Double slit experiment - doesn't measurement affect photons?

[Saiker]

All these videos and articles about the Double Slit Experiment say that if we "look" where the single photons go, they act like particles and if we don't "look" they act like waves, creating the interference pattern...

BUT

What does it mean to "look"? We're not using our eyes or any camera after all. How do these detectors work? Don't they have to interact with the photon in order to measure it? If so, doesn't that interaction change the behaviour of the photon and affect the experiment results?

 

[SergioPL]

It means that you set place in the experiment something that would interact with the photons or electrons you are using in the experiment. No matter that you are able to detect the interaction, if it happens, it will destroy the interference pattern.

 

[Buzz Bloom]

What does it mean to "look"?

Hi Saiker:

"look"is a metaphor. It does not mean "seeing" where the photon goes. It means the experimenter has blocked one of the slits so the photon only has one slit to go through.

Therefore the experimenter knows the photon went through a specific slit. In the case in which no slit was blocked the experimenter has no way to know "which slit the photon when through". In that case, the metaphor also says the photon has an equal chance to go through each slit, so the math calculates probabilities of the end point of a "path" based on the statistical combinations of all possible paths. However, ordinary combination of probabilities are not used. The values used which correspond to the probabilities are complex numbers, each being a combination of a real number and an imaginary number. When these complex number are combined for all possible paths, one gets an interference pattern result like what one would expect from two interfering waves from a wave going through both slits.

So, the two cases do not change the way photons behave. Only the calculation results about the distribution of probabilities change.

The following Wikipedia article will give you more details.

https://en.wikipedia.org/wiki/Double-slit_experiment​

Regards,
Buzz

 

[StevieTNZ]

Hi Saiker:

"look"is a metaphor. It does not mean "seeing" where the photon goes. It means the experimenter has blocked one of the slits so the photon only has one slit to go through.

I would dispute that statement, as it is with two slits (neither blocked) with detectors by them which causes the interference pattern to -apparently- disappear.

In principle the photon becomes entangled with the detectors, so does not go through either one slit or the other. When it does is the crux of the famous measurement problem.

 

[Buzz Bloom]

I would dispute that statement, as it is with two slits (neither blocked) with detectors by them which causes the interference pattern to -apparently- disappear.
In principle the photon becomes entangled with the detectors, so does not go through either one slit or the other. When it does is the crux of the famous measurement problem.

Hi Steve:

I don't think we disagree, since I agree with what I interpret you to be saying.

If instead of fully blocking one slit, suppose the experimenter puts some detecting device into the apparatus. This will effect the probability distributions in a manner which cannot be predicted, even probabilistically, unless the specifics are given regarding how the detectors are constructed and where they are placed. However, if the detectors detect anything (other then the location of the photon when its ends it's path) which the experimenter then is able to know, the path probabilities will change, and consequently the interference pattern will also be changed, and possibly even destroyed.

What I was trying to explain to Saiker is that his understanding is based on one or more metaphors which by the nature of QM never completely capture any intuitive understanding of the phenomena. That is only captured by understanding the math. That is what I believe Richard Feynman meant when he said, "No one understands quantum mechanics."

Regards,
Buzz

 

[Physics Footnotes]

There is a lot of confusing, misleading, and outright incorrect language used in discussions of Quantum Theory (including, frustratingly, most textbooks!), and "look" is one of the more popular culprits.

You should interpret the term "look" as to invoke any process that creates an indelible record from the presence of a quantum system, such as a blotch on a photographic plate, a track in a cloud chamber, a flash of a scintillator, a click of a Geiger counter, etc. Such a process is usually referred to as a "Measurement" or "Observation", although even those preferable terms can also be misleading.The main requirement is that everybody agrees that in the presence of a certain quantum system (photon, electron, muon, etc.) 'such-and-such' definitely happened, which means we can talk objectively about experimental results and their probabilities. The rest is story-telling.

When somebody says that incompatible things happen according to whether or not you look, I find it more meaningful to say something like: "When we perform Measurement A, the results lead us to tell Story A about what the system was doing prior to measurement. When we perform Measurement B, the results lead us to tell Story B about what the system was doing prior to measurement." (Loosely speaking, we replace "Whether or not you look" by "Whether you look here or look there".)

Let's take that double-slit experiment, for example. It is not a question of looking or not looking; it is instead a question of deciding whether to place your detectors just behind the slits (Measurement A => Story A) or at the screen some distance behind the slits (Measurement B => Story B).

It doesn't matter how much a measurement interferes with the system being measured, because you are not making both measurements on the same system (at least not in the situations you are talking about). You prepare a single beam of electrons, say, and you make measurement A on some of the electrons and measurement B on other electrons and then try to reconcile the results by creating a narrative that goes beyond the marks on the screen. It is in the apparent incompatibility of the narratives that the legend of the quantum resides.

Most physicists respond by not taking the narrative seriously, and just focusing on those objective indelible records.

 

[MichPod]

As it was already said by other people, many popular descriptions of QM are confusing.
What I can add to the discussion, you do not even need to "measure" a particle at the slit to make the interference disappear. It is enough to have one additional particle B which state changes if your original particle A passes through slit #1 (of two slits), while the B's state does not change if the particle A passes through slit #2. After passing the slits, the particle A is entangled with particle B and that already should destroy the interference.
I took this idea from the popular lectures of L. Susskind.

 

[StevieTNZ]

Hi Steve:

I don't think we disagree, since I agree with what I interpret you to be saying.

If instead of fully blocking one slit, suppose the experimenter puts some detecting device into the apparatus. This will effect the probability distributions in a manner which cannot be predicted, even probabilistically, unless the specifics are given regarding how the detectors are constructed and where they are placed. However, if the detectors detect anything (other then the location of the photon when its ends it's path) which the experimenter then is able to know, the path probabilities will change, and consequently the interference pattern will also be changed, and possibly even destroyed.

I agree with that. I guess your use of "blocking the slit" meant, to me at least, there only being one slit for the quantum system to go through (apparatus or no apparatus).

 

[vanhees71]

All these videos and articles about the Double Slit Experiment say that if we "look" where the single photons go, they act like particles and if we don't "look" they act like waves, creating the interference pattern...

BUT

What does it mean to "look"? We're not using our eyes or any camera after all. How do these detectors work? Don't they have to interact with the photon in order to measure it? If so, doesn't that interaction change the behaviour of the photon and affect the experiment results?

It's not right but popularizations of science. Photons never behave like particles nor like waves. Photons are single-particle Fock states of the electromagnetic field. You cannot simplify it from this statement without getting it somehow wrong. Forget classical pictures about photons. They are as far from classical objects as something can be. You are not even able to define position observables (neither mathematically nor experimentally in the real world). All you can predict and finally also measure in the real world are detection probabilities given a single-photon state, and comparing experiment with theory (Quantum Electrodynamics, QED) in this case leads to results that are among the best of entire physics today. Some quantities like the Lambshift of hydrogen or the anomalous magnetic moment of the electron coincide between theory and experiment at an accuracy of ~12 siginificant digits or so.

The interference patterns, you talk about, occur for doing a double-slit experiment with single photons for many times, if you do not register somehow, through which slit the photon has gone. Of course, given that we don't have a position observable for photons, we have to define, what we mean with that, and that can be done only by describing concrete experimental setups.

One way to get which-way information is to begin with linearly polarized photons and then set a quarter-wave plate in each of the slits, one oriented with $+\pi/4$, the other with $-\pi/4$ angle relative to the polarization direction of the photons. Then, if the photon goes to the first (second) slit it's right-circular (left-circular) polarized, and thus its polarization state (helicity $h=\pm 1$) clearly indicates through which slit the photon came. But now the "partial beams" of photons cannot interfere anymore, because the polarization states are orthogonal to each other, and if you have a superposition of two orthogonal states, they cannot interfere, because
$$\langle 1 + -1|1+-1 \rangle=\langle 1|1 \rangle + \langle 1|-1 \rangle+\langle -1|1 \rangle+\langle -1|-1 \rangle=\langle 1|1 \rangle+\langle -1|-1 \rangle,$$
i.e., there are no mixed terms between photons coming from slit 1 (helicity $+1$ state) and those coming from slit 2 (helicity $-1$ state). So gaining "which-way information" (in this physically well-defined sense) destroys the interference pattern.

If you distort the relative angles of the quarter-wave plates in the slits, you gain only some which-way information, i.e., it's not certain any more through which slit each photon comes, but for a given polarization indicates that it with some more probability it came from a specific slit. Now the polarization states are not orthogonal anymore and you get some interference back, i.e., a somewhat washed-out interference pattern. If you put the quarter-wave plates in the same direction all photons coming to either of the slits have the same polarization, and thus you cannot get any which-way information anymore, but you get the best contrast for the interference pattern possible.

 

[DrChinese]

All these videos and articles about the Double Slit Experiment say that if we "look" where the single photons go, they act like particles and if we don't "look" they act like waves, creating the interference pattern...

BUT

What does it mean to "look"? We're not using our eyes or any camera after all. How do these detectors work? Don't they have to interact with the photon in order to measure it? If so, doesn't that interaction change the behaviour of the photon and affect the experiment results?

There is no requirement that you know which slit(s) the photon traversed. It is enough that you COULD have learned that information. You do not need to block the slits to see the effect. Instead, place a polarizer screen by each slit.

a) If the polarizers are aligned PARALLEL, there WILL be interference.
b) If the polarizers are aligned PERDENDICULAR, there will be NO interference.

The intensity of the output will not change, and the pattern will be symmetric left and right either way. This setup does not tell you which slit the photon passed, but it could be modified to yield that information in the b) version. That would not change the result. Conclusion: there was no specific disturbance of any single photon going through any specific slit in the b) variation, and yet the interference ceased. Presumably that is because it could not go through both slits (in some sense)!

Note that this example is similar to what vanhees71 has stated above.