Few-Cycle Driven Relativistically Oscillating Plasma Mirrors: A Source of Intense Isolated Attosecond Pulses

The conditions required for the production of isolated attosecond pulses from relativistically oscillating mirrors (ROM) are investigated numerically and experimentally. In simulations, carrier-envelope-phase-stabilized three-cycle pulses are found to be sufficient to produce isolated attosecond pulses, while two-cycle pulses will predominantly lead to isolated attosecond pulses even in the absence of carrier-envelope stabilization. Using a state-of-the-art laser system delivering three-cycle pulses at multiple-terawatt level, we have generated higher harmonics up to 70 eV photon energy via the ROM mechanism. The observed spectra are in agreement with theoretical expectations and highlight the potential of few-cycle-driven ROM harmonics for intense isolated attosecond pulse generation for performing extreme ultraviolet-pump extreme ultraviolet-probe experiments.

Figure 1

1D PIC simulation of the laser-plasma interaction using the code LPIC [14]. (a) and (d): $E$ fields of incident (red) and reflected (blue) three-cycle laser pulses for a CE phases of $\phi =\pi /2$ and $\phi =\pi /40$. (b) and (e): Spectra obtained by fast Fourier transform of the $E$ fields after spectral selection by the filter shown as dotted line. (c) and (f): Time domain behavior of the filtered spectra. Note the appearance of the attosecond pulses at the points of the steepest slopes in the individual cycles. The intensity of the attosecond pulses that comprise the attosecond pulse train is in red. The gray Gaussian curve is a fit to the attosecond pulse train to quantitatively deduce its width. The dashed line corresponds to $1/e^2$ of the strongest attosecond pulse.

Figure 2

Variation of the number of attosecond pulses with the CE phase of the laser pulse. (a) The color-coded plot shows in logarithmic scale the train of attosecond pulses as a function of time of occurrence for values of $\phi =0$ to $2\pi$. The yellow bars for each value of $\phi$ indicate the region outside which the satellite pulses are less than $1/e^2$ of the main pulse. (b) Variation of the pulse train width (FWHM) with $\phi$ (points from simulation data, dashed line from fit to the data).

Figure 3

(a) Single-shot raw data of the emitted harmonic radiation. The spectrum (yellow line) is obtained after vertical binning of the record. The horizontal structure is due to the supporting mesh of the Al filter. (b) Recorded spectra in the frequency domain with laser pulses of different duration but comparable laser pulse intensity on target and central wavelength of ${\lambda }_{\mathrm{L}}=815\mathrm{nm}$; in yellow with a pulse duration of ${\tau }_{\mathrm{L}}=3T_{\mathrm{L}}\sim 8\mathrm{fs}$ and in green with a pulse duration of ${\tau }_{\mathrm{L}}=10T_{\mathrm{L}}\sim 28\mathrm{fs}$ (see Ref. 24).

Figure 4

Sample recorded spectra in the frequency domain after background subtraction acquired with random carrier-envelope phase but otherwise identical experimental conditions, i.e., with 60 mJ laser pulse energy on target, a pulse duration of ${\tau }_{\mathrm{L}}=3T_{\mathrm{L}}\sim 8\mathrm{fs}$, and a central wavelength of ${\lambda }_{\mathrm{L}}=815\mathrm{nm}$.