# Photon spin

1. Sep 16, 2007

### jostpuur

I just realized I don't know a simple thing. How do you experimentally verify, that photons are spin 1 particles? At least you cannot do it with the Stern-Gerlach experiment, because photons are not charged. The polarization effect is usually somewhat identified with the spin, but is it really the same? How do you calculate the theory of polarization measurements from the quantum mechanical principles? I've seen only classical explanations.

Suppose I throw a wild idea, that the electromagnetic potential is a 2nd-rank tensor $A^{\mu\nu}$, and four component potential has worked as an approximation of this, because they both give the Coulomb's force similarly, and the weak magnetic effects are the same (hopefully, I'm not 100% sure). What's the experimental results that contradict this?

2. Sep 16, 2007

### genneth

SG should work on photons -- it depends on the magnetic moment, not the charge. However, I don't know if the experiment can be done that way.

Does your wild idea have any predictions which are different from the usual models? If not, then occam's razor gets used on the extra components.

3. Sep 16, 2007

### jostpuur

I haven't thought about this much yet, which of course makes this a bad post. But I thought I could use some information, before trying to calculate stuff.

The 2nd-rank potential would be analogous to the gravitation, so I might guess the photon would be spin 2 then, as the gravitons are. I haven't really solved angular momentum of 2nd-rank field with Noether's theorem yet, but I've heard that the spin 2 is the result.

4. Sep 16, 2007

### ZapperZ

Staff Emeritus
Er.. shouldn't this be verified simply via conservation laws?

For example, the dipole transition in atoms that resulted in the emission of photons requires a specific selection rule as far as the change in the angular momentum quantum number between the initial and final state. Unless one doesn't believe in that conservation law, one already has the evidence that the emitted photon must have a spin angular momentum of 1.

Zz.

5. Sep 16, 2007

### jostpuur

mhh... okey then.

6. Jun 1, 2009

### katsiusa

Hi everybody,
Can I ask about the photon spin. Photon as the theory has spin 1, in which -1 for left-circular polarization and +1 for right-circllar, but the spin 0 state data actually exists. So, the spin of photon in the ellipse polarization and linear polarization, and with natural light we can understand how? I hope you can help me and explain for its. Thank you very much!

7. Jul 7, 2009

### malawi_glenn

are you not familiar with forum behaviour?

i) first try to google it

8. Jul 7, 2009

### conway

Er...I don't think so. You would need to show that there is a particular transition which ought to otherwise occur, and the only reason it doesn't occur is because of spin conservation.

From what I know of thermal emmission, as long as a transition between two states has a net oscillating electric or magnetic moment, the transition will procede. Regardless of the net change of spin.

9. Jul 7, 2009

### ZapperZ

Staff Emeritus
Er... that's what the selection rule is!

Zz.

10. Jul 7, 2009

### strangerep

Circular polarization just specifies a particular pair of basis vectors. Photon
polarization can be expressed wrt other bases, eg linear. The circular
polarization basis is convenient because those states are eigenstates of
the angular momentum operator. A linear-polarized state can be
regarded as a superposition of those two circularly polarized eigenstates.

"Natural" light (by which I presume you mean incoherent light from the sun,
or some other hot substance) needs to be described by a mixed state.
Cf. http://en.wikipedia.org/wiki/Mixed_state

http://en.wikipedia.org/wiki/Stokes_parameters

Jackson, "Classical Electrodynamics", also explains a bit more
about the relevance of Stokes parameters in measurements
of polarization.

HTH.

Last edited: Jul 7, 2009
11. Jul 8, 2009

### conway

I don't believe you can name a specific transition which illustrates your point.

12. Jul 8, 2009

### ZapperZ

Staff Emeritus

Do you dispute this?

Zz.

13. Jul 8, 2009

### conway

Yes, I dispute the "evidence" part. I don't believe the transitions, say, of the Hydrogen atom prove anything about the spin of the photon. Is that what you're claiming?

14. Jul 8, 2009

### ZapperZ

Staff Emeritus
Yes.

So you're claiming that the atomic spectrum, as described in standard QM, is wrong?

Zz.

15. Jul 8, 2009

### ytuab

If the standard QM is correct, the spin 1 of the photon is correct.
If we deny the spin 1 of the photon, we must insist other theories.

For example, in Sommerfeld's theory, the spectrum of the hydrogen (fine structure) is explained by the relativistic mass change (s and p orbital).

Later, in QM, using the selection rule (2S -- x -- 1S, 2P --O - 1S), the fine structure was explained by the spin-orbital interaction.

The energy difference between 2P1/2 and 2P3/2 state was accidentally consistent with the the energy difference by the relativistic mass diffference between S and P orbitals.

16. Jul 8, 2009

### conway

We are not arguing whether quantum mechanics is correct or not. The claim was made that you can find experimental evidence for the spin of photons in the thermal spectrum of hydrogen gas. I believe this is incorrect.

17. Jul 8, 2009

### Civilized

Zz gave a standard correct texbook argument, cf the wiki article on http://en.wikipedia.org/wiki/Selection_rule" [Broken], and this argument can also be found as an example in books like Sakurai's, near the discussion of Wigner-Eckart.

Last edited by a moderator: May 4, 2017
18. Jul 9, 2009

### ZapperZ

Staff Emeritus
I know that you are incorrect. Open any QM text. You disputed what I said, which is right out of standard QM for a standard selection rule.

Based on the responses given by several others, and now based on the nature of YOUR responses, I believe this thread is now done.