/de/research/projects/2.3/topics/XUV photoelectron spectroscopy/photoelectron imaging/index.html
2.3 Time-resolved XUV-science
Project coordinator(s): A. Rouzée, S. Patchkovskii

" XUV/X-ray photoelectron imaging "

A. Rouzee

Main Goal

This research aims to follow all aspects of a chemical reaction, i.e. the time-dependent changes in (i) the atomic positions and (ii) the electronic structure of an evolving target molecule. We are developing novel experimental approaches towards this goal that make use of the wave character of electrons generated by means of XUV or X-ray ionization. In theses experiments, information is stored in the photoelectron angular distribution that can be measured using the well known velocity map imaging technique.

What is photoelectron imaging?

When a molecules is exposed to a XUV or X-ray laser pulse, an electron can be release by single photon absorption. The kinetic energy of the outgoing electron (Ek) is given by the photon energy (Ep) and the ionization potential (IP) of the target molecule by the relation:


At high photon energy (typically above 150 eV), the photoelectron wavepacket leaves the molecule with a rather short de Broglie wavelength (1 Angstrom or below) and becomes sensitive to the surrounding atoms. As an example, photoelectrons can be scattered off one of the neighbouring atoms in the molecule on their way out and leave an imprint of this scattering event in the photoelectron angular distribution (PAD) that can be used to retrieve information on the molecular structure similar to a photon diffraction experiment. Electron-molecule scattering was theoretically studied by Dill and Dehmer and a number of successful experiments have already been performed at synchrotrons for stable molecules relying on this concept. Measurements of laser-generated high kinetic energy photoelectrons offer the advantage that they can be generated now with ultrashort XUV/X-ray laser pulses, enabling time-resolved molecular studies with sub-10 fs resolution. The photoelectron imaging technique is based on this concept: we measure the electron generated "within the molecule" by an ultrashort XUV/X-ray pulse in order to record a movie of a molecule undergoing structural changes.

Experimental setup

bottom-up Science

The experiments are performed in a lab-based environment, making use of a home made, carrier-phase stabilized, TW laser system or at free electron laser facilities, such as FLASH, the free electron laser of Hamburg, and LCLS, the linac coherent light source at Stanford. Our laser system is a commercial CEP-stabilized amplifier systems from KMlabs delivering 2.5 mJ at 3 khz repetition rate, with 30 fs pulse duration. Part of this beam is further amplified in a home-made amplifier build in collaboration with amplitude technology. After second amplification stage, the laser typically delivered 35 mJ, for a 35 fs pulse duration at 50 Hz repetition rate.

bottom-up Science

In order to probe ultrafast molecular dynamics by means of high kinetic energy photoelectron, a soft X-ray beamline has beam constructed at the Max Born Institute. This beamline is composed of four high vacuum chambers:

  • A high harmonic source where a portion of the beam is used to generate high harmonic following the strong field ionization of an atomic gas target.
  • An in vacuum interferometer used to recombine the XUV pulse with a VUV/visible or near IR pulse needed to trigger the photochemical reaction of interest.
  • A velocity map imaging spectrometer with a gas injection system where we record the photoelectron angular distribution after XUV/soft X-ray ionization of the target molecule
  • A soft X-ray spectrometer that allows to monitor the harmonic yield during the experiment.

Velocity map imaging

bottom-up Science

The basic idea of a velocity map imaging spectrometer is to accelerated the charged fragments resulting from ionization and fragmentation towards a 2D position-sensitive detector consisting of a microchannel plate(MCP)/phosphor screen assembly and a camera system. Charge and mass selection is made possible through the application of a short gate to the MCP detector. In the so-called "velocity map imaging" mode, the detector is operated under a condition where the position of the particle impact on the detector only depends on its initial transversal velocity. Accumulating an image that contains a large number (typ. 106) of these impacts then allows to retrieve the 3D velocity and angular distribution of the selected charged fragments using standard mathematical tools based an Abel inversion [1].

Laser-induced molecular alignment

For all diffraction experiments, it is an absolute requirement that the measurements are done in the molecular frame, meaning a coordinate system that is fixed with respect to the molecular axes. In experiments performed in a randomly oriented sample, most of the structural information that is contained in the photoelectron angular distributions is averaged out. We solve this problem by using the laser-induced molecular alignment technique. Laser-induced alignment techniques operate on the basis of dynamically driving the molecular rotation using a strong laser pulse, by exploiting the interaction of the laser field with the molecular polarizability. When the pulse duration is shorter than the rotational period of the molecule, a rotational wave packet is created through a cascade of two photon Raman processes. Due to the quantization of the rotational energy level, this rotational wave packet rephases periodically after the laser pulse has ended leading to a transient alignment, where molecules switches between an alignment along the laser polarization and an anti-alignment perpendicular to the laser polarization. We have been pioneering experiment on laser-induced field free molecular alignment [2,3], and we have recently exported this technique at free electron laser facilities [4].

Recent results

bottom-up Science

Figure: Difference of 2D projections of the 3D electron momemtum distribution recorded with a velocity map imaging spectrometer between aligned and antialigned CO2, N2, O2 and CO molecules photoionized with an high harmonic comb. Two configuration of polarization are shown between the alignment laser pulse and the XUV pulse.

A series of small molecules (CO, N2, CO2, and O2) were exposed to a sequence of an IR laser pulse that dynamically aligned the molecules, and an XUV pulse generated by HHG that ionized the molecules at a variable time delay. Photoelectron angular distributions were recorded using a VMI spectrometer at times corresponding to maximum alignment or anti-alignment. This experiment has shown that ionization of aligned molecules leads to a large modulation of the molecular frame photoelectron angular distribution that depends of the initial molecular orbital and on the wavelength. This result reflects the sensitivity of the PAD with respect to the orbital geometry (electronic structure) and molecular structure via the electron-molecule interaction after ionization (diffraction of electronic wave on surrounding atoms).

Next steps

In order to get structural information, the electron ejected during photoionization need to be localized around one of the atomic core, which rationalize the use of soft and hard X-ray laser pulse. We are currently working on the generation of high harmonic up to the water window using a combined IR and mid-IR driving field.

In parallel to the soft X-ray beamline, we are working on a pump-pump-probe arrangement system, with a first pulse that dynamically align the molecule, a second pulse that trigger a chemical reaction which is then investigated at a later time delay using few 100 eV, few fs laser pulse. It should allow to monitor time-dependent changes in molecular structures with ultra-high spatial and temporal precision making use of the high energetic outgoing electrons.






  • Part of the research is funded by an ERA-chemistry proposal, where a collaboration has started between the MBI and Szeged University (Hungary) aiming to develop an HHG source using two-color 800 nm and 5-10 um radiation in order to generate wavelength up to the water window (2.3-4.4 nm, i.e. 280-540 eV).

  • Our team has a strong collaboration with the Junior theory group led by Olga Smirnova.
  • Experiments at FEL are performed in collaboration with many institutes and universities, among them the MPQ-ASG, the CFEL, PULSE (Stanford), DESY, Aarhus university, Tohoku university, A&M University, LASIM (Lyon university), Helmholtz-Zentrum Berlin, and Western Michingan University.


[1] M. Vrakking,Phys. Rev. Lett.,107,3605, (2011).
[2] F. Rosca-Pruna, and M. J. J. Vrakking,Phys.Rev. Lett. 8, 153902 (2001).
[3] O. Ghafur, A. Rouzée, et al., Nature Physics 9, 289 (2009).
[4] A. Rouzée, et al., New Journal of Physics 11, 105040 (2009).
[5] A. Rouzée, et al., in PUILS, Springer series in Chem. Phys. 99, 45 (2010).
[6] P. Johnsson, et al., J. Phys. B 42, 45 (2009).
[7] N. Berrah, et al., J. Mod. Opt. 1362, 45 (2010).
[8] F. Kelkensberg, et al., Phys. Rev. A 84, 051404 (2011).
[9] A. Rouzée, et al., J. Phys. B 45, 074016 (2011).