Thierry Ruchon's Research pages

At the base of attosecond pulse generation is the process of high harmonic generation (HHG). This phenomenon, discovered in the late 80's simultaneously in Saclay and Chicago, is a highly non linear upconversion process by which an initial visible-IR photon is converted to the XUV spectral domain (120 nm wavelength and below). It is observed when focusing a short laser pulse in a gaseous target, usually at a few to some tens of mbar pressure. As any nonlinear process, it results from the combination of a microscopic response of the medium to the excitation and a macroscopic one through phase matching. When HHG is performed with multicycle laser pulses, as illustrated above, the XUV spectrum appears as a series of odd harmonics of the fundamental laser beam. This is a consequence of the repetition of the process every half cycle of the driving laser. For more details, a lecture on HHG is available in Tutorial/lecture page. A very complete theoretical lecture was made available by Armin Scrinzi. A very brief introduction is also available in the insets below the figure.

Typical harmonic spectrum obtained in neon gas on the LUCA laser in Saclay (SLIC laser servers) . The vertical dimension represents the divergence of the harmonics while the horizontal axis is the spectral dimension. This spectrum was obtained with a 800 nm wavelength laser, with 50 fs duration.

Microscopic response

The process of HHG is best described by quantum mechanical methods. A long term collaboration with Richard Taïeb, Jérémie Caillat and Alfred Maquet from LCPMR provide us the theoretical support on the resolution of the Time-Dependant Shrödinger Equation (TDSE). Locally, Thierry Auguste is implementing the Strong Field Approximation (SFA) version of TDSE, but includes full 3D propagation of the incoming field and XUV field in his computations.

In a pictorial model, the process may be split in three steps. The first one, occurring close to an extremum of the field, is the tunnel ionization of the outermost electron of the target gas. In a second step, this electron is driven by the laser field during about half of the period of the laser field, which is 1.3 fs with our most used Ti:Sapphire 800 nm wavelength lasers. This excursion in the continuum may drive the electron back to the ionic core where the collision (3rd step) may result in the emission of its kinetic energy as an XUV photon.

Illustration of the three step model. The attosecond structure of the XUV radiation comes from the very narrow time window, close to the field extremum, during which tunnel ionization is significant.

Macroscopic response

Like all nonlinear conversion phenomena, phase matching plays a crucial role in HHG. Here we upconvert IR photons to XUV in a gas cell with a laser beam intense enough to partially ionize the medium. Since the ions, the electrons and neutral do not have the same complex index, phase matching will depend on both time and space, as the medium gets progressively ionized at one point in the gas target. This makes the result of HHG very much laser dependent, even on a daily basis. We studied temporal aspects of this phenomenon (see T. Ruchon et al., NJP 2008 and X. He et al., Phys. Rev. A 2009). As a benefit, it also yields the good directionality of the XUV beam, which is emitted perpendicular to the IR wavefront, it may be used to cut on or off the emission during a pulse (ionization gating) and may be used to even split the emission from successive half cycles in the attosecond light house configuration (see e.g. Kim et al.)).

Illustration of the phase matching for HHG. The progressive ionization of the medium while the laser goes through leads to a varying index in time, both for its absorption and refractive parts (right graph). The same happens in space.