Supplementary Materials

Spatiotemporal imaging of 2D polariton wave packet dynamics using free electrons

Yaniv Kurman, Raphael Dahan, Hanan Herzig Sheinfux, Kangpeng Wang, Michael Yannai, Yuval Adiv, Ori Reinhardt, Luiz H. G. Tizei, Steffi Y. Woo, Jiahan Li, James H. Edgar, Mathieu Kociak, Frank H. L. Koppens, Ido Kaminer

Materials/Methods, Supplementary Text, Tables, Figures, and/or References

Download Supplement
  • Materials and Methods
  • Supplementary Text
  • Figs. S1 to S7
  • Table S1
  • Captions for Movies S1 to S4
  • References

Images, Video, and Other Media

Movie S1
Video showing the evolution dynamics of a PhP wavepacket. The video is composed of energy-filtered electron measurements with varying time delays. When the laser pump illuminates the left edge of the sample, the wavepacket is formed and propagates toward the right side of the sample. Here, the wavepacket is formed and remains “stuck� near the edge until a time delay of 122.25 ps. Then, as the pump decays (estimated by the signal outside the sample), the wavepacket begins to propagate along the sample and finally decays at a time delay of approximately 123.25 ps. Bottom inset: image of the sample, acquired without energy filtering of the electrons. Top inset: wavepacket’s signal averaged parallel to the sample’s edge versus the distance from the edge.
Movie S2
Video showing the evolution dynamics of a multi-branch PhP wavepacket. The video is composed of energy-filtered electron measurements with varying time delays. When the laser pump illuminates the left edge of the sample, the wavepacket is formed and propagates toward the right side of the sample. Here, a multi-mode wavepacket is formed, which then splits into two distinct wavepackets (at delay time of 122.4 ps) having different group velocities. At time 123.1 ps, a second pulse begins triggering a second excitation at the left edge of the sample, while the first wavepacket still propagates and reaches the right edge. Bottom inset: image of the sample, acquired without energy filtering of the electrons. Top inset: wavepacket’s signal averaged parallel to the sample’s edge versus the distance from the edge.
Movie S3
Simulation of wavepacket dynamics with long-pulse excitation. (a) FDTD simulation of the z component of the electric field in a 55 nm thick, isotopically pure hBN flake on top of a 20 nm Si3N4 membrane. (b) Excitation field profile as a function of time. A red dot denotes the field at each time delay τd (between the electron pulse and excitation laser pulse). The excitation used in the simulation is a dipole located on the left side of the sample. (c) Excitation wavelength over time, showing the chirped profile of the excitation dipole. (d) EFTEM signal, calculated based on the field in (a), extracted using the continuous-PINEM theory in the Materials and Methods. (e) z component of the electric field at the surface of the hBN sample, related to an SNOM measurement. (f) Wavepacket location (using the Gaussian fitting for each time point) according to the free-electron probe signal (blue) and the expected SNOM signal (orange). The comparative analysis of the two approaches provides similar results and a similar estimated group velocity.
Movie S4
Simulation of wavepacket dynamics with short-pulse excitation. (a) FDTD simulation of the z component of the electric field in a 55 nm thick, isotopically pure hBN flake, on top of a 20 nm Si3N4 membrane. (b) Excitation field profile as a function of time. A red dot represents the field at each time delay τd (between the electron pulse and excitation laser pulse). The excitation used in the simulation is a dipole located on the left side of the sample. (c) Excitation wavelength over time, showing the chirped profile of the excitation dipole. (d) EFTEM signal, calculated based on the field in (a), extracted using the continuous-PINEM theory in the SM. (e) z component of the electric field at the surface of the hBN sample, related to an SNOM measurement. (f) Wavepacket location (using the Gaussian fitting for each time point) according to the free-electron probe signal (blue) and the expected SNOM signal (orange). Unlike in Supplementary video 2, here the comparative analysis of the two methods provides different group velocities, which arise from the fact that the different probing methods weight each frequency component differently.