as which bonds are broken first or how atoms and electrons rearrange during a chemical reaction. This technique's main advantage is that the extreme ultraviolet radiation from a Free-Electron Laser or FEL can act as a site-specific probe because it only ionizes specific atoms within a molecule. Learning to achieve spatial and temporal overlap between the FEL and optic laser beams benefits from visual demonstration because very specific diagnostics are used and the effects can be subtle.

Demonstrating this procedure will be Demitrios Rompotis, physicist at the FLASH Free-Electron Laser at DESY. First, verify that the ion detector, the electron detector and the high voltage power for the ion spectrometer electrodes are turned off. Close the FEL and optical laser shutters using the instrument software.

Configure the filters and attenuators installed in the beam line so that the FEL pulse energy and the optical laser power are reduced to less than 1%transmission. Then insert the Cerium YAG beam viewing screen into the interaction region. Open the FEL shutter and examine the screen via a CCD camera.

If the beam spot is not detectable on the screen, slightly increase the beam intensity. Once the beam spot has been located, mark the FEL beam position as a region of interest in the camera data acquisition software. Then open the optical laser shutter and close the FEL shutter.

Adjust the steering mirrors to align the optical laser beam with the marked FEL beam position. Repeat this beam blocking process to refine the spatial overlap and to verify that the overlap is stable. Once the beams are aligned, remove the beam viewing screen.

Turn on the detectors and the spectrometer electrode power. Ensure that a fast photodiode connected to a fast oscilloscope is installed perpendicular to the FEL beam along with a movable mesh to divert a small quantity of scattered photons to the diode. Reduce the FEL pulse energy and the optical laser power to 1%transmission.

Then close the FEL and optical laser shutters. Insert the scattering mesh into the beam. Adjust the mesh position, the FEL pulse energy, and the optical laser power so that each individual beam produces a clear signal and both signals have the same height.

Then close the optical laser shutter. Configure the fast oscilloscope to use the finest time base available and to collect about 100 averages for a trace. Record and save a reference trace of the FEL beam alone.

Then close the FEL shutter and open the optical laser shutter. Compare the trace from the optical laser with the FEL reference trace. Then shift the optical laser pulse arrival time so that the onset of the optical laser signal precisely matches the onset of the FEL signal.

Repeat the beam blocking and signal onset comparison to confirm that the FEL and optical laser pulses are precisely aligned. Note the time when the FEL and optical laser pulses overlap as the initial estimate of T0.To begin fine tuning T0, attenuate the FEL and the optical laser to a sufficient degree to avoid damaging the ion and electron detectors when xenon gas is introduced into the system. Ensure that the spectrometer is in time of flight mode.

Then introduce xenon gas into the chamber either through the gas jet or by allowing xenon gas into the evacuated chamber through a needle valve. If the latter method is used, attain a chamber pressure between one times 10 to the negative seven and one times 10 to the negative six millibars. Record a xenon ion time of flight spectrum.

Then close the FEL shutter and adjust the FEL pulse energy so that xenon two plus and xenon three plus are among the strongest xenon charged states in the time of flight spectrum and higher xenon charged states are suppressed as much as possible. Then close the FEL shutter and open the optical laser shutter. Adjust the optical laser power so that the laser pulses produce primarily xenon plus with only a small amount of xenon two plus.

Open the FEL shutter when the adjustment is finished. Based on the previously determined rough T0 value, set the FEL and optical laser pulse timing to have the optical laser pulses arrive about 200 picoseconds before the FEL pulses. Acquire a xenon ion time of flight spectrum and determine the ratio of xenon two plus to xenon three plus from the peak areas.

Then configure the lasers such that the optical laser pulses arrive about 200 picoseconds after the FEL pulses. Acquire another time of flight spectrum and determine the ratio of xenon two plus to xenon three plus. Verify that the xenon three plus signal is significantly stronger in this spectrum than in the previous spectrum.

Sometimes the difference between laser early and laser late in the xenon signal is very small because of insufficient spatial overlapping. In such a case, one should repeat the procedure of spatial overlapping in order to achieve a big difference in the two signals. Set the laser timing to halfway between the previous two values and acquire another time of flight spectrum.

Compare the ratio of xenon two plus to xenon three plus to determine whether the optical laser pulses are arriving before or after the FEL pulses. If the optical laser pulses are arriving before the FEL pulses, set the timing to halfway between the current value and the value at which the optical laser pulses arrived 200 picoseconds after the FEL pulses. Acquire another time of flight spectrum and examine the ratio of xenon two plus to xenon three plus.

Continue adjusting the laser pulse timing until T0 has been approximated with a precision of better than 500 femtoseconds. Then set up a delay scan over a region of plus or minus one picosecond around the approximate position of T0 in steps of no more than 50 femtoseconds. Acquire a time of flight spectrum and determine the ratio of xenon two plus to xenon three plus for each step.

Plot these ratios with respect to the delay times, derive a step function and calculate the center of the step function to obtain the exact temporal position of T0.Xenon ion time of flight spectroscopy could be used to determine whether an 800 nanometer near-IR pulse arrived in a xenon gas target before or after an FEL pulse with a photon energy of at least 67.5 electron volts. Post-ionization of excited metastable xenon two plus occurred when the near-IR pulse arrived after the FEL pulse increasing the xenon three plus yield. Plotting the ratio of xenon two plus to xenon three plus as a function of delay time provided a step function from which T0 could be determined.

Iodine ion momentum images were also used to determine T0 with a photon energy of at least 57 electron volts. A low energy contribution was visible as a spike only when the UV pulse arrived before the FEL pulse. T0 was extracted from a plot of the spike ion yield as a function of delay time.

Shot-by-shot data recorded by a Bunch Arrival Time Monitor was used to correct for the jitter in the relative arrival time of the FEL pulses with respect to the optical laser pulses. This produced a noticeable improvement in the data quality particularly in the temporal resolution. Once mastered, establishing the temporal and spatial overlap between the optical laser pulses and the FEL can be done in about two to three hours while the pump probe measurement that follows typically takes several days.

Although this procedure was developed for atoms and molecules in the gas phase, it can also be applied to other samples such as nanoparticles or liquids and solids. Don't forget that working with high-power femtosecond lasers can be extremely hazardous. Specific safety training is mandatory.

And when working with high-power lasers, always wear your protective laser safety goggles.