Interference from a thin plate is well known. I wonder why it should be thin. The distance between the upper boundary and the lower boundary is always a multiple of some wavelength/2 from 400 to 750. nm. So there always should be interference for some wavelength which is positive and for the others which are negative. Why don't we see interference from thick plates?
Have you considered the coherence length of the source ? For a He-Ne laser at 633nm with a spectral width of 0.1 nm the coherence length is about 4mm. If the optical path difference within the thick plate exceeds the coherence length, you will not see any fringe pattern. For the laser above mentioned, the coherence length is equivalent to a BK7 plate of thickness 1.3 mm.
Have you considered the coherence length of the source ? For a He-Ne laser at 633nm with a spectral width of 0.1 nm the coherence length is about 4mm. If the optical path difference within the thick plate exceeds the coherence length, you will not see any fringe pattern. For the laser above mentioned, the coherence length is equivalent to a BK7 plate of thickness 1.3 mm.
Thank you all very much! For all reasons you pointed out I have serious doubts.
1. Absorption: i dont think it will be problem for a transparent material such as glass of thickness 5 mm
2. Non parallelism: are you sure a thin film is perfectly parallel; there is a problem to measure the thickness of a small object placed between two plates so they build an air wedge and there is interference from the plates. The plates are neither parallel no perfect (roughness)
3. Coherence length: as far as Dirac pointed out only separate photons interfere with themselves. What is CL of a photon - i think infinity. So its wavefunction is reflected from both boundaries without change of perfect sin form and the photon should reflect where WF interfere positive.
I remembered there was an example in Feynman QED strange theory of light about an experiment of Newton with very thick plates and as far as I remember there was interference. Maybe they were highly polushed? I try to check.
Interference from a thick plate can be seen. But normally a spectrally narrow (i.o.w. - highly coherent, see answers above) light is needed. Such interference had an adverse effect on the spectrometers - because most sensors have a protective glass window (typically something like 500 micron of BK7 + ARC - do you consider it thick enough?) - that's why in particular cases it is advised to remove this protective window (at your own risk and voiding any warranty of course). But even with the protective window removed, some pattern will still be recognisable - because of the semiconductor technology sensors behave like having a thick layer on their surface. I have seen a very interesting document on the implication of this effect on the image sensors.
If you want to observe the interference in thick plate on your own - take the photocamera, remove all the optics and shine by the more or less monochromatic point source - e.g. put a pinhole between LED/LD and image sensor. You will immediately observe the interference fringes (due to the interference in protective window). The layout which I have described is used for the lensless in-line holography - so if you look at images, which you can find in works by A.Ozcan, in some cases you will recognize such a fringe. Look e.g. at Fig. 4 in http://www.opticsinfobase.org/oe/fulltext.cfm?uri=oe-18-9-9690&id=198385. I might be wrong with such attribution of the fringe in this particular case - but in general what you will observe is very similar to what is shown on that Fig.
Ilian,
1- in classical physical optics the absorption of the plate will have an impact on the fringes contrast (assuming a perfectly monochromatic and spatially coherent source,i.e. infinite coherence length), the two wavefronts interfering having different amplitudes.
2- non-parallelism is not necessarily an obstacle to interference. In fact interferometry on a plate is a good way to control its parallelism (fringes of equal thickness).
3- I assume you are referring to Newton's experiment on thick plate colors. This is another topic : colored rings can be seen only if you have very rough surfaces. What causes the rings color is the interference of the light scattered by surface roughness or particles. The same effect can be achieved with powder spread in front of a mirror.
If your question refers to single photon interferometry then it is another story. In single photon interferomerty what interferes is not partial wavefronts but the probability amplitudes of a linear superposition state. We can not use the wavefront formalism anymore.
In QED strange theory of light Feynman describes
interference experiment made by Newton where the
plate was 30 000 wavelengths thick. So interference from
thick plate is no more different than from thin?
QM rules everything. There not separate problems. EM
waves are appriximations to photons obeying QM.
It might be misleading to try to describe these phenomena with a naive form of Quantum Mechanics, and it is meaningless in this context to say that the coherence length of "a photon" is infinity. This is a simple classical problem, and the main problem is the coherence length of the light: in a transparent medium if the coherence length is larger that the double of the thickness, you see fringes, otherwise not. Otherwise, you can observe fringes even with white light if you can detect the light with a suitable filter, like an interferometer: that is what is called "white light interferometry" or "low-coherence interferometry".
- Badges
- Science topic
- Similar topics
- Chemistry
- Nanotechnology
- Nanostructures
- Thin Films