my question is as I believe quite simple since I'm new to physics. However here it is: if we take a double slit and constantly shoot helium atoms on it with a constant speed one by one we will see a certain interference pattern on a properly set up screen behind the double slit (scenario 1). If we now set up a detector that tells us which of the slits each atom is passing the pattern on the screen will change (scenario 2), so far so good. But if we were to set up the same detector but without looking at what it tells us and the information wouldn't be saved what would happen? So the machine detects where each atom is passing but nobody looks at the information and it goes basically instantly lost, would we see the pattern from scenario 1 or 2? (Btw I'm not a native English speaker so I apologize if I used unprofessional terms at some point)
-
Hint: in your scenario 2 that you seem to accept, will the interaction between the detector and the atom change in any way depending of whether a human learns about the new state of the detector after the interaction has taken place? – Marius Ladegård Meyer Mar 22 '23 at 16:53
4 Answers
I hope the following explanation isn't too advanced for you. The gist is - in real quantum mechanics you don't need to philosophize about what is and is not a measurement. An interaction with a detector ruins the particles' ability to make an interference pattern.
Interaction with a detector introduces an uncontrolled random phase on the wavefunction which makes the interference pattern disappear. A more sophisticated understanding of quantum mechanics recognizes that there is no distinct moment that can be called a "measurement." Instead, when you have your particles interact with something like a detector, "coherence" (defined as having a well-defined, reproducible phase between two parts of a wave function) is lost, and although the particle remains in a superposition, the two halves of the wave function dont consistently add and subtract from eachother the way they did when they were making the interference pattern. And the result of the experiment is the same as if you had just sent particles through either particular slit one at a time and averaged the results.
- 7,337
-
"n real quantum mechanics you don't need to philosophize about what is and is not a measurement".I disagree.The measurement problem is a real problem because it affects the interpretation of reality. – appliedSciences Mar 22 '23 at 18:22
-
2I recommend this lecture series - it's where I get most of my understanding of loss of coherence and how (particularly for answering a question like this) it's a more useful way of understanding "measurement" https://ocw.mit.edu/courses/8-421-atomic-and-optical-physics-i-spring-2014/resources/lecture-22-coherence-ii/
To some extent there remains a philosophical question that goes something like "what is a conscious entity and when does it decide what part of the wave function to be aware of" but for the purpose of this question there is no such issue.
– AXensen Mar 22 '23 at 18:26 -
PBS spacetime seems to have propagated this misunderstanding that the detector adds a random phase. That's not necessary (see https://physics.stackexchange.com/questions/204100/entanglement-and-coherence/205121#205121, e.g.). Also, if it did add a random phase, then we would expect to see a smeared-out interference pattern, and not two bright lines like we actually do. – A_P Dec 02 '23 at 02:47
-
@A_P Not sure about pbs spacetime, but I linked a lecture series from a nobel prize winner where this idea is presented (although this was a long time ago so I no longer remember where in the lecture). But your second sentence is completely wrong. If the distribution of landing positions through each single slit do not overlap, there will be no interference pattern. Just look at the wiki for the double slit experiment where it compares the single slit result to the double. You do not get two separate bright lines when you add the detector https://en.wikipedia.org/wiki/Double-slit_experiment – AXensen Dec 02 '23 at 22:21
-
@AXensen Do you mean my third sentence? Yes, you're right that if there are actually detectors there, then there will be no interference pattern at the screen whether or not they add a phase shift. What I mean to say is that if they only add a phase shift (and don't entangle with the electron) then you will not reproduce the two bands; you will merely smear out the interference pattern. Are you certain that this Nobel laureate really claims that the decoherence here is caused by adding a random phase? Because it's really not necessary, as my link shows (and as all QC students learn). – A_P Dec 04 '23 at 02:30
-
To be even more clear: if the introduce phase shift is uniform on [0, 2*pi), then you get a perfectly smeared out band with no "interference pattern" at all. This is indeed a kind of decoherence, but not the kind that's interesting in the two-slit experiment (where we see two solid-ish bands, one for each slit). – A_P Dec 04 '23 at 02:34
-
@A_P I think you're repeating the misconceptions of your previous comment. When you add a detector to a slit on the double slit experiment, you do NOT get two distinct bands, each corresponding to the particles that went through one or the other slit. You DO get a "smeared out interference pattern." If particles that go through slit one, and particles that go through slit two do not have any chance to land in the same place on the screen, there will never be an interference pattern. This is the limit where diffraction is insufficient to make an overlap; this is not how double slits are set up. – AXensen Dec 04 '23 at 02:52
-
@AXensen Oh right, what I wrote made no sense. Still, what's bothering me is that a random phase shift isn't necessary for the effect; simple entanglement will do. In that case, why even call out the phase shift (if it indeed happens)? The way decoherence is normally taught in QC, the punchline is that entanglement makes it look as if a random phase has been added. I assume that's where the source of confusion is in the PBS Spacetime segment, but if a Nobel laureate is also saying it, I'd be interested in learning why. – A_P Dec 04 '23 at 16:23
-
@AXensen OK, I watched the talk. Thanks, that was insightful, even though it's about a different experiment. I'm not aware of any detectors used in two-slit experiments that actually work in this way (though if they did, they would lead to decoherence). A simpler setup is to add a waveplate in front of one of the slits, to "tag" the photon with which-way information. This way, there's nothing random about it, and yet we still get decoherence. The interesting part of the two-slit experiment is that simply knowing the path destroys interference. Adding random phases doesn't give path info. – A_P Dec 04 '23 at 17:29
The type of experiment you are thinking are known as Quantum Erasure experiments:
https://en.wikipedia.org/wiki/Quantum_eraser_experiment
The short answer is, yes, you can create an experiment where each photon goes through slits and gets tagged on which slit it passed. If you look at the tag, the interference pattern disappears. However, you can "untag" the photon and you will restore the interference pattern.
In my experience, though, you need to look at the details of each experiment: you can't make a general conclusion. If you really look at the setup of each experiment, you find that it is less "surprising" than the abstract made it look like.
You can see these videos for a better discussion:
https://www.youtube.com/watch?v=l8gQ5GNk16s&ab_channel=Fermilab https://www.youtube.com/watch?v=RQv5CVELG3U&ab_channel=SabineHossenfelder
Hope it helps!
- 19
In general you do not need to make the conscious observation to destroy the pattern, interaction with the light will be enough of an interaction to alter the particle path after the slits ... the slits are no longer part of the wave function.
Interference patterns whether they be photons themselves, electrons or even the Buckyballs ..... are all a result of forces that favour the energy (photons) or masses (electrons, particles) to move to certain areas (bright spots) and not other areas (dark lines). The responsible force is the EM (electromagnetic force) which we say is governed by the EM field which is everywhere.
For photons and electrons the EM field is already active even before the photon or electron leaves the atom, i.e the excited electron in the atom is fully interacting with the EM field over distances. For buckyballs the EM field interaction is more subtle.
- 2,570
-
The easiest double slit experiment to set up is with light - not charged particles. Here they do it successfully with buckyballs (C60) from an oven at 900K https://documents.epfl.ch/groups/i/ip/ipg/www/2016-2017/Traitement_Quantique_de_l_Information/doubleslitwithc60.pdf. But moreover, I think this answer is just wrong - why do you think it matters whether we use voltages or gravity or just velocity? – AXensen Mar 23 '23 at 08:32
-
That's a good question. Another article also proves your point. https://arxiv.org/pdf/1402.1867.pdf . In your referenced paper there is an ionization step (!), this reference is cleaner in that the observation involves no ionization. I'll make some changes above. – PhysicsDave Mar 23 '23 at 16:27
-
-
I only brought up the massless particles to make the broader point that this answer is totally wrong about what causes interference. The dark lines in interference are not in any way due to forces pushing the particles out of those regions. It is because of two halves of the wavefunction cancelling out because they have opposite phase. This is particularly clear in the case of photons (no forces whatsoever act on photons). I'll link to one source explaining the effect https://pressbooks.online.ucf.edu/phy2053bc/chapter/youngs-double-slit-experiment/ – AXensen Mar 24 '23 at 16:11
-
@AndrewChristensen If you can agree that the wave functions are virtual then I think we are in agreement. I'm a Feynman fan, every photon's path is pre-determined ... and this can only be due to the EM field. The EM field is full of virtual as well as real photons .... and the EM field guides everything (photons, electrons, buckyballs). – PhysicsDave Mar 24 '23 at 19:43
-
We are absolutely not in agreement. Your answer says "Interference patters are all a result of forces that favour the photons to move to certain areas." And that's unquestionably wrong and completely at odds with any reputable description of the effect available in any textbook or online. Just because photons are electromagnetism doesn't mean electromagnetic forces act on them. And in the case of neutral particles it's tremendously clear that it's just the base nature of wavefunctions, and not electromagnetic forces, that stops them from landing in dark spots in interference patterns. – AXensen Mar 24 '23 at 20:53
-
@AndrewChristensen "Just because photons are electromagnetism doesn't mean electromagnetic forces act on them" ...... how do photons reflect?, scatter? ... yes we can not deflect a photon with an E or M field but it is the fundamental interactions of electrons thru the EM field that cause photons to change momentum. The EM field is full of real photons (energy carriers) and virtual photons (force carriers) .... both of which have real momentum. – PhysicsDave Mar 25 '23 at 15:29
-
@AndrewChristensen And there are no such things as truly neutral particles, especially if you try and observe them using the EM field (EM field) .... the De Broglie wavelength is tied to the state of the electrons in the matter (https://en.wikiversity.org/wiki/De_Broglie_wavelength). Are you proposing a different force/field to explain matter wave diffraction? ... strong gravity? – PhysicsDave Mar 25 '23 at 15:30
This question can be answered without reference to human interaction, see the reference below for the experimental demonstration using photons.
The general rule for double slit interference is: There will be interference UNLESS there is the possibility, in principle, for determining which-slit information. It does not matter whether you know which slit the particle goes through, it is enough to eliminate the interference that you could have obtained this information.
In the cited experiment, a polarizer is placed in front of each of the 2 slits. When the polarizers are oriented parallel, the traditional interference pattern appears (there is interference). When the polarizers are crossed (orthogonal), there is no interference pattern. (Note that you can vary the angle between the polarizers from 0 to 90 degrees and get a mixture of more or less of the interference pattern.)
In this scenario with slit filters crossed (90 degrees apart), it would be possible to further filter the particles hitting the detection screen to determine which slit the particle went through. Therefore no interference occurs. It matters not that you don't actually obtain this information, it is enough that you could have. Note that in this experiment, you don't need to consider that the human does or doesn't look at the results.
Experiment: https://sciencedemonstrations.fas.harvard.edu/files/science-demonstrations/files/single_photon_paper.pdf
Theory only: https://arxiv.org/abs/1110.4309
- 1,036