In a recent paper published in Physical Review Letters, we propose a new approach to observe the correlated motion of two electrons on the attosecond timescale. In this pump-probe setup, two identical extreme-ultraviolet light pulses with a duration of just a few hundred attoseconds (1 as = 10-18 s) are sent onto a helium atom. We expect that with the continuing development of intense attosecond pulses, this kind of experiment could be performed in the next few years. Most current experiments use a strong few-femtosecond infrared field in combination with an extreme-ultraviolet attosecond pulse. These rely on highly nonlinear effects to attain subcycle time resolution within the infrared pulse. In contrast to such setups, the wave packet dynamics are not modified by the fields when using two extreme ultraviolet pulses. The proposed measurement would thus be one of the first experiments to directly observe field-free correlated electron dynamics in atoms on their natural attosecond timescale.
In our proposed setup, the first (pump) pulse excites a coherent wave packet of doubly excited states. These are prototypical examples of highly correlated states where the two electrons influence each other strongly. After letting this wave packet evolve for some time, the second (probe) pulse ejects both electrons. By repeating the sequence many times with different time delays between the two pulses, a "movie" of the doubly excited wave packet can be created frame by frame.
In the paper, we show that by measuring only one of the two ejected electrons and counting only those electrons within a specific energy interval, it is possible to gain direct access to an observable related to the dynamics of both electrons: the distance between them at the moment of ionization.
There is one further problem to overcome: Both steps in the pump-probe sequence only occur with small probabilities. Both in double excitation (pump) and in double ionization (probe), absorption of one photon has to lead to a two-electron transition. As a photon only "talks" to one electron directly, these two-electron transitions are quite unlikely. We show that one can exploit quantum interference to increase the magnitude of the signal: Since the pump and probe pulses are identical, absorption of two photons from just one of them leads to the same final states as absorption of one photon from each pulse. This "direct" pathway, where each photon ejects one electron, is orders of magnitude more likely than the more interesting pump-probe pathway. However, it does not represent an incoherent background that masks the signal of interest. Instead, it provides a coherent reference pathway that the pump-probe pathway through the doubly excited states interferes with. The amplitude of the interference term is about a hundred times larger than the magnitude of the pump-probe signal by itself, thus providing an experimentally more accessible signal.
Reference: J. Feist, S. Nagele, C. Ticknor, B. I. Schneider, L. A. Collins, and J. Burgdörfer, Phys. Rev. Lett. 107, 093005 (2011)