Attosecond Electron Correlation Dynamics in Double Ionization of Benzene Probed with Two-Electron Angular Streaking

With a novel three-dimensional electron-electron coincidence imaging technique and two-electron angular streaking method, we show that the emission time delay between two electrons can be measured from tens of attoseconds to more than 1 fs. Surprisingly, in benzene, the double ionization rate decays as the time delay between the first and second electron emission increases during the first 500 as. This is further supported by the decay of the Coulomb repulsion in the direction perpendicular to the laser polarization. This result reveals that laser-induced electron correlation plays a major role in strong field double ionization of benzene driven by a nearly circularly polarized field.

Figure 1

(a) The basic principle of the $3D\text{−}2eAS$. The ionization time delay of two electrons is mapped onto the relative ejection angle in the plane of polarization ($YZ$ plane), while the momentum correlation in the $X$ direction provides field-free electron correlation, which is often dominated by Coulomb repulsion at short time delays. (b) The experimental setup for 3D electron-electron-ion coincidence detection capable of attosecond pump-probe measurement. It features a 3D imaging system with a zero dead-time detection of electron-electron coincidence.

Figure 2

(a) Angle- and time-dependent double ionization (DI) yield of benzene in intense elliptically polarized light field. The angle is defined as the relative ejection angle between electrons in the plane of polarization. The data are plotted as a relative yield between true coincidence events and false coincidence events. False coincidence events are generated by mismatching coincidence electron pairs. As a comparison, the inset shows the result of xenon double ionization, which is dominated by sequential double ionization. (b) The ratio between back-to-back and side-by-side ejection events in the direction perpendicular to the plane of the polarization. The $n$ in the top axis label is a non-negative integer. The black curves are lines connecting data points, while the red ones are smoothed data. The inset is from xenon double ionization.

Figure 3

(a) Angle- and time-dependent double ionization rate from classical ensemble calculations, showing double ionization preferring short time delays and small relative angles. The counts are normalized to that at 110°. The inset is a representative trajectory in the plane of polarization, showing that, after the ionization of the first electron (blue), it returns to the core and knocks out the second electron (green). (b) The ratio between back-to-back and side-by-side events in the direction perpendicular to the plane of polarization, showing at short time delays, back-to-back ejections are preferred due to Coulomb repulsion. The inset shows the same trajectory as in (a) but in the $XY$ plane.

Figure 4

(a) Calculated angle-dependent ionization rate assuming a sequential double ionization at two different laser intensities. These results explain the second feature in Fig. 2 at $\sim 85°$ but not the first and the third feature. (b) 1. The degenerate highest occupied molecular orbitals of benzene, HOMO and HOMO-1. 2. Angular dependence of the total yield for the first ionization of benzene. 3. Angular dependence of the yield of benzene monocation resulting from the ionization of an electron in the HOMO. 4. Angular dependence of the second ionization; benzene monocation with one electron in $\alpha$ HOMO, $\alpha$ HOMO-1, and $\beta$ HOMO-1 is ionized to the dication by ejection of an electron from $\alpha$ HOMO-1 (yellow), $\alpha$ HOMO (green), and $\beta$ HOMO-1 (red). 5. The same as 3 but for ionization of an electron from HOMO-1 to produce benzene monocation with one electron in $\alpha$ HOMO, $\alpha$ HOMO-1, and $\beta$ HOMO. 6. The same as 4 but for ionization of monocation by ejection of an electron from $\alpha$ HOMO (yellow), $\alpha$ HOMO-1 (green), and $\beta$ HOMO (red).