ESTADIOS DE JEAN PIAGET.

1.5

Above threshold ionization (ATI)

Not all electrons driven back to the atomic core by a strong laser field recombine to release XUV photons. Actually the cross section of the recombination process is rather small as dictated by the conversion of the fundamental to XUV photons. Those electrons that do not re-combine, are re-scattered by the core, leave the atom vicinity with kinetic energy corresponding to a photon number much higher than necessary to reach the ionization threshold. The process is manifested in the spectral domain by a series of photoelectron peaks separated by the energy of the fundamental photon. The plateau behavior discussed for harmonics so far is also present for ATI electrons, implying the existence of the same mechanism behind the two phenomena. Today the effect is understood within the frame-

work of the three step model as well [39]. The fact that the same mechanism governs XUV

light generation and ATI, implies that at the few-cycle limit a series of interesting phenom- ena are to appear when the interaction of atoms with short pulses is taking place. Recent experiments with the involvement of phase stabilized pulses, have unveiled a plethora of interesting features with a huge range of applications for the determination of the absolute

phase of short pulses [65], [66], or even the demonstration of the double slit experiment for

Chapter 2

Schemes for temporal measurements

on an attosecond time scale

The theoretical approaches concerning the generation of attosecond pulses in the XUV and X-ray regime, as well as for steering the generation process by means of waveform control of few-cycle light pulses, have been introduced in the preceding chapter. However, generation of attosecond pulses shall be always combined with a competent measuring technique for their qualitative and quantitative characterization. Characterization of attosecond pulse trains, has been achieved up today by means of the techniques mentioned in chapter 1. For the experiments discussed in chapter 5, aiming the characterization of isolated attosecond pulses and in chapter 6 for direct tracking of the evolution of light waveforms, the technique of the atomic streak camera is employed. This chapter is devoted to the introduction of this concept, and how its employment could allow gaining temporal access to phenomena evolving on an attosecond time scale.

2.1

Principles of a streak camera

Before presenting the concept of a streak camera operating on an attosecond time scale, a short introduction to the conventional streak camera is essential. The streak camera was

invented by [68], [69] and in principle, incorporates ideas overtaken from an oscilloscope

as well as from a conventional film camera. While the conceptual impact of the former to the design of a streak camera is more apparent, it shares several common features with the latter as well. Here instead of using a sequence of film frames to record the time evolution of an event, the image is streaked and projected in a single slide. The streak camera principle is illustrated in figure 2.1. An optical pulse or waveform to be characterized, impinges on a photocathode and generates a backside photoemission of an electron bunch, that in principle imitates the temporal structure of the laser pulse. In the illustrated case for example, a pulse consisting of a double feature generates two electron bunches of similar temporal shape. The generated electrons enter a capacitor, where a fast temporally rising voltage applied to its plates forms a fast ramping electric field between them. If the

2.1 Principles of a streak camera Cathode Laser light Electron pulse Phosphor Screen

E

Fast rising electric field

-

Figure 2.1: Schematic of a conventional streak camera. A laser pulse with arbitrary shape (a double optical burst here for illustrating the concept) impinges on a thin cathode and results in a back side photoemission. A fast ramping voltage, well timed with the laser pulse is applied to the plates of a capacitor. Electrons localized at the leading or at the trailing edge of the bunch are deflected asymmetrically while passing through the plates. A detector with spatial sensitivity allows resolving this asymmetry which is directly associated with the pulse duration.

field ramp is temporally faster or at least comparable with the electron bunch duration, the overall lateral displacement of an electron at the leading edge of the bunch will be significantly different than that of one at the trailing edge. By projecting the resulting bunch to a phosphor screen, a transversally elongated spot resulting from the effect of the field to the bunch will be formed. Under identical voltage conditions, a short pulse will result in a less elongated trace, as compared to that of a long one.

The deduction of the temporal information about the bunch from this trace, requires the precise knowledge of the temporal evolution of the field between the plates, as well as the degree of synchronization between the electron bunch and this field. Alternatively, a temporally characterized bunch could allow the retrieval of the fast electric field ramp.

At a very first glance, the concept does not seem to face any severe limitation as re- lated to the shorter pulse that could be characterized. The experimental implementation of these ideas however, confronts multiple bottlenecks that so far have resulted to a maximum achieved temporal resolution for streak cameras, of the order of a few hundreds of femtosec- onds. A key limitation arises from the lack of perfect synchronization between the ramping electric streaking field and the generated electron replica of the optical pulse, as well as in-

2.2 Streak camera on an attosecond time scale