Dynamic Double-Slit Experiment in a Single Atom

A single-atom “double-slit” experiment is realized by photoionizing rubidium atoms using two independent low power lasers. The photoelectron wave of well-defined energy recedes to the continuum either from the $5P$ or $6P$ states in the same atom, resulting in two-path interference imaged in the far field using a photoelectron detector. Even though the lasers are independent and not phase locked, the transitions within the atom impart the phase relationship necessary for interference. The experiment is designed so that either $5P$ or $6P$ states are excited by one laser, before ionization by the second beam. The measurement cannot determine which excitation path is taken, resulting in interference in wave-vector space analogous to Young’s double-slit studies. As the lasers are tunable in both frequency and intensity, the individual excitation-ionization pathways can be varied, allowing dynamic control of the interference term. Since the electron wave recedes in the Coulomb potential of the residual ion, a quantum model is used to capture the dynamics. Excellent agreement is found between theory and experiment.

Figure 1

Irradiation of Rb with two independent photons generates a photoelectron passing through an intermediate state either via path 1 (transition amplitude ${\psi }_1$) or path 2 (${\psi }_2$), resulting in interference related to the relative phases of ${\psi }_1$ and ${\psi }_2$. The (very weak) nonresonant two-photon transitions are represented as dashed lines.

Figure 2

Excitation, ionization, and decay processes in Rb relevant to the experiments.

Figure 3

Different pathways to ionization producing photoelectrons with 0.36 eV.

Figure 4

Results from experiment and theory, normalized when both lasers are on resonance. The data are compared to theory for path amplitudes equal (dashed line) and to allow for cascades (solid line). The red curves are fits to determine peak amplitudes. (a) Results from path 1. (b) Results with only the blue laser resonant, and also shows cascade contributions (inset). The data have these contributions removed. (c) The sum of (a) and (b), and (d) is when both lasers are resonant. (e) The difference between (c) and (d), and so measures the interference term [Eq. (2)]. (f) The relative path phase shifts [Eq. (3)], the theoretical curve showing this difference derived from individual pathways.