Recent efforts to achieve time-resolved imaging of chemical reactions or other structural changes with sub-femtosecond resolution
possess a strong interdisciplinary character, since the goal is of interest for physics, chemistry, and biology. Pioneering experiments with
near-infrared ultrashort laser pulses like the ones in [1,2] in which
structural information about the chemically most relevant valence
electrons was extracted from the emitted high-harmonic radiation
or electron spectra have been of great interest, since these approaches should intrinsically have the potential for providing also
the required time resolution. However, in contrast to the initial
assumptions, recent experiments have demonstrated that the
molecular strong-field response depends, at least for some molecules, on more than one orbital [3–6]. While such multi-orbital effects clearly complicate simple imaging schemes, they can also be
the source for even richer information that can be gained from such
experiments. An example is the electron–hole dynamics in the lasergenerated ion that may be observed by analyzing the high-harmonic
radiation [5]. Though very exciting by itself, this appears to make direct imaging of the valence electrons and their field-free dynamics
during, e.g., a chemical reaction more complicated.
The proposed imaging schemes based on linear-polarized ultrashort near-infrared laser pulses may roughly be divided into two
categories, rescattering-based schemes and direct imaging. The
first category is based on the celebrated three-step model of
strong-field physics in which (1) an electronic wavepacket leaves
the molecule around the local maxima of the electric field by tunneling ionization, (2) this wavepacket is accelerated in the laser
field and reverses its direction as the field direction changes, and
(3) the electronic wavepacket may recollide with its parent ion.
As a consequence of this recollision, the electronic wavepacket
may partly scatter elastically (diffraction) or inelastically (leading
to excitation or further fragmentation), or recombine by the emission of high-harmonic radiation. Clearly, all these processes should
depend on the structure of the molecular ion and thus have the potential for revealing structural information. This includes both
electronic structure as well as nuclear geometry. Corresponding reviews may be found in [7,8]