Generation of gigawatt-scale isolated attosecond pulses

Summary form only given. To make a breakthrough for attosecond science, one of the most important issues is the development of high-power isolated attosecond pulses (IAPs) and/or attosecond pulse trains (APTs). At present high-power APTs have been successfully generated thanks to research on harmonic energy scaling using a loosely focusing geometry [1]. Although IAP were already obtained using a state-of-the-art Ti:sapphire laser system, output energy is still not sufficient [2] for inducing nonlinear phenomena. Since a phase-matching technique can´t be utilized due to high ionization rate of target gas, a efficiency is just 10-⁵ level [2] even if Xe gas was used.In this paper, we demonstrate the first generation and full characterization of microjoule-energy isolated attosecond pulses[3], which have sufficient power for inducing nonlinear phenomena in atoms and molecules as well as for attosecond-pump/attosecond-probe spectroscopy. Our generation scheme is based on the two-color (TC) infrared laser field synthesis [4] and the energy-scaling method [1] of HHG. As we previously reported [4], our optimized TC scheme for generating IAPs enables us not only to relax the requirements for the pump pulse duration (30fs) but also to reduce ionization of the harmonic medium. Because neutral media allow us to use the phase-matching technique, we can realize the generation of intense IAPs with high conversing efficiency. To increase both the interaction length and the pump energy, we employed a loosely focused pumping geometry (f = 4000-mm) according to the energy scaling method of HH. To correct the different focusing points between main pump pulses (800 nm, 30 fs) and supplementary infrared pulses (1300 nm, 40 fs), both pulse were focused by two separate focusing lenses. Target Xe gas was statically filled in the interaction cell with 12-cm length. Under optimized phase matching condition, the total pulse energy of continuum harmonic spectrum (28 - 35 eV) is eva- uated to be 1.3 μJ/pulse, which is almost 100-fold higher than ones ever reported. The conversion efficiencies attained was 1.1×10-4 at the cutoff region. Moreover, the output energy of mid-plateau region (14 - 29.5 eV) having a quasi-continuum spectrum attained to 10 μJ/pulse. By autocorrelation method using a Coulomb explosion of N2, the pulse duration of HH was measured directly. Figure 1 shows the measured AC traces (dashed profile) for the cutoff HH selected by Sc/Si mirror (28 - 35 eV). The vertical axis is the detected signals of N+ via the Coulomb explosion of N2. The measured pulse duration was evaluated to be 500 as. The peak power of our table-top light source reaches 2.6 GW, which even surpasses that of a compact free electron laser in the extreme-ultraviolet region, the peak brightness is estimated to be 1030 photons/(mm2mrad2s). Furthermore, in our TC-HHG experiment, we created a mid-plateau region exhibiting a quasi-continuous spectrum with an energy of 10 μJ. By utilizing this quasi-continuous spectrum at the mid-plateau (14 - 27eV), we also demonstrated the generation of a quasi-IAP. We evaluated the pulse duration of the main pulse of this quasi-IAP to be 375 as with 0.25 intensity satellite pulses at ± 6.7 fs. Our TC method has good energy scalability and high conversion efficiency thanks to the phase matching technique. This demonstration also shows our developed IAP source has enough pulse energy making breakthrough for nonlinear attosecond optics.