Angular asymmetry and attosecond time delay from the giant plasmon resonance in ${\mathrm{C}}_{60}$ photoionization

This combined experimental and theoretical study demonstrates that the surface plasmon resonance in ${\mathrm{C}}_{60}$ alters the valence photoemission quantum phase, resulting in strong effects in the photoelectron angular distribution and emission time delay. Electron momentum imaging spectroscopy is used to measure the photoelectron angular distribution asymmetry parameter that agrees well with our calculations from the time-dependent local density approximation (TDLDA). Significant structure in the valence photoemission time delay is simultaneously calculated by TDLDA over the plasmon active energies. Results reveal a unified spatial and temporal asymmetry pattern driven by the plasmon resonance and offer a sensitive probe of electron correlation. A semiclassical approach facilitates further insights into this link that can be generalized and applied to other molecular systems and nanometer-sized metallic materials exhibiting plasmon resonances.

Figure 1

Two-dimensional (2D) projection of the photoelectron velocity distribution (left) and corresponding angularly integrated photoelectron kinetic energy spectrum (right) for ${\mathrm{C}}_{60}$ measured at synchrotron photon energy of 20 eV. The two peaks at 12.4 and 11.1 eV correspond to ionization from the HOMO and HOMO-1 orbitals, respectively.

Figure 2

Experimental and theoretical variation of the asymmetry parameter, $\beta$, around the plasmon resonance for HOMO (left) and HOMO-1 (right). Dots: experimental data points. Curves: TDLDA calculation with the inclusion of the plasmon resonance at 20 eV (plain) or without including the plasmon resonance (dashed).

Figure 3

Variation of photoemission time delays for HOMO (red [light gray]) and HOMO-1 (blue [dark gray]) photoelectrons with photon energy calculated in TDLDA. Corresponding delays without the correlation phase ${\phi }_{\text{c}}$ are shown in dashed curves.

Figure 4

Comparison between TDLDA and semiclassical calculations of variation of the asymmetry parameter $\beta$ around the plasmon resonance for HOMO (left) and HOMO-1 (right). Plain curves: semiclassical model calculation. Dashed curves: TDLDA calculation. Dots: experimental data points.

Figure 5

Photoemission time delay variation associated with change in the screening potential. (a) Photoemission time delay for HOMO (red [light gray] curve) and HOMO-1 (blue [dark gray] curve) photoelectrons in the semiclassical model (plain curves). Filled areas correspond to the classical electron photoemission time from HOMO (red [light gray] area) and HOMO-1 (blue [dark gray] area). (b) Real part of the potential ${\mathrm{V}}_{\mathrm{scr},r}$ as a function of the photon energy and distance [23]. The purple (dark gray) dashed line represents the ${\mathrm{C}}_{60}$ radius (${\mathrm{R}}_0$) and positions 1 and 2 are respectively the maximum and minimum of screening.