Consecuencias del Maltrato Infantil

Cavities

In pump-probe experiments, probing with single pulses rather than a pulse train provides the best time resolution because the moment (or period) of observation is well-defined. In the frame of attosecond physics, the availability of isolated attosecond (as-) pulses at MHz repetition rates would enable new insights in the dynamics of the microcosm and is therefore highly desirable.

XUV as-pulses are generated with every linearly polarized optical half-cycle of the fundamental radiation driving the HHG process. Therefore, to obtain one isolated as-burst per driving fs pulse, either a single as-burst has to be isolated from the harmonic radiation generated by the driving pulse (which is usually a train of as-pulses), or the HHG process has to be confined to a single half-cycle of the driving pulse. The operation of methods following the first approach (e.g. amplitude gating [126]) is inherently limited to few-cycle pulses. Enhancing pulses this short in a passive resonator imposes extreme challenges to the mirror design. While the construction of such a setup would be a sensation, the second

3_{Fig. 8.a in the paper included [37], see Chapter 6, shows the power enhancement vs. detuning from} the stability range center. The width of this curve is given by the coupling among transverse modes, which increases again as the position in the stability range approaches an instability boundary.

5.3 Towards Isolated as-Pulses Using fs-Enhancement Cavities 75

approach seems more realistic for obtaining isolated attosecond pulses from fs-enhancement cavities in the near future. In the following we give a brief overview of the most prominent techniques in line with this approach and discuss aspects relevant for implementing HHG gating in an enhancement cavity.

Polarization gating (PG) uses the fact that the HHG process is strongly polarization- dependent (see [127] and references therein). By combining two delayed counter-rotating circularly polarized pulses (with slightly different wavelengths), the polarization of the driving pulse is modulated in such a way that close-to linear polarization, necessary for HHG, is only achieved over a short time window within the resulting pulse, on the order of a half-cycle. The upper limit of the driving pulse duration is set by the ground state population depletion: if the pulse is too long, then the atoms will be fully ionized by the leading edge before the linear polarization gate starts and the as-burst can be generated.

Two-color gating (TCG) employs waveforms synthesized from a fundamental-radiation pulse and its second harmonic (SH), which can increase the period between the generation of as-bursts to a full optical cycle of the fundamental, see e.g. [128]. Another technique relying on the few-cycle duration of the driving pulses isionization gating (see e.g. [129] and references therein). Here, the first cycle reaching the intensity required for HHG generates an as-pulse and the subsequent cycles fully ionize the gas atoms so that HHG cannot occur anymore. A powerful method working also with multi-cycle driving pulses isdouble optical gating (DOG) [130], which combines PG and TCG. Finally, generalized double optical gating (GDOG) [131, 132] employs two elliptically (instead of circularly) polarized pulses to reduce the ground state population depletion at the leading edge of the pulse. GDOG allows the generation of isolated as-bursts with fundamental radiation pulses as long as 28 fs, generated by a Ti:Sa laser system. This value already lies in the pulse duration range for which enhancement in a passive cavity has been demonstrated [17]. Due to the larger wavelength, for an Yb-based system the same number of optical cycles corresponds to a pulse duration of approximately 36 fs, which can be achieved with nonlinear compression, cf. Section 5.1.2. Therefore, from the point of view of the pulse duration, the prospect of intracavity gating is given.

Another encouraging result is provided by a test we performed in our cavity. To verify the resonance of the cavity for elliptically polarized seeding light, we placed a quarter-wave plate in the input beam, just in front of the cavity input coupler. While we rotated the plate over 360◦, the cavity scan pattern did not change observably, implying that the cavity losses for any ellipticity of the input polarization are comparable. Therefore, we assume that any linear combination of elliptically polarized pulses can be resonant in the cavity. This is a prerequisite for the majority of the gating methods mentioned here. The unique power regime achievable in enhancement cavities, which provides multi-cycle pulses with large intensities at MHz repetition rates, might also open the door to novel combinations of the existing techniques (e.g. a combination of PG with ionization gating seems promising) or to completely new gating mechanisms.

It should also be mentioned that superimposing a SH portion to the fundamental driv- ing the HHG process can be readily implemented in a single-pass fashion. Standard highly reflecting dielectric mirrors usually have a transmission band at the second harmonic fre-

76 5. Outlook

quency of the fundamental light. Thus, the mirror just before the HHG focus could serve as the input coupler for the fundamental radiation, while a SH portion is transmitted through this mirror and overlaps coherently with the intracavity circulating pulse in the HHG focus (but is not resonant in the cavity).

Recently, Durach et al. [133, 134] predicted a plasmonic metallization of thin dielectric films, illuminated by intense single-cycle laser pulses. Such a film could be used as a reflective surface onto which the intracavity generated as-pulse train impinges. If a second pulse, with a slightly different wavelength (e.g. a Ti:Sa-generated) impinges on this surface, the metallization follows the electric field of the pulse virtually instantaneously and thus, could switch the reflectivity of the surface. This phenomenon could in principle be used as an alternative to the above methods for isolating an as-pulse from a train.