Watching Electrons: Experiments in Feynman's Lectures

[Philip Van Hoof]

As an engineer I was interested in the setup of the experiment described in '1-6 Watching the electrons' of Feynman's recently published lectures.

I understand that a method that is used to detect whether the 'particle' passes through hole 1 or 2 is to use a wave of a wavelength that is either longer than or shorter than the distance between the holes.

While trying to map this experiment to the real world I imagined this as making a wave in an shallow pool to see it get disturbed by a person who's slowly walking over the bottom of the pool (the 'particle') and then at the sides of the pool measuring the arriving waves to locate the walking person and to know about his momentum. To complete the mapping to a pool it has two entrances to it and people get asked to walk in a straight line over the bottom while not caring about external influences (our waves might influence them, we don't know).

The chapter explains that a 'terrible thing' happens depending solely on the wavelength of our waves being shorter or longer than the distance between holes 1 and 2 (doing a measurement with a definable answer or not doing one that will yield a definable answer).

A 'disturbance' is caused when it is shorter (it makes the walker stick to his earlier decision which hole he used to enter the pool). When it is longer we see a interference pattern: the walker acts like a wave coming out of the pool's gates, even interfering with future walkers. He's no longer a particle but a wave as for all we know he acts like one. I hope I didn't mix the two results up :-)

In '1–8 The uncertainty principle' an alternative method is discussed allowing the entrace holes to experience a (measurable) recoil up or down based on which hole the 'particle' chose. I'm not convinced that the recoil-movement of the wall has no influence on the 'particle' and its final path.

I wonder if ever experiments have been done where the direction of the wave to detect the 'particle' doesn't come from left or right of the holes, nor from in the middle of the holes where it's left for 2 and right for 1 (like described in Fig 1-4) but from above and/or below 1 and 2.

I also wonder if ever experiments have been done where the time between two 'particles' is very long (does this still create interference).

ps. I realize this is related to this thread, I'll read all of it later: https://www.physicsforums.com/threads/double-slit-experiment-and-watching-the-electrons.727431/

 

[DrChinese]

Welcome to PhysicsForums, Philip!

Look at Naima's post #13 in the thread you mention, which references this:

http://arxiv.org/abs/1110.4309

The act of observation does NOT physically disturb the particle so that the interference is lost. You can see that either way, there is a polarizer in place and it is the relative settings of the polarizers that controls whether there is interference. So I would recommend you start from there.

 

[Philip Van Hoof]

Welcome to PhysicsForums, Philip!

Look at Naima's post #13 in the thread you mention, which references this:

http://arxiv.org/abs/1110.4309

The act of observation does NOT physically disturb the particle so that the interference is lost. You can see that either way, there is a polarizer in place and it is the relative settings of the polarizers that controls whether there is interference. So I would recommend you start from there.

Doesn't a polarizer already assume the 'particle' to be a wave? Isn't that a problem when we're trying to measure wave vs. particle behavior? It ignores 50% of the possibilities :-)

But I agree it's an interesting experiment in the paper.

 

[DrChinese]

Doesn't a polarizer already assume the 'particle' to be a wave? Isn't that a problem when we're trying to measure wave vs. particle behavior? It ignores 50% of the possibilities :)

But I agree it's an interesting experiment in the paper.

If that were true, you couldn't get interference patterns when polarizers are in place... but you can! If the polarizers are crossed, you get no interference. But you do when they are parallel.

 

[mfb]

I also wonder if ever experiments have been done where the time between two 'particles' is very long (does this still create interference).

Yes, and it still gives interference (if nothing else prevents it, of course).
I think the other questions are covered by the previous posts.

 

[Philip Van Hoof]

Yes, and it still gives interference (if nothing else prevents it, of course).
I think the other questions are covered by the previous posts.

Is it the case that the interference gets reset when the light source in Fig. 1-4 of 1.6 gets turned on (when a measurement is made)? I understand this answer as the interference being recorded (a 'wave' that somehow keeps existing between the hole and the backstop) as apparently the time between the shooting of 'particles' doesn't matter: past ones cause interference with new ones anyway.

 

[DrChinese]

Is it the case that the interference gets reset when the light source in Fig. 1-4 of 1.6 gets turned on (when a measurement is made)? I understand this answer as the interference being recorded (a 'wave' that somehow keeps existing between the hole and the backstop) as apparently the time between the shooting of 'particles' doesn't matter: past ones cause interference with new ones anyway.

Interference in these cases is SELF interference. Not interference between different "waves". As mentioned already, the measurement apparatus does NOT in and of itself "reset" the interference. Only if it leads to potential knowledge of "which slit" the particle travels through does interference disappear.

 

[Philip Van Hoof]

Interference in these cases is SELF interference. Not interference between different "waves". As mentioned already, the measurement apparatus does NOT in and of itself "reset" the interference. Only if it leads to potential knowledge of "which slit" the particle travels through does interference disappear.

Aha, so the 'particle' goes through both holes, creates the wave behavior on both sides, and that results in 'self' interference caused by the two waves. No 'recording' needed for that nor resetting by the measurement apparatus, indeed.

Ok, I think I get it. Time for me to go to the next chapter in the lectures.

 

[vanhees71]

I don't know, what's meant by "self-interference". Let's first discuss classical electromagnetic waves. You have one electromagnetic field, which evolves dynamically according to the laws of classical electromagnetism, described by Maxwell's equations. Interference patterns behind a slit, double-slit, gratings and all objects are due to the boundary conditions that describe (effectively) the interactions of the electromagnetic wave with the material of these obstakles.

In odd-dimensional space the pattern can be understood intuitively by making use of Huygen's principle: You can understand the wave field as a superposition of spherical waves originating from each point of the setup which is transparent to the em. wave. In this intuitive picture the interference comes from the relative phase shifts of the single waves hitting a point at the screen, where you observe the em. field (intensity).

Now, in quantum theory you deal not with classical fields, but with a description of probability distributions which are defined by the wave function squared: $P(x)=|\psi(x)|^2$. The math is pretty similar to solving a classical field equation, but what you get is a probability distribution, which you can realize in your experiment by shooting a lot of single particles through your slits. If quantum theory is right, and up to now there's no reason to doubt that, you'll find this distribution on your screen. Mathematically it's clear that there is interference due to the boundary conditions imposed on the wave function by the obstacles making up your slit structure. Each single particle will give a single point on your screen, not some smeared thing looking like the interference pattern. That's the historical reason for Born's rule, stating precisely that the square of the wave function is the probability distribution for a single particle's position under the circumstances defined by your experiment set up (here the double slit hit by a single particle and detecting at which position it may occur at the detector, or you don't find it if it's absorbed somewhere).

This is what quantum theory states, and there is (at least in my opinion) no better "explanation" for the observed facts than this "minimally interpreted" quantum theoretical formalism. Trying to make intuitive pictures in either a classical-particle picture or a classical-field (wave) picture are misleading. Since Bell we know that a classical picture about the behavior of particles, if it exists at all, must become a very complicated non-local theory. So it doesn't make too much sense to say "the interference pattern occurs by interfering the particle or parts of it with itself." Also Schrödinger's original field interpretation fails according to observations. If I shoot a single electron through the slits, I don't get an extended picture showing the interference pattern on the screen but just one spot due to one "pointlike" electron hitting the material in the photo plate. So a classical wave picture in the sense that an electron is in reality some continuously spread entitity like a classical electromagnetic field does not describe the facts either. So it's in some way as Feynman says in his marvelous introductory chapter on wave mechanics in vol. III of the famous lectures: There's no explanation why it works, but quantum theory is like this, and it's very successful in describing nature.

 

[naima]

I wonder if ever experiments have been done where the direction of the wave to detect the 'particle' doesn't come from left or right of the holes

Excited atoms were used in the young experiment. They could decay between the slits and the screen, so a random photon was emitted without a source at a given position.