Historical research guide · 0424 · 1989
Demonstration of single-electron buildup of an interference pattern
Akira Tonomura · Junji Endo · Tsuyoshi Matsuda · Takeshi Kawasaki · Hiroshi Ezawa
In one minute
Feynman called the two-slit experiment with single electrons impossible to do — the apparatus, he said, would have to be built on an impossibly small scale. Tonomura and his colleagues at Hitachi built it anyway, and filmed it. Their electron microscope fires electrons one at a time past a fine charged filament, an electron biprism, which splits each electron’s wave in two and folds the halves back together; a detector that registers essentially every arrival paints the electrons onto a screen as they land. At about a thousand electrons a second the average gap between one electron and the next is 150 kilometres of flight path, so there is effectively never more than one electron inside the instrument. The first arrivals look like scattered dust. By three thousand a banding is visible; by seventy thousand the interference fringes are sharp. Each electron arrives as a particle at a point nobody can predict, and the pattern they build is exactly the one the wave predicts.
Connection
Chapter 2 asks the reader to accept that the vacuum is a field with structure, and this is the cleanest evidence that a single quantum object really does explore both paths through it — the interference is built one electron at a time, with no second electron anywhere in the machine to interfere with. Chapter 10 uses the same instrument: the electron biprism interferometer that resolves this pattern is the tool Tonomura’s group used to read the phase an electron picks up from a vector potential.
From the source
The experiment Feynman described as impossible was performed. The two-slit interference experiment with electrons, called impossible, absolutely impossible to explain in any classical way, and has in it the heart of quantum mechanics, had never been done in just this way because the apparatus would have to be made on an impossibly small scale — and this article reports a realisation of it in the form of biprism interference, recorded as a movie.Abstract; Sec. I, Introduction
The instrument is an electron microscope with a Möllenstedt–Dücker electron biprism and a two-dimensional position-sensitive electron-counting system. Coherent waves from a sharp field-emission tip are accelerated to 50 kilovolts, giving a wavelength of 0.054 ångström; the two beams passing either side of the filament interfere with a fringe spacing of 7000 ångström, magnified two thousand times by two projector lenses onto a detector twelve millimetres across, where the fringes are 1.4 millimetres apart.Sec. III, Experiments; Eqs. 7 to 9; Fig. 3
The electrons really do arrive one at a time. The arrival rate is about one thousand electrons per second across the whole field of view; the source-to-screen distance is 1.5 metres, so the average interval between successive electrons is 150 kilometres of flight path, and each electron’s wave packet is only about one micrometre long. There is very little chance of two electrons being present between source and detector at once, and much less chance of two wave packets overlapping.Sec. III, Experiments
Every electron is counted. A 50-kilovolt electron striking the fluorescent film makes about five hundred photons, which are converted at a photocathode, multiplied in a microchannel plate and positioned by a position sensor, with counting loss and detection noise both held below one percent and detection efficiency approximately one hundred percent.Sec. III, Experiments; Fig. 4
The buildup was photographed frame by frame. At ten electrons and at one hundred the arrivals appear quite at random; a dim figure of the biprism fringes begins to emerge at three thousand; at twenty thousand it is plain, and at seventy thousand — about fourteen thousand electrons per fringe — the fringes are clear. Repeating the run at intensities from five thousand down to two hundred electrons per second left the fringe contrast unchanged within a ten percent experimental error.Sec. III, Experiments; Fig. 5
The authors read the result as an unambiguous demonstration of wave–particle duality. A single electron passes the two paths as a wave and forms a probability interference pattern, and electron–electron interaction plays no role, since the next electron is not even produced from the cathode until long after the preceding one is detected. At the detector the same electron is observed as a localised particle: a position is selected, it cannot be predicted, and it occurs in the probabilistic way dictated by the probability amplitude.Sec. III, closing paragraphs; Sec. IV, Conclusion
The way in
https://doi.org/10.1119/1.16104How to cite it
Akira Tonomura, Junji Endo, Tsuyoshi Matsuda, Takeshi Kawasaki, Hiroshi Ezawa (1989) Demonstration of single-electron buildup of an interference pattern doi:10.1119/1.16104