# How do photons make interference patterns

1. Mar 27, 2005

### Aardwark

I understand that the consensus of opinion regards photons as particles. My difficulty is in picturing how particles can produce interference patterns, even if only one photon is involved? In the double slit experiment for instance?

2. Mar 27, 2005

### ZapperZ

Staff Emeritus

Pay particular attention to the citation of the Marcella paper.

Zz.

3. Mar 28, 2005

### DrChinese

ZapperZ,

I tried googling that paper to find a copy to look at, but had no luck. Do you have a link to it somewhere? Thanks.

4. Mar 28, 2005

### ZapperZ

Staff Emeritus
As far as I know, it is not available for free online. You need to go to the journal website and get it there (via a subscription).

Zz.

5. Mar 28, 2005

### Aardwark

"Most people often do not realize that one CAN describe interference effects (a typical wave phenomena) using photons![1]"

[1]

[1] T. Marcella, Eur. J. Phys., v.23, p.615 (2002).

Quoted by ZapperZ

https://www.physicsforums.com/showth...54&page=1&pp=15 [Broken]

I also couldn't find this reference.

Please, expain how marcella explains interference and diffraction using the photon model of light.

Last edited by a moderator: May 2, 2017
6. Mar 29, 2005

### DrChinese

Apparently this is not available without subscription, but what he does is start from the perspective of a conventional QM description of a prepared state in which a specific observable is being watched. Classical wave theory is completely ignored. When there are 2 slits, there are 2 terms in the resulting probabilty equations. Marcella then calculate and plots the net results (with terms for wavelength and slit separation) and the normal interference pattern appears as a probability amplitude.

Zapperz's point is that the entire derivation is done on the momentum of an individual photon as a particle, not as a wave. The wave-like behavior essentially is a direct result of HUP. Ergo, the usual wave/particle duality is simply an artifact meant to tie back to classical treatments of light. However, it is completely unnecessary since formal QM give results identical to experiment.

7. Apr 1, 2005

### DrChinese

Zz,

So here is my question: Key to this paper is the idea that double slit results are directly tied to the particular setup, and could be predicted using the QM formalism. Jumping over to the Afshar experiment, it would appear that those results would similarly be a conseqence of that particular setup, and conceptually could be predicted in advance by a calculation using the same basic ideas as in the Marcella paper. If so, how could the Afshar experiment still be seen as demonstrating something new or unexpected?

8. Apr 1, 2005

### ZapperZ

Staff Emeritus
Are you trying to force me to read the Afshar non-paper? :) :)

Zz.

9. Apr 12, 2005

### DrChinese

Far be it from ME to force you to read it...

OK, let's try this. As I read the Marcella paper, an interference pattern forms from the build-up of the probabilties of particles hitting various spots. There are terms for each of the two slits. So each slit contributes a certain probabiltiy amplitude at various points on a detector screen. Those amplitudes are expressed in terms of the uncertainty relations for position and momentum.

When the two terms are combined in a certain way, it becomes clear that the terms mimic exactly the effects of overlapping waves. So the math "happens" to match the wave description. This makes sense to me. (That is scary by itself.) They *should* be equivalent views!

Under that viewpoint, it should also be possible to make predictions for hypothetical experiments in which many different permutations of the HUP are tested. For example, the Afshar experiment would now be an exotic permutation of a test of the HUP. Presumably, you could add terms which accounted for the presence of a wire at various spots and that would be equivalent to the observed results. Thus there is no contradiction demonstrated vis a vis complementarity.

10. Apr 13, 2005

### rainbowings

for the benefit of those who don't have access to the papers in question, could you please state the question independently of the reference?

and could you also please explain what you mean by "the usual wave/particle duality is simply an artifact meant to tie back to classical treatments of light" .... ? and I would think that if at all you to had to pick one of the two, it would be the particle like behaviour that becomes more and more manifest as you go to the classical limit (for massive objects, of course), right?

11. Apr 13, 2005

### ZapperZ

Staff Emeritus
Oh, force me anyway..... :)

I reread the Marcella paper, and skimmed through quickly again the Afshar's non-paper - at least as far as I can get it. I will first of all admit that I didn't quite understand the "big deal" with Afshar's paper, so it is possible that I'm missing the subtleties of the experiment. Having said that, I don't find what you described as unreasonable. It did cross my mind when the Afshar's brouhaha came up to write to Marcella and see what he thinks, or if he can reformulate his paper to clearly show the Afshar's results. It appears that you have thought about this more than I have. Maybe you should write to him instead! :)

Zz.

12. Apr 18, 2005

### Aardwark

Of course my original question was unfair as it addressed the central enigma of quantum mechanics. If there had been an answer then I think we’d all know about it anyway.

Morcella’s paper I suspect with respect to the build up of the probabilities of particles hitting various spots it is the equivalent of the wave theory. Since the energy and momentum necessarily will be linked to a wavelength and frequency. In the double slit experiment the slits will precisely define the position of the two paths in the direction at right angles to the slits, HUP then requires that probability for the potential paths that the particle can take to radiate uniformly from each slit. If we assume the momentum is h/wavelength and the intensity of the momentum varies sinusoidally with time we can then obtain the probabilty distribution for the build up of the interference pattern, which is the equivalent of assuming we had a wave function in the first place.

Again if Morcella had clarified the distiction between the wave and particle characteristic of quantum objects he would be as famous as Micky Mouse.

13. Apr 18, 2005

### ZapperZ

Staff Emeritus

Zz.

14. Apr 18, 2005

### DMuitW

Id first like to agree with ZapperZ that indeed, in the case of photons, we can't talk about wave/particle duality, when assuming the notion 'particle' as being (m0>0).
I would like to remark that saying this rules out a duality is contextually irrelevant. If the two slit experiment is conducted with particles having a well defined rest mass (m0>0), the same interference pattern is obtained, thus proving particles with real rest mass to show wave properties.
Furthermore, that leaves it with the unanswered question what relativistic mass as energy density is fundamentally made of , as well as where rest mass is fundamentally made of.
They have to be equal to certain degree, otherwise Einstein's formula would be worthless.

15. Apr 26, 2005

### Aardwark

Sorry , no I didn't. I based my comments on DrChinese's description.

Which suggested we could analyse the behaviour of light in two ways.

1. We can regard light as it passes from a donor atom to an absorber
as a wave. The development of the interference patterns then develop
according to classical mechanics as the photon wave passes through
whatever obstructions there might be between source and detector. Of
course in QM the momentum and energy is delivered to the detector,
wherever it is, all at once as an impulse. Our interference pattern
must act as a probability
distribution governing where the photon is likely to be detected.
Here we have use a deterministic method to develop the interference
pattern and then we are forced into interpreting the result as a probability
distribution in order to make any meaningful predictions about where
we are likely to find the photon. In effect we have to break the
causal chain once we have determined the interference pattern for
that particular experimental situation.

2. The second way is to assume the photon is a particle. If we
take Young's double slit experiment the particle must go through one
or other of the slits or at least if we place detectors in the slits
we will find the photon only at one of them. Assuming a symmetrical
arrangement there will be a 50% chance of finding the photon in one or
the other of the slits. When the photon passes through a slit it's
position in a direction at right angles to the slit is precisely
fixed, HUP then demands that the component of the momentum in that
direction is completely uncertain. From either of the slits the photon
has an equal probability of taking a path at any angle from zero to
180 degrees. However since there are two slits and the photon has an
equal chance of going through either; the position of the photon as it
goes through the slits is not precisely fixed, it can still go through
either one. This increases in uncertainty with respect to position
allowing the uncertainty in momentum to be reduced. The information
we need to calculate the reduction in uncertainty is contained in the
equation p = h/wavelength. And the knowledge that even though the
component of momentum at right angles to each of the slits is
completely unknown the overall magnitude of the momentum will be
unaffected when the photon passes through the slits. For every
possible path demanded by HUP we can plot along the path the intensity
of the particle's momentum. By combining the results from the two
slits, this will allow us to obtain the same probability density
assumed the photon to be a wave initially.

I hope I've got the general idea right, otherwise please put me right.
Unfortunately this solution throws up many new questions. One of
which I would like to be discussed.

How can the probability distributions from the two slits interfere
with each other? Does this mean somehow under the cloak of uncertainty
the particle is actually taking all possible routes and existing at
more than one place simultaneously and at any one place and time there
is more than one version of the particle? Or does it mean some how the
probabilities are built into the experimental arrangement and the
particle with a given momentum just follows the normal laws of
probability?

May be that's three questions, just counting the ?'s ?

16. Apr 26, 2005

### ZapperZ

Staff Emeritus
I'm sorry, but back up a bit. You have zero discomfort about making definitive statements about something BEFORE you actually read it?

I then seriously question (based on what I've read so far) your understanding of the QM treatment in general. If you do not see any problem with forming ideas about something you only understand superficially, what does it say about the source of the rest of your understanding. It explains why your so-called "QM description" of the 2-slit experiment is all wrong, and why your posting is full of errors. Example: "We can regard light as it passes from a donor atom to an absorberas a wave" <--- What IS this?! You are discarding light generated by accelerating charges? What kind of "light" are all those we created from the synchrotron sources all over the world?

I strongly suggest, if you wish to comment on the Marcella paper AND to understand how interference patterns are generated using the QM/photon picture, that you READ it FIRST!

Zz.

17. Apr 26, 2005

### seratend

A simple experiment to understand better the double slits interference patterns

Let's try to give a less known double slit experiment (ghost interference), just to understand better interferences (at least get another view) as well as the meaning of QM measurement and collapse.

A PDC source S sends entangled photons, which we call photons 1 and photons 2 (propagation direction x). A double-slit screen is placed in the path of photons 1 (vertical direction y). We have 2 detectors (D1,D2) that measure the presence of the photons 1 and 2 along the vertical direction (CCD or whatever we want).

-------------------------------------------------------------> x direction

D1 [ ---<--- [double slit screen]-----<-----[PDC source]--->-----]D2

We also assume that the photons emitted by the PDC source S have a large momentum uncertainty so that no first order interference is observed by D1 for photons 1.
For photons 2, first order interference is neither expected, nor is it seen by detector D2.

Now, from D1 detector, we use one signal (a pixel at position y) as a synchronisation signal for detector D2. In other words, we use this D1(y) signal to filter the detector D2 events: D2 now only sees a photon if D1(y) clicks: we have build a new detector D2[D1(y)] that measures (some of) the photons 2.

Question: Does the new detector D2[D1(y)] see any interference pattern? (D1 does not see an interference pattern) Why?

If we close one slit, does the new detector D2[D1(y)] see any interference pattern?

Seratend.

Last edited: Apr 26, 2005
18. May 8, 2005

### Sherlock

There's an experiment that sends light through a double-slit and
no interference effects are observed? When relatively few photons
have been detected ... ok. But, won't an interference pattern
eventually emerge at D1?

19. May 8, 2005

### El Hombre Invisible

ZapperZ, when you gave me this much c--p I thought it was personal. A few scans across various threads have proved me wrong. Aardwark stated his summation as a 'suspicion', validated that it was based on what he thought Dr Chinese had said, and even said 'Sorry'. He doesn't deserve your diatribe. Wise in the ways of quantum mechanics you are; well-versed in the social niceties of polite society you are not. It seems almost as if your intention is to scare off from this site those people whose ignorance of the majority of your knowledge is ugly to you, and in this case it seems to have worked. Aren't mentors meant to display more patience than this?

20. May 31, 2005

### Aardwark

Hola Hombre

Thanks for your support. Zz's comments haven't put me off I just
suffer from a severe addictive condition, Michael Palin calls it
"dromomania". One of the symptoms is frequent lack of access to the
internet.

Among other things its seems Zz does not like my assertion that by
regarding a photon as a particle and invoking HUP we obtain a
description that is mathematically equivalent to regarding the photon
as a wave. It is merely a matter of interpretation. Either way we are
still left with the fundamental problems of causal discontinuity and
uncertainty.

Zz's seems to think my argument is full of errors. I would like him
to clarify where I'm making the errors. The only example he gives so
far seems to be more of an omission rather than an error. I used an
atomically bound electron as the source of the photon he argues that
this does not include light generated by accelerated charges. This is
true but their inclusion would, I believe, not substantially add to the argument?
I would like Zz to demonstrate how their inclusion can help us better
understand the fundamental problems associated with the
electromagnetic transmission of energy.

A