Wow, this conversation seems to have gotten really heated lately, and I'm not quite sure I understand why. There are some rather straight forward comments from @Georg Gesek in particular, that have somehow spawned an argument about reputations - but perhaps I'm misunderstanding @CharlesHirlimann's comment in particular.

So, with deep breath, let me try to add some comments that are hopefully useful.

Wave-particle duality is, to large extent, a construct whereby we attempt to impose our classical notions of wave-like and particle-like characteristics on quantum objects. Naturally, real quantum objects should not be thought of as being either waves or particles, both of these abstractions undermine and confound our understanding, but can nevertheless be useful in certain systems.

What is really happening? We have quantum systems evolving according to a wave equation, that's all.

@Harry ten Brink asked about what the particle properties of photons were: here is a small set of them, and it is important to stress that these are dependent on the measurement process used to determine them (I'll come back to that in a minute, below).

We measure photons as detector clicks. Their energy is quantised. We localise photons at points on screens (or photographic plates). Blackbody radiation spectrum. We observe anti-bunching (Hanbury Brown-Twiss effect) and Hong-Ou Mandel two-photon interference. (To be precise, only the latter two really show particle-like phenomena, the former could all be explained by purely wavelike photons interacting with particle-like detectors).

Observe that all of these phenomena deal with the measurement of the photons, and in particular, measurements that force position localisation or number-state type readout, and this measurement is critical.

What do we think of with wave-like phenomena? Superposition, interference. And these phenomena are also routinely observed through interferometers and diffraction effects that can only be understood on the basis of wave-like properties.

But again, we must return to the fact that the single photon (I'll use the particle-like name because I'm more comfortable with it, but it is just a name) is a quantum entity. Like any quantum particle, if I perform a measurement on it, it must give me a result that is an eigenstate of the measurement apparatus. If I choose that measurement to be such that it gives a particle-like response, don't be surprised that you infer that the photon is a particle. If instead you choose wave-like (i.e. superposition-type response), well, you're going to infer a wave.

As I recall, Ulrich's initial idea was something quite elegant, where both the wave-like and particle-like properties could be seen in the same experiment, and where one should be able to smoothly transition between the two. This is elegant and entirely feasible. What it reinforces is the fact that 'wave-like' and 'particle-like' are classical props that we are imposing to help us understand the results of the measurement. The paradoxes typically arise from assuming that quantum objects really are either particles or waves.

I don't think I've added anything new here, but I hope that I have nonetheless helped.

The photon goes through both slits. Otherwise there would be no interference pattern. That's the wave character of photons. But it is only detected by one of the camera pixels. That's the particle character.

@Dieter:Of course, I agree with the quantum basics that you recall. (Aha, late insight: you probably read only the question and to that you gave the perfect answer; actually the question was intented to be an eye-catcher only and the following proposal was planned to be the message.) Nevertheless, when looking to the slit aperture (as the proposed experiment allows us to do) you will see individual photons seemingly coming out of only one slit each (as you say, you can't detect a single photon on more than one camera pixels). What might surprise even the informed at least at a first glance is that the place where the observational paths split (one for observing the interference pattern, the other watching the slit apertures) is downstreams the slits when folowing the propagation of light. One could expect (although one should'nt, as we will see) from a naive notion of causality that all observations referring to the 'upstream part' of the process must represent 'facts'. This would mean that each photon had in fact come through only one of the slits - which is, of course, not correct.

As I see it, generations of physicists improved their understanding of the 'quantum world' by following Richard Feynman in his vivid discussions of the double slit experiment. With the augmentation proposed here (which had to wait for the availability of sufficiently sensitive electronic cameras) we can add, in an impressive way, to the lessons we got from Feynman.

Dear mr mutze,

yes it's possible with the help of a sensible detector like ccd camera as the intensity of the diffraction is very very poor.

Thanking you.

@Dillip:Would you mind thinking also about the experiment which is proposed in the lower case text after my question? Perhaps your organization can manage to do the experiment? I would be glad to get your thoughts.

Our Institution don't have at present not having such a sensitivity detector of very small quanta photon, If we can think over this issue then I hope we can succeed for the proposed experiment.

Take a look at Sir Geoffrey Ingram Taylor, "Interference Fringes with Feeble Light", Proc. Cam. Phil. Soc. 15, 114 (1909).

It is a very famous series of experiments.

@Peter: Thanks for your reference to a very early low light experiment. What do you think that my proposed experiment will show?

It will show that if there is no attempt to measure, you get interference effects. But if you do make a measurement, there will be no interference.

For more recent research on this topic I refer you to the Aephraim M. Steinberg research group at the University of Toronto: http://www.physics.utoronto.ca/~aephraim/

"A double-slit `which-way' experiment on the complementarity--uncertainty debate" a paper from 2007 goes into this question in detail, including new types of experiments. They have continued work in the area of "weak measurement".

Dear Dr Ulrich Mutze,

Thanks for the explanation that matches my feeling also but let's watch and see when such experimented will be observed. I hope the collaboration among the colleagues will resolve the issue in a much lucid manner. and hoping for a successful experiment that we can design to do so as soon as possible.

@Ulrich and Dillip

The dual nature of the photon has always intrigued me.

Can you tell me what kind of a resolution camera we need for this experiment? Does it exist? And if not do we have the technology to build one?

Dear Sinjab,

Thanks a lot for your idea but I am not sure of it whether it is available any where or not but it is possible to go for it. That is my perception and confidence to go for one first to have on trial basis.

@Dillip and Sinjab: A resolution of 256x256 pixel would be sufficient. There are several cameras with 'electron multiplying CCD' (EMCCD) on the market (search for 'single photon EMCCD') which would be very suitable. The Company Hamamatsu would probably be a good supplier since they have a good name in low light imaging since decades.

@Peter: Many thanks for the Toronto/Steinberg reference. Will read it today!

Ulrich

Do you have an idea of how much this type of camera cost?

Microchannel plates have been used for single-photon detection for a long time: http://www.ncbi.nlm.nih.gov/pubmed/20555472

The plates in my vacuum system pick up 260 nm photons at close to single photon rates, and pick up electrons at single count rates ... imaging quality systems with 50 cm active area run $4,000-$6,000. Plus you need a good camera, and a vacuum chamber.

But a modern camera should cost a lot less, and not require the vacuum chamber!

Thanks Peter.

Ulrich ... I built an ultrafast photo-electron diffractometer during my PhD work. I spent a lot of time shooting very weak pulses of electrons through the worlds tiniest slits - Bragg reflections from thin metallic crystals. The electrons were generated by the photo-electric effect from very weak laser pulses. The laser pulse was split via a polarizing thin film, and the power through each arm was controlled by rotating the polarization with a half-wave plate. The "pump beam" was then attenuated to control the power density applied to the thin metal film - these were non-destructive pump-probe experiments to detect and measure changes to the material caused by the pump pulse.

The other pulse was focused through a second harmonic generating crystal, resulting in a mixture of 780 nm and 390 nm light in each pulse. These were immediately sent through a mixing crystal to generate some 260 nm light, which is mid-UV. Dichroic mirrors were used to split out the UV portion, which entered an ultrahigh vacuum chamber, and struck the back of a fused silica window which had a thin gold coating. The UV is strongly absorbed by the gold, but with an individual energy of 4.5 eV, which is just above the work function for gold, electrons are emitted. Most are scattered away internally, and only the electrons emitted at the very end, in in the correct direction make it out of the film, and enter the strong electric field of the electron gun.

I used a high quality Faraday cup and an electrometer calibrated to picoamps to measure the average charge in the pulses for a series of angles of the laser beam polarization. I was able to generate from 1,000 to 10,000,000 electrons by rotating the initial wave plate.

When the Faraday cup was removed, I carried out some experiments with the UV beam in the vacuum chamber. Though almost all of the UV was absorbed by the gold photocathode film, a tiny amount made it through and was diffracted by the extraction grid at the anode of the electron gun. This grid was a gold TEM grid, 400 mesh squares. This resulted in a cross-like pattern of dots on the MCP, which was, of course, strongly enhanced by the MCP. When calculated, there were only a few photons generating each of the spots in the pattern for the normal intensities that I used.

My main experiments were to measure changes in electron diffraction patterns as the delay between pump pulse and electron pulse were varied. There was always some background noise because the MCP was not gated - 2 ns gating was possible if we had enough money - but the electron diffraction patterns were always visible if I integrated over long enough pulses.

Thus very low counts, perhaps down to single electron, and single UV photon, generated diffraction patterns in these experiments. This is not a surprise - it is exactly what the quantum theory predicts, and has been seen for generations.

Your proposal is to measure which-way information. I don't see how that could be done with my setup ... everything is too small to manipulate.

You can see most of these results here:

http://hdl.handle.net/2027.42/63758

Thanks, Peter, for your vivid sketch of your PHD work. I should have considered MCP myself. Probably such a plate combined with a off-the-shelf CCD camera would be the best solution.

Ulrich

I understand from the introduction to your question that you will do this experiment if you had the required camera. I am getting some quotes on a single phton EMCCD camera. If the price is reasonable and within my means I would be willing to buy the camera. Would you be willing to do this expeiment together? - I will be more than happy to be your student/assistant!

@Sinjab. The most that I could do is to give advice by email. But setting things up had to be done at your site. You better have available the infrastructure of an university optical lab.

@Mutze . I shall consult with few of my collegues at the university to see if this is possible.Well let you know.Thanks any way Ulrich.

@Bradley:Sorry for coming back to your very justified question so late. Unfortunately I have a bad cold and would prefer not create the drawing that certainly would help the most. Let me draw instead by words.

Use usal orthogonal axes x,y and unit 1 mm in our drawing plane.

1. Place a point-like light source L at the origin (0,0).

2. draw a straight line from L to S=(200,0).

3. Draw a camera which looks towards L, has S as the center of its entrance pupil and has its optical axis coincident with the x-axis.

4. Set the focus of the camera such that you have a sharp image of L.

5. Place a double-slit screen symbolized by the line from (100,-25) to (100,25)

6. Place a beam-splitting mirror symbolized by a line from (125,-25) to (175,25).

7. Draw a second camera of the same type with entance pupil at (150,50) looking

downward (i.e. anti-parallel to the y axis).

8. Set the focus of the camera such that you have a sharp image of the double slit screen.

This should clarify the proagation paths and the cameras under consideration. As you see, my 'twin camera' was intended to mean 'camera of the same type as the one used earlier'. Actually it is not important that they have objective lenses of the same focal length, and that they are placed at same optical distance from the light source.

Only important is that one focusses the screen and the other the lightsource.

Of course the beam splitter does not split photons. Just taking this into account makes the expected result so surprising at a first glance: The photons that decided to go to the 'screen view camera' are seen to have passed the screen at a well-defined point of the screen aperture (i.e. one of the two slits). The photons that decided to go to the 'lamp view camera' paint an interference pattern than can't be understood assuming that each photon came trough just one slit.

Sorry for your inconvenience with my drawing by words.

@all: A sketch is now available through the attached link

http://www.ulrichmutze.de/misc/P1020638.JPG

Thank you, Bradley.

100% d'accord.

Although it is not more than an educational demonstration (that it can't involve something scientifically new is clear from the fact that we both know what the expensive cameras will show before we ordered them) it may be educational not only for greenhorns. At least it was educational for me. When thinking about 'which way information' I was not immediately aware that in optical experiments there is a way to obtain such information by simply looking at the way under consideration.

It was educational for me too. I have not ordered the camera because 1) the experiement is not new and 2) the camers was available at the university . I must thank you Ulrich anyway because your idea has put me in touch with few people at the university which has opened up new ideas.

@Issam Sinjab: Please don't misunderstand what Bradley said. The experiment as described may well be new but it's results can be predicted from existing knowledge. So it does not reveil 'new physics'. This is, however, the case with many modern experiments in optics. Also the EPR-type spin correlations in photon pairs generated in spontaneous parametric down-conversion (SPDC) are not new physics. (SPDC itself, of course, is a new version of optical technology.) I see it more in parallel with the demonstration of optical basics by means of a wave trough, which I remember well to have impressed me deeply at school. So doing the experiment in an appealing manner that Bradley envisioned (making a youtube video out of it?) could well be legitimate even for a university department.

Youtube video is definitely a good idea.I shall speak to the physics department.Thanks again Ulrich.

Yes quantum mechanically a probability is there, so let's try and see the results.

@Vesselin: You should be aware that in the description of the proposed experiment the 'photon' enters only as a 'facon de parler'. Actually it deals with 'photo ionisation events'. That these happen at least with the degree of localization given by the pixel structure of the CCD camera is obvious. The questions that you brought up can be be answered in many ways since they refer to theoretical constructs. When I was young enough to consider such questions important to physics I studied localization of particles in terms of covariant projector-valued and positive operator valued measures. So I know of many theoretical frameworks which can be used to generate intelligently looking but useless statements.

This is quite a nice demonstration proposal (took me a while to get what you were aiming for until I saw the picture). I must confess that I don't think that there is a distinct necessity for single photon imaging in this case, although you could certainly build up the coherence pattern with such imaging. The point here is that because the readout is destruction, the photons are all independent, and hence for a demonstration (at least for people who believe in photons) the difference between the patterns on each screen should suffice.

It brings to mind another experiment that I have used in public demonstrations, namely quantum erasure. In that experiment, you also tag the photon pathway, but instead of using the beamsplitter, you use polarisation. Let me explain.

To observe interference you place a needle or other sharp edge in the path of a laser (say diagonal polarisation), so that it is imaged as a shadow on a CCD. The light diffracts around the needle, and so within the shadow region you will see faint fringes. These fringes are the signature of the which path interference. Now replace the needle with another needle, but where an H polariser and V polariser have been attached to the left and right side of the needle respectively. Now the fringes must disappear because the polarisation encodes the which-path information - hence you have tagged each photon as having propagated on the left or right of the needle. Now for the last demonstration, you place a polariser at 45 degree to the horizontal on the far side of the laser. This erases the which path information and restores the interference fringes.

I do agree with Dr Ulrich, but the details of the ray diagram needs complete understanding on the experiment and then we can design it for testing.

Let us have a good team so that the problem can be solved and contribute a innovative experiment for the researches.

@Andrew: Thanks for this nice description. Just to be sure: 'place ... on the far side of the laser' is to place between the needle and the CCD? When you are doing optical demonstration experiments anyway, you are invited to realize also the one I described. You are right that you need not employ low light conditions. You could use direct laser light instead with projection screens istead of CCD's.

@Dillip: If anything is unclear so far, please ask specific questions and I'll try to anwer them. I would really like to see the experiment realized!

I am clear on your experiment and want to be involved with the experiment which looks to me an innovative and precise.

Don't hesitate to contact me for any kind of academic contributions from side.

@Ulrich, Yes, absolutely correct - you place the final polariser between the needle and the CCD. It also helps to use this to prove the which path information by aligning it vertically, then horizontally and showing the response to the audience. The difficulty is in ensuring you pull the polariser out of the path of the beam before rotating it - otherwise you can tend to give the game away!

I had not getting the ray diagram for the same experiment to have an thought on the experiment.

So if Ulrich can display it then it will have a good impact on all of us.

@Dillip: Here is the link to the ray diagram again:

http://www.ulrichmutze.de/misc/P1020638.JPG

Dear Ulrich,

Thank you very much for the cooperation.

Since I am familiar with some of the DS-experiments, I’d like to offer an answer to the outcome of the experiment.

The setup reminds me on the one by Shariar Afshar from 2004. The only difference is the beam splitter, which requires two collecting lenses instead of one. But the principle of focusing on the two slits (or pinholes, like in Afshar’s setup) is the same. The outcome in 2004 has been the insight that the wave-like behavior and the particle-like one occur at the same time with the same photons. If you focus on the light source, you will produce a clear picture of it on your screen. Instead, if you focus on the pinholes (slits) the screen is to some extent out of focus of the light source and the picture of it will be less clear. (The slits can never be recognized as light sources, because they are only sources for the probability wave.) In both cases, if the picture results in an area on the screen, you will achieve the typical interference pattern within the picture – as long both pinholes are open. (If this area is very small, in case of a punctual light source, it will be more difficult to resolve the interference pattern). In case of Afshar’s experiment, the pictures on the screens showed some “dark fringes”, if both pinholes were open. He could install thin wires in front of the collecting lens which had no impact on the picture, when they were in the right place. But the impact showed as soon as he closed one pinhole by diffraction, which resulted in a darker picture.

@Georg: Thank you very much for this 'just to the point' reference, which I did not know. I was just thinking about a very similar augmentation of my proposal. Shariar Afshar's work gives me a lot to think. Thanks again.

Dear Ulrich,

I had gone through the ray diagram and found it is quite possible to arrange the experiment.

A link to the work Georg refers to is as appended:

There is also a Wikipedia article about the Afshar experiment.

http://www.arxiv.org/ftp/quant-ph/papers/0702/0702188.pdf

Dear Ulrich,

I congrats and wish you all the best.

@Georg: Your ideas on what you will see when focussing the light source and when focussing the pinholes differ from what I wrote and so it will not come as a surprise to you when I say they are wrong. Let me explain. When you focus the light source, your camera has the pinhole screen in front of it. It would make virually no difference if you would place the screen inside the objective lens of the camera in the plane where the diaphragm ('Blende' for us Germans) is. So you have the Fourier transform of the pinlhole screen transmission function on the sensor plane. This is just the interference pattern under consideration in the experiment. If you focus on the pinholes you will see those (luminous on a dark background) theoretical abstrusities can't prevent this.

@Dillip: My proposal would be to start with a version with intensive light (laser beam+neutral reduction filter), so that off-the shelf consumer cameras are sufficient. Pinholes instead of slits. Are you interested in really doing the experiment. I would be glad, and would help if this would be desirable.

You are right Ulrich, that's what I meant to explain, I had something else in mind.

Dear Ulrich,

When you are going to experiment the fact that we all agrees. I will be happy to observe such a nice and fundamental experiment.

Dear Ulrich, Me too, I am interest to follow up your results

Dear all,

In the explaining text to my question I said I would do the experiment if I had a suitable low light camera.

Several insightful contributions made clear that the single quantum operation mode is not essential here and that the experiment makes sense for normal light conditions too. Here consumer electronic cameras can be used, and for tentative experimentation, a human observer's eyes.

Such tentative experimentation showed me how inconvenient the missing of an optical table and the of the usual equipment of an optical lab is, even for an experiment so simple as the proposed one.

So I decided not to try to realize a professional version of the proposed experiment and, as a consequence, not to publish results in a scientific journal.

I would like to invite all scientists who have access to an optical lab to consider realizing the proposed experiment or some creative variation of it. As was pointed out by some contributers, the least controversal value of the experiment is as an educational demonstration. So anybody responsible for demonstration experiments in a physics department of a university could make a fine addition to the demos already in place. Of course, I would be glad to get notice of any accomplishments in this direction.

Every idea of experiment is worth realizing, but here is how I see the outcome. If many photons are accumulated, we'll see with one camera the interference pattern, and with the other the two slits. Very good, but what we are interested in is the correlation between a spot in the interference pattern and the slit the photon passed through (Welcher-Weg.) So the experiment should be done one photon at a time. But if there is only one photon, it is recorded either by one camera, eiher by the other, and no correlation can be observed. The reason is both detections are destructive.

Claude, I agree completely.

Ulrich, I am already trying but with some difficulties. Let me reach the final achievement then it will be a book mark experiment contributed by those scientists contributed to this nice idea in particular and the scientific community in general.

Martin,

Experimentation all the way are like that only but we should not deprived of it with out observation.

This question has been deeply experimentally discussed by Alain Aspect and Philippe Grangier who used a beam splitter and a Mach-Zender interferometer.

One photon light states are obtained using one of the component of a pair of correlated photons*.

- Using one photon light states, there is a strong anti-correlation in the photo-detection signals on both sides of a beam splitter. This means that a unique photon is either reflected or transmitted by the beam splitter with a probability that is in accordance with the splitter reflection coefficient.

- One photon, launched through a Mach-Zender interferometer still gives rise to the observation of an interference figure having the right amplitude. This means that the wave associated to the one photon state has passed both arms of the interferometer. - Since then many other experiments have been performed, showing the non-locality of pairs of photons, the interference of non correlated photons, verifying the Bell's inequalities etc.

A good start reading point could be:

- Experimental Evidence for a Photon Anticorrelation Effect on a Beam Splitter: A New Light on Single-Photon Interferences, P. Grangier et al 1986 Europhys. Lett. 1 173

- Wave particle duality for a single photon A Aspect, P Grangier and G Rogerl 1989 J. Opt. 20 119 doi:10.1088/0150-536X/20/3/003

For french readers here is Grangier's thesis:

http://tel.archives-ouvertes.fr/docs/00/05/11/69/PDF/tel-00009436.pdf

* see: http://en.wikipedia.org/wiki/Spontaneous_parametric_down-conversion

Charles,

Nice to see your comprehensive solution to the question raised by many for this question.

Christian,

Your concept should be appreciated by me if not by the scientific community.

Christian,

your view of wave-particle duality is the same as mine.

Since you write

>

you may appreciate a little help for our limited brains which, when it formed in my brain during a nice summer walk, surprised and enjoyed me:

The photon is an inherently 'opportunistic' entity. It lives with the fact that people always ask: 'Are you a particle' or 'are you a wave' although it has a much more complicated internal life (encoded in the wave function) so that any answer to any of these questions is wrong. However, our photon knows that people think and ask in these silly categories and, since it is in some sense dependent on them, it gives the answer which it thinks, that people expect in the given situation. It also knows that people expect consistency and so it memorizes its answers and allows itself to change them only if the situation has changed too.

Or, as a even more personalized metaphor: replace our photon by an agnostic journalist, who works in a country in which two fanatic religious groups struggle for supremacy. Whenever he contacts people he has to show sympathy with the group he thinks these people belong to. Poor journalist, poor photon.

Does this help?

@Charles,

thank you so much for the link to Grangier's thesis. What a pleassure to read!

Hi Christian,

I fear you see the chances of our photon for a self-determined life better than they are. These experimental physicists with their torture equipment are merciless. They enforce answers to their questions. No help!

Ulrich,

With out thinking on cross and prone let's design and experiment the reality with believing on the comments of others.

In my point of view, I find it rather interesting to see very intelligent people with a passion in physics speculating about the "true" nature of particles like photons since many decades. And the only persistent conclusion seems to be the statement of duality of every particle in this beautiful universe. Of course this is the outcome of many intelligent thoughts, but is it wise? Should we ask, whether an apple is an apple or a wave? Is it useful at all to identify the waves in the air which hit our eardrum as particles and call them phonons?

In my humble impression scientists are very creative in finding mathematical expressions for relations between entities in our universe, but frequently mismatch these formulas with the entities themselves.

This seems to be true with the wave-particle dualism. A particle is a part of matter or radiation, as the word already expresses. And a wave is a common behavior of many particles, which carry the waves energy. A wave relies on a medium. How can these very different concepts become mixed up? How can scientists be so careful with their statements most of the time, but using the "wave-particle dualism" indefatigable?

The answer should be that we can calculate some of the behavior of a photon with formulas which are used for both, waves and particles. But this cannot imply such a dualism on the physical entity itself. In my opinion, the "wave-particle dualism" is useless in order to understand the nature of any entity in the universe and should be avoided in order to prevent endless discussions about the true nature of photons like this one.

Anyway, that QED is working gives some hints to the nature of the entities on which our universe is built: 1st, they are capable of interaction and thus defining dots in space. And 2nd they are evolving between two interactions and thus defining periods in time.

Both capabilities represent the particle- and the wave-like qualities, but without any strange dualism of contradicting concepts. In fact, if we follow this path, we could understand how space-time is constructed, quantum-mechanically!

Georg,

I can't see that your essay adds in any way to the question under consideration.

How your 'insight' of particles as points in space and intervals in time tell you what a camera will show if it is focussed on the slits of a double slit experiment?

Urich,

I don't agree with the discussion of Georg. Further may I request you to please accept our Invitation for ICSDTMC2013 .The details are there in our Institute website "www.nitrkl.ac.in".Your presence will make us proud and we can discuss upon many such issues of importance.

@Georg. You seem to believe that physics is a matter of decision but it is not. In some circumstances photon do behave like particles and in others like a wave. This is not a matter of discussion either, it is a fact.

Apple are not quantum objects. Even if you could prepare an apple in a unique intricated state this state would not last long enough for you to make a meaningful observation.

Unfortunately most people do not pay enough attention to what people say, but rather who is saying it.

Quite straight forward Ulrichs experiment is capable to show the coincident qualities of a photon as a particle and a wave. It is not a matter of "circumstances" which decide the behavior.

And the innocent apple, which ought to be no quantum object is simply made of such things. Anyway it was a symbol, you can use it for very "tiny apples" too.

Georg

Can you then please define the particle properties of photons?

@georg. You seem to be confusing science and journalism. Science is about facts that are independant from : i) the place were they are established, ii) the time they are established, iii) the person who finds them.

Yes off course, I agree with Prof. Charles Hirlimann. Science are based on blinded judgment, on the scientific evidence itself only. Otherwise it will be a catastrophic.

Prof. Tawfik,

You are wrong so for as science is concerned because it is the only truth every body believe and apply in their respective fields.

Wow, this conversation seems to have gotten really heated lately, and I'm not quite sure I understand why. There are some rather straight forward comments from @Georg Gesek in particular, that have somehow spawned an argument about reputations - but perhaps I'm misunderstanding @CharlesHirlimann's comment in particular.

So, with deep breath, let me try to add some comments that are hopefully useful.

Wave-particle duality is, to large extent, a construct whereby we attempt to impose our classical notions of wave-like and particle-like characteristics on quantum objects. Naturally, real quantum objects should not be thought of as being either waves or particles, both of these abstractions undermine and confound our understanding, but can nevertheless be useful in certain systems.

What is really happening? We have quantum systems evolving according to a wave equation, that's all.

@Harry ten Brink asked about what the particle properties of photons were: here is a small set of them, and it is important to stress that these are dependent on the measurement process used to determine them (I'll come back to that in a minute, below).

We measure photons as detector clicks. Their energy is quantised. We localise photons at points on screens (or photographic plates). Blackbody radiation spectrum. We observe anti-bunching (Hanbury Brown-Twiss effect) and Hong-Ou Mandel two-photon interference. (To be precise, only the latter two really show particle-like phenomena, the former could all be explained by purely wavelike photons interacting with particle-like detectors).

Observe that all of these phenomena deal with the measurement of the photons, and in particular, measurements that force position localisation or number-state type readout, and this measurement is critical.

What do we think of with wave-like phenomena? Superposition, interference. And these phenomena are also routinely observed through interferometers and diffraction effects that can only be understood on the basis of wave-like properties.

But again, we must return to the fact that the single photon (I'll use the particle-like name because I'm more comfortable with it, but it is just a name) is a quantum entity. Like any quantum particle, if I perform a measurement on it, it must give me a result that is an eigenstate of the measurement apparatus. If I choose that measurement to be such that it gives a particle-like response, don't be surprised that you infer that the photon is a particle. If instead you choose wave-like (i.e. superposition-type response), well, you're going to infer a wave.

As I recall, Ulrich's initial idea was something quite elegant, where both the wave-like and particle-like properties could be seen in the same experiment, and where one should be able to smoothly transition between the two. This is elegant and entirely feasible. What it reinforces is the fact that 'wave-like' and 'particle-like' are classical props that we are imposing to help us understand the results of the measurement. The paradoxes typically arise from assuming that quantum objects really are either particles or waves.

I don't think I've added anything new here, but I hope that I have nonetheless helped.

@Charles

Feynman did not perform the experiment; he described it in his lectures without apparently knowing that the experiment had been performed in the lab.

@Andrew

I was actually proving Georg because exactly what you write the photoelectric effect, clicking is not a proof at all. Correlation, rather anti-correlation of emitted photons by a single molecule is proof

If we are confused by the wave-particle duality of the photon, I think the matter wave function is the real facet of the physical world.

Bobbs,

The question is not wave & particle that come to any one slit rather whether the particle reaching the slit can be detectable or not?

@Andrew Greentree, thank you very much for your precious time reminding everybody on the basic principles of wave-like and particle-like observations. Since I am an entrepreneur, unfortunately I am under-equipped with time for the beautiful science, but @ Charles, I am not a journalist. On the other hand I'd like to advise against neglecting the human factor in science, which is obviously in place by any human scientist conducting research. Sure, over a long period of time, compared to a human lifespan, scientific insights should become commonly accepted, but in human history there have been many detours on this path.

I find it very important to hold tight to @Andrews statement, that any quantum entity is neither a wave nor a particle, as we have defined these concepts. Why? Simply a particle has a well defined trajectory, which any object with observable quantum probabilities has not. And a wave resides on an arbitrary excited medium and is therefore explicitly distributed on this. Quantum entities do not need such a medium a priory and in addition can transmit all their energy (e.g. the absorption of a photon on a detector) to literally one point in space-time.

@Bradley - I guess we have the same comprehension on this. Just for a single photon, which shows a particle-like behavior in space-time in the moment of detection, the location of the single hit is defined by a so called "probability-wave-function", but cannot be foreseen beyond this given information. Therefore the information about this probability and the one about the effective spot, where the interaction takes place are coincident in exactly this manner. This is also the principal answer to @Ulrich.

@Charles, I agree that there is an underlying reality to the interference pattern, which is more than QED tells us. Incidentally this was one of Feynmans most known declarations about this theory, that it lets us calculate remarkably precisely, but has nearly no contribution to our insight what quantum entities really are and where this probabilities arise from.

For my personal part, I have found a graph-database description which explains the origin of any elementary particle (which is an informational entity in this database) in our universe without gaps. The quantum behaviors we observe emerge from he evolution of the connections in this database, which represent together all the information stored in the universe. (just a try to understand)

Georg

"Just for a single photon, which shows a particle-like behavior in space-time in the moment of detection"

Again you create your own classical picture of a "particle/photon". The photo-electric effect on which the detection is based is not proof for that. The click is caused by the quantized nature of the detection material

HARRY,

It should be either wave or particle nature simultaneously for validation.

I haven't had time to follow the many answers. But my take is this: first it doesn't have to be a photon, it could be an electron. So I'll just refer to it generally as a particle. It seems to me that any experiment we devise to tell which slit the particle went through always destroys the information and so we never know.

It's like nature wants to hide a secret from us :-)

However, for me this is the *real* point of a quantum computer. I don't believe for one minute a quantum computer will do anything commercial. However, I do believe it will answer fundamental science questions. For one, it will directly confirm that superposition is a resource you can harness and manipulate. For me, this is indirect evidence that that particle is in a superposition of going through both slits. (Don't forget an n-qubit quantum computer is isomorphic to a binary tree of n-cascaded double slit experiments).

But having said that, I am a true believer in doing experiments, because nature can often show us when we are mistaken in our belief. Therefore, I think Ulrich's proposal is a great idea and if this experiment hasn't been done, it should be tried despite what we "think" we know.

Ulrich,

Why to wait, let us do the experiment and see the consequences of it to establish the truth.

Just coming in with a couple of quick responses to the last few posts:

@Harry ten Brink is absolutely correct regarding the detector click not being proof of the particle-like nature of the photon (although it is certainly consistent with the particle-like nature). The result could either be due to the quantised nature of the absorber or the quantised nature of the photon (or both, but not neither). But I should clarify that photon anti-bunching experiments (Hanbury Brown-Twiss experiment) does confirm the particle-like nature of photons.

@Georg, if you haven't done so already - I hadn't appreciated that you were an entrepeneur at VUT. If you haven't done so already, you should go and visit the excellent groups at VUT of Rauschenbeutel and Schmiedmeyer. Their groups (and I suspect some of the other VUT groups as well) do some of the most beautiful experiments on quantum physics, and will clarify your understanding far more than these (unfortunately imperfect) conversations can.

@Charles +1

@Dilip I suppose one way of doing this would be to have a double slit type interferometer (to generate wave-like interference) with a Hanbury Brown-Twiss detector to confirm that indeed the single photons arrived as particles, but I find the experiments that interpolate between these limits to be far more satisfying [Kocsis et al. Science 332, 1170 (2011)]

@Derek - we can argue commercial, but of course until practical quantum computing is proven, it is only a hypothesis! My feeling is that ultimately, when we have enough control, all computation will be quantum, but then I am a quantum optimist!

But I agree that we are going to get far more fundamental science. We already now have (admittedly contrived) an example of a computation being performed in a quantum sense that is not efficient classical - boson sampling.

The isomorphism you mention is certainly correct, although you do need to be careful as the N-qubit quantum computer is isomorphic to a 2^N path interferometer - it is the exponential that prevents the classical scalability (if it were N cascaded elements, then we could build such a system trivially).

And finally, with regards Derek's point about electrons - absolutely, yes! For some reason we are (collectively) getting hung up about wave-particle duality of photons, but wave-particle duality of electrons (and Bose-Einstein Condensates) is really just as surprising and contains exactly the same physics.

Charles

Reference to Wikipedia does not add to a scientific discussion. In the paragraph you refer to there is a sentence that only makes sense when you are fully familiar with the experiment or a wave-function for those not informed the sentence is Sanskrit:

"(Note that it is not the probabilities of photons appearing at various points along the detection screen that add or cancel, but the amplitudes. Probabilities are the squares of amplitudes."

Andrew

I thought a combination of interferometry and HBT has been done but cannot retrace the publication