Particle-only description of single-photon double-slit experiments

[Count Iblis]

http://arxiv.org/abs/0809.0616"

 

[edguy99]

"From these general considerations, it is clear that a minimal
model should be able to account for the memory and
the threshold behavior of real photon detectors. As it
is our intention to perform event-based simulations, the
model for the detector should, in addition to the two features
mentioned earlier, operate on the basis of individual
events."

Interesting, as I read this, a partial hit of a photon on an electron in the detector may not cause ionization but sets that electron up to ionize the atom on a second hit from a photon coming later from the other slit. Counting these combo hits over time that would never have ionized an atom at this location through one slit (since all the photons are coming from the same slit) accounts for the interference pattern with two slits.

The key is the detectors atom's "remember" partial hits. The intensity peak occurs on a line directly from the middle of the two slits since this area would get the most "partial" hits that add up to the most ionizations.

This interpretation is IMO only and will probably change, but it sure is an interesting simulation. Thanks for the post.

PS: I think this means that the interference pattern would only build up over time. If you ran it for a short time, then let the detector atoms settle down, then run it for another short time, then let the detector atoms settle down in sequence like this - would not result in an interference pattern.

 

[lightarrow]

"Messenger. In our simulation approach, we view each
photon as a messenger. Each messenger carries a message,
representing its time of flight. Compatibility with
the macroscopic description (Maxwell’s theory) demands
that the encoding of the time of flight is modulo a distance
which, in Maxwell’s theory, is the wavelength of the
light. Thus, the message is conveniently encoded as a twodimensional
unit vector ei = (e0,i, e1,i) = (cos(phi_i), sin(phi_i)),
where (phi_i) is the event-based equivalent of the phase of the
electromagnetic wave and the subscript i > 0 labels the
messages."

I don't see a significative difference between this and the wave description.

 

[peter0302]

Ok... So in other words if we treat the double slit like a Fresnel biprism, we can reproduce the quantum effects using a classical particle model.

Except the double slit is not a Fresnel biprism. And the fact that interference results in counteless circumstances - irrespective of the physical structure of the device making the choice (whether it be a double slit, a half-silvered mirror, etc) - leads me to believe that no physical explanation will ever suffice for all the various experiments we've seen supporting orthodox QM.

 

[lightarrow]

Furthermore: isn't the fact that "each photon carries a message" a hidden variable description?

 

[peter0302]

Right. If someone constructed a local realistic computer model of entanglement that reproduced the quantum results, I'd be highly impressed.

 

[edguy99]

I don't see a significative difference between this and the wave description.

The difference comes more in what the drawing or animation looks like. If you show a wave as a circle leaving the electron or atom, you should use at least two expanding circles representing the photon at a particular wavelength. You can imagine how difficult this is if you are trying to represent a photon shot from an atom on the sun, arriving at the Earth 8 minutes later. I agree that the calculations are the same.

Ok... So in other words if we treat the double slit like a Fresnel biprism, we can reproduce the quantum effects using a classical particle model.

From the article: "Simulation results. – First, we show that our eventby-
event simulation model reproduces the wave mechanical
results Eq. (1) of the double-slit experiment. Second,
we simulate a two-beam interference experiment
and demonstrate that the simulation data agrees with
Eq. (3). Finally, we present the results for the simulation
of the single-photon interference experiment with a
Fresnel biprism [15], as depicted in Fig. 1."

I think they have done three different ones, but you are right, there are lots more to do.

Furthermore: isn't the fact that "each photon carries a message" a hidden variable description?

As I see it, they are only tracking the normal properties of a photon, perhaps "message" is not the best word. They are really only keeping track of the photons energy/wavelength, its initial polarity (which they assign randomly when the photon is generated) and how far it has gone (used to calculate the probability of a refraction vs a reflection).

 

[peter0302]

Ok I think I understand the paper better, but I would very much like to see the source code for the computer simulation.

I'm most interested in the simulation of the double-slit.

So the hidden variable is the flight path modulo the wavelength? Sort of like Feynman's rotating arrows?


Detector memory was definitely, at one time, a promising candidate for a loophole. I don't know if it still is. But I've said for a long time that there should be - if there isn't already - an experiment whereby 100 different double slits are lined up in a row and a source moves along the length of the row shooting one photon or electron at a time, and seeing if an interference pattern emerges when the images behind all of the slits are superimposed.

 

[edguy99]

... an experiment whereby 100 different double slits are lined up in a row and a source moves along the length of the row shooting one photon or electron at a time, and seeing if an interference pattern emerges when the images behind all of the slits are superimposed.

The results of this experiment would be interesting. The animation would get no interference pattern.

 

[lightarrow]

Furthermore: isn't the fact that "each photon carries a message" a hidden variable description?

As I see it, they are only tracking the normal properties of a photon, perhaps "message" is not the best word. They are really only keeping track of the photons energy/wavelength, its initial polarity (which they assign randomly when the photon is generated) and how far it has gone (used to calculate the probability of a refraction vs a reflection).

I don't think you can consider it as a "normal property of a photon"; phase is a property of the wave, not of the photon; all the photons must be exactly equal, indistinguishable, they cannot have an internal "clock" which signs different times for each photon, if you don't want a hidden variable description.

 

[peter0302]

The results of this experiment would be interesting. The animation would get no interference pattern.

Why do you say that? Wouldn't orthodox QM say there WOULD be an interference pattern?

 

[edguy99]

Why do you say that? Wouldn't orthodox QM say there WOULD be an interference pattern?

Yes, I agree with you that QM says there would be an interference pattern.

"... an experiment whereby 100 different double slits are lined up in a row and a source moves along the length of the row shooting one photon or electron at a time, and seeing if an interference pattern emerges when the images behind all of the slits are superimposed."

As I understand your experiment, if done with this animation, you have wiped out the effect of "detector" memory, hence the superimposed images would not show an interference pattern. Ie. the atoms down the middle of the slits are not getting hit by both sides (they only get hit once from one direction and don't ionize, because the next photon is going through the next experiment slit) and you don't get that bright pattern in the middle of the slits when the images are superimposed.

 

[peter0302]

Well, yes, my experiment would prove or disprove the possibility of detector memory. I suspect, however, that orthodox Qm would prevail as it always does, and that there would be an interference pattern in my experiment.

Another reason I don't think detector memory is on the table any longer is that in many entanglement experiments (DCQE, etc.) a very small detector scans along an axis. If there were some kind of detector memory you'd expect very strange results as the same detector was used to scan the entire length of the interference pattern, as opposed to a single wide detector or an array of detectors.

 

[Kilt]

I'm interested in this statement from the paper:

"In the simulation model, the photons have which-way
information, never have direct communication with each
other and arrive one by one at a detector. Although the
photons know exactly which route they followed, they nevertheless
build up an interference pattern at the detector ..."

Does this contradict the QM expectation that which-way slit "consciousness" should destroy the interference pattern?

But wait: this is a computer simulation of the double slit experiment. What has "consciousness" of what within a computer simulation? What does it mean for the simulated photon to have "information" about its simulated route?

Maybe the interference pattern builds because a human which-way consciousness is required, and there is no human consciousness within the framework of a computer simulation of an experiment.

On the other hand, a human consciousness designed the entire simulation experiment and interprets its outcomes.

Any thoughts on the significance of this experiment for which-way knowledge or consciousness and the Copenhagen Interpretation.

 

[edguy99]

But wait: this is a computer simulation of the double slit experiment. What has "consciousness" of what within a computer simulation? What does it mean for the simulated photon to have "information" about its simulated route?

Maybe the interference pattern builds because a human which-way consciousness is required, and there is no human consciousness within the framework of a computer simulation of an experiment.

As I understand this simulation, you may be reading too much into it. The interference pattern in the simulation is caused because the detectors ions have a memory. Ie, if an ion in the detector gets hit with a small energy photon, that amount of energy is stored with the ion (maybe the electron bounces around the ion faster without escaping) and if it gets hit a second time with another small energy photon it goes off (ionizes). In this way, you get an interference pattern (ie. a peak of energy) right down the center between the 2 slits.

I don't think most people would run the simulation in this way as the photo-electric effect would suggest that the lower energy photons have no effect in ionizing a molecule and that you need a full hit with a photon with enough energy to ionize the molecule. It is an interesting way to produce an interference pattern treating photons purely as particles.

 

[DrChinese]

Where do they show that blocking off one side destroys the interference pattern? I looked but did not see that result, which of course is required to even begin to make their case.

(And of course, anything like "detector memory" is laughable to begin with.)