Attosecond pulse source development and characterisation

18/08/05 – C A Haworth

We are aiming to create and characterise a source of single attosecond pulses at Imperial College. To achieve this we need to carry out high harmonic generation with a few-cycle driving laser.

Attosecond pulses are created from phase locking large bandwidths of XUV harmonics and spectrally filtering out parts of the harmonic spectrum close to the cut-off. The high harmonics are produced by focussing the IR laser into a gas target at intensities greater then 1 10¹⁴ Wcm^-2. This produces the large bandwidths necessary in the XUV. Attosecond pulses can only be created at short wavelengths because the pulse durations start to reach the single cycle limit in the IR. For example at 800nm a single optical cycle of the e-field lasts 2.6fs and we need to generate pulses shorter than this. High harmonics are also useful as they are intrinsically in phase and coherent as a result of the generating process.

Single isolated pulses were first reported in 2002 by Ferenc Krausz’s group in Vienna where they used a CEP stabilised IR drive laser to produce 650as pulses[1,2].

Currently we are in the process of developing a number of gas targets from kHz piezo driven valves to differentially pumped static cells and needles. We need to find a stable source capable of providing enough gas density to generate the very highest harmonics efficiently whilst not adversely affecting the vacuum conditions required.

Another challenge in the field of attoscience is detection and metrology of the attosecond pulses. At Imperial College we intend to primarily use an atomic streak camera method[3] to measure the attosecond pulses. After generation we will have an extensive XUV beam-line to manipulate the harmonics. Allowing us to spectrally filter, focus, measure and deliver the beam to an experiment.

Once the harmonics are produced they will enter a filter chamber where we can apply a number of different material filters, this allows spectral and special filtering and can also provide chirp compensation[4]. After filtering we can either deflect the beam into a spectrometer or it can be focussed onto a target using a gold coated toroidal mirror. With this we should be able to achieve focussed intensities of 1 10¹⁴ Wcm^-2 with the harmonics.

Fig. 2:Image of harmonics on MCP

Detection of the harmonics is currently carried out using a flat-field XUV spectrometer setup to measure from 40nm-10nm. The light is dispersed off an variable line spacing Hitachi grating, and detected using a micro-channel plate. At the back of the MCP we use a cooled 12bit CCD camera (Photonic Science, Coolview FDI), matched to the phosphor wavelength of the MCP to image the harmonics.

[1] Steering attosecond electron wave packets with light Kienberger, R., et al. Science , 297 (5584), 1144-1148, (2002)

[2] Attosecond metrology Hentschel, M., et al. Nature , 414 (6863), 509-513, (2001)

[3] Atomic transient recorder Kienberger, R., et al. Nature , 427 (6977), 817-821, (2004)

[4] Amplitude and phase control of attosecond light pulses Lopez-Martens, R., et al. PRL , 94 (3), (2005)