Can a Photon be Split into Pieces?

If a photon goes into a beam splitter (a half-silvered mirror), does it split in half or does it "choose" one of the two directions? In other words, does the whole particle-wave packet stay together in one piece or does one half go one way and the other half the other way?  This should be an easy experiment.  Maybe this could finally  decide once and for all whether it is more useful to think of the photon as a tiny indivisible "point-particle" or a short spherical burst of smooth continuous electromagnetic waves.  Most people today believe the answer is BOTH.  It's commonly called wave-particle duality.

John Clauser's 1973 beam splitter experiment at Lawrence Berkeley seems to prove beyond any doubt that photons are indivisible particles.   It's very convincing, you can easily find his paper online.   John was awarded the 2022 Nobel Prize in Physics, so we have to take his work seriously.

In his experiment, the correlation between the two detectors was nearly zero after subtracting the background count.  His conclusion was that one single photon particle always hits only one detector never both, proving beyond any doubt that it has chosen only one path.  Single photons can not be split in half in a beam splitter.

On the other hand if an ordinary electromagnetic light wave splits equally 50-50, it should cause both detectors to detect the radiation pulse with nearly 100% coincidence.   John Clauser's data says this never happens.   Since 1973 similar beam-splitter experiments have been repeated thousands of times.  Nearly all profesional physicists today would agree, a photon can't be split.

John Clauser's experimental apparatus with beam splitters 
at Lawrence Berkeley, 1973

But is there a way that we can we explain the odd behavior using Maxwell's classical electromagnetic field equations?   Can the classical wave theory model explain why there is zero coincidence?

For this we have to look at another experiment known as the HOM Effect (Hong-Ou-Mandel, 1987).

They discovered that when two photons hit a beam splitter together at the same time, both photons always go one way.  It depends on the relative phase of the two photons.  It can be understood as classical interference of two waves.   Clauser's experiment used only one photon.

What Clauser neglected is that every beam splitter always has two inputs, like it or not.  Just because Clauser left the second port open does not mean there is no second input.  There is always ambient noise or "vacuum fluctuation" coming into the other port.

So if the noise is in phase with Clauser's single input photon, it will go one way,  if it's the opposite phase it will go the other way.  That explains why the photon never goes both ways, i.e. the measured coincidence rate is zero.

This wave theory (HOM effect) explanation for the lack of correlation is only true if the input power is very low, comparable to the noise level.   If the input power level is much higher (like a laser) then both detectors will fire and the coincidence will be high, in agreement with experiments.

So what is my final conclusion?  A photon can be modeled as a short spherical burst of electromagnetic waves produced by an atomic transition.  Sometimes it is useful to think of a photon as a "point-particle" but underneath that QM statistics model is an electromagnetic wave.  Electromagnetic waves can usually be split in half but if the power level is extremely low, comparable to the background noise, the whole wave will go in one or the other direction in a beam splitter due to interference.  That's the HOM effect.

I am NOT saying that quantum mechanics is "wrong."   Far from it.  QM accurately and reliably predicts the outcome of all known photon behavior.   I am saying that it is often useful to understand how light and photons behave by using a classical electromagnetic field model.   When an atom emits a photon, you can choose to imagine it as a statistical point-particle or as a smooth spherical wave pulse.  The advantage of the wave picture is that it is much more intuitive.   Real waves don't "collapse" upon measurement.  Real waves don't get "entangled."  Waves can't be in two places at once.   Waves don't ever violate locality or causality.   Waves are not "spooky."  

For a recent overview of photons see this reference: Zeilinger's Photon Centennial article [Nature Feb 2005]. 

Zeilinger admits, in Box 2, page 2, "Many phenomena thought to be due to the quantum nature of light can actually be explained by using a classical electromagnetic field and by assuming that only the processes of absorption and emission are quantized."

Al Kordesch

Sept. 12, 2023