Can a sinlge photon be polarised?

[_PJ_]

As the title! :)

 

[dentedduck]

Yes. The complication in reconciling the wave properties of light with the particle-like photons is in understanding that a single photon can be considered as a short pulse of a wave, or a wave packet. That means that it still has a direction for the oscillation of the electric field, and hence a polarization.Dave

 

[_PJ_]

Thank you.

 

[khemist]

Meaning you could only polarize the light in the way it was already polarized?

 

[_PJ_]

Meaning you could only polarize the light in the way it was already polarized?

The question was regarding the "possibility" of polarising a photon in isolation, not necesarily that when first observing a photon, might it already be polarised.

However, if a photon already has undergone polarisation, note that polarising does not necessarily restrict the polarity to a singular, definite degree, but a range of values, which are affected by quantum probabilities. Also, being a quantum effect, polarised photons can have their polarisation history 'erased' if no observation is made.

 

[A. Neumaier]

Meaning you could only polarize the light in the way it was already polarized?

No. If a photon passes a polarizing filter, it acquires the polarization determined by the filter. If that is different from the original polarization, only a fraction survives the filter, though.

 

[dentedduck]

However, if a photon already has undergone polarisation, note that polarising does not necessarily restrict the polarity to a singular, definite degree, but a range of values, which are affected by quantum probabilities. Also, being a quantum effect, polarised photons can have their polarisation history 'erased' if no observation is made.

I think that the real question is, what does it mean for a photon to _not_ have a polarization? My understanding is that the quantum mechanical description of an unpolarized photon would be one where the quantum state is in a superposition of different states such that an ensemble measurement would yield each polarization 50% of the time. But this is only if you didn't have knowledge of the atom which emitted the photon and didn't know the polarization state originally. I don't know about polarization states being "erased". Not sure what that means. Check out this thesis for a brief description of the superposition state.
http://www.uni-saarland.de/fileadmin/user_upload/Professoren/fr73_ProfEschner/TwoPhotonInterference.pdf


Dave

 

[A. Neumaier]

I think that the real question is, what does it mean for a photon to _not_ have a polarization? My understanding is that the quantum mechanical description of an unpolarized photon would be one where the quantum state is in a superposition of different states such that an ensemble measurement would yield each polarization 50% of the time.

No. An unpolarized photon is a uniform mixture of all possible polarization directions - not superposition. It becomes polarized f it passes a polarization filter.

 

[dentedduck]

Wouldn't that be the semiclassical description? A fully QM description would need to define a quantum state for the photon and that would be a superposition state.

 

[asheg]

As I know photons only can have circular polarization (left rotating or right rotating) and linear polarization are superposition of being in these state. i.e. when
|Photon> = .5 |LHR> + .5|RHR >
it behaves like linearly polarized.
That's why when a molecule emits photon, the electron (which moves from excited state to lower energy state) loses or obtains angular momentum (difference between excited state and ground state's angular momentum).

 

[A. Neumaier]

Wouldn't that be the semiclassical description? A fully QM description would need to define a quantum state for the photon and that would be a superposition state.

QM describes pure states by superpositions and mixed states by mixtures (of superpositions). Most states occurring in Nature are mixed states - pure states usually require special preparation.

 

[_PJ_]

I don't know about polarization states being "erased". Not sure what that means.

The states aren't erased, the quantum history is erased.

Essentially, this claims that where there may be multiple filters in the path of a photon, regardless of assumed 'previous' polarisation, and provided no decoherence occurs through observation until after the photon ought to have passed through ALL filters, only the latter filter will affect the probability of polarisation states, as if any previous filters were simply not applied.

Read up on "Quantum Eraser", although the experiment was tytpically performed as the DOuble-Slit experiment, producing interference patterns rather than examining the polarisation.