On Wheeler's delayed-choice Gedankenexperiment and its laboratory realization

Abstract

Here, we present an analysis and interpretation of the experiment performed by Jacques et al. (2007 Science 315, 966), which represents a realization of Wheeler's delayed-choice Gedankenexperiment. Our analysis is based on the evolution of the photon state, since the photon enters into the Mach-Zehnder interferometer with a removable beam-splitter until it exits. Given the same incident photon state onto the output beam-splitter, BS_output, the photon's state at the exit will be very different depending on whether BS_output is on or off. Hence, the statistics of photon counts collected by the two detectors, positioned along orthogonal directions at the exit of the interferometer, is also going to be very different in either case. Therefore, it is not that the choice of inserting (on) or removing (off) a beam-splitter leads to a delayed influence on the photon behavior before arriving at the beam-splitter, but that such a choice influences the photon state at and after BS_output, i.e., after it has exited from the Mach-Zehnder interferometer. The random on/off choice at BS_output has no delayed effect on the photon to behave as a wave or a corpuscle at the entrance and inside the interferometer, but influences the subsequent evolution of the photon state incident onto BS_output.