A photon is supposed to arrive as one count, not as two neat halves. A new theoretical paper in Physical Review Letters argues that a carefully timed attempt to interrupt a single photon’s reflection would not merely send it down one path or the other. It would create new light at other frequencies.
The work, by three physicists in Norway, examines a deceptively simple setup: send a single photon toward a mirror, then switch that mirror from reflective to transparent while the photon’s wave packet is interacting with it. The calculation says the result is stranger than the usual classroom picture of a photon meeting a beam splitter.
In the familiar beam-splitter case, a single photon has some probability of being transmitted and some probability of being reflected. Quantum mechanics describes that as a superposition until measurement. Put a detector on each path, and only one detector fires. You do not get two half-energy clicks, because the photon is not being carved into tiny classical chunks.
The moving-target mirror case changes the problem. The photon is extended in time, so the mirror can be reflective for part of the interaction and transparent for another part. According to the Physical Review Letters paper, that abrupt change gives the electromagnetic field a sharp temporal edge. Sharp edges are expensive in frequency space.
The mechanism is the same basic Fourier bookkeeping that makes a short electrical pulse spray energy across a broad band of frequencies. A long, clean tone occupies a narrow frequency range. A sudden click needs many frequencies to describe it. If a reflected optical field is cut off quickly, and a transmitted field appears quickly, the field no longer looks like the original narrowband photon.
The Norwegian team’s calculation says that the interrupted photon remains in a quantum superposition involving reflected and transmitted light, but the abrupt switching also produces additional photons at other frequencies. In plainer terms: the attempted photon guillotine does not yield two half-photons. It can generate a small rainbow of new photons, because the field has been forced to change too fast for the original frequency to carry the story alone.
That distinction matters for any experiment claiming to catch this effect. The expected signature is not two detectors each registering half a photon. It is extra light in new frequency components, potentially appearing in both the reflected and transmitted directions.
The experiment is unpleasantly hard
The paper’s proposed physics is not something to test with a normal mirror. The authors calculate that the switch from reflective to transmissive would need to happen in roughly 10 femtoseconds. A femtosecond is 10^-15 seconds, which is well outside the useful range of yanking hardware out of an optical path.
The practical route would use a material whose optical properties can be changed rapidly. The report notes that some semiconductors can be driven from reflective to transparent in about 30 to 100 femtoseconds using ultrafast laser pulses. That is close enough to be relevant, but it creates the obvious lab nuisance: the switching laser is far brighter than the few photons researchers would be trying to measure.
Any real test would also need an on-demand source of single photons with very narrow spectral bandwidth. Narrow bandwidth stretches the photon out in time, making it easier to see whether the mirror’s sudden change has added new frequency components.
There is indirect support for the idea. Similar fast-switching mirrors are already used to shorten ultrashort pulses, and shortened pulses contain broader spectra after reflection. The single-photon version, according to the report, has not yet been observed.
This story draws on original reporting from Ars Technica.
