How does quantum tunneling occur without an observer?

[Demystifier]

I'm going to cut this off here, because you have a very devious, sneaky way of carrying this discussion. Instead of directly addressing what you were asked, you throw off distracting items and avoid answering the questions completely. I don't know what you're trying to pull off here, but I don't have the time to entertain you anymore.

Sorry, but I cannot explain why collapse is needed in this special case before explaining why the collapse is need in the general case. Since I don't know how much about the collapse in general you already know, I asked you to give an example of a phenomenon where the collapse is definitely needed. In that way I would better understand your way of thinking about the collapse, so I could adjust my answer to your way of thinking. Such adjustment is definitely needed, because clearly there is a lot of mutual misunderstanding here.

 

[atyy]

In very few words- tunnelling is a phenomenon associated with $\Psi$, not with <$x$>.

That's fine as a matter of definition. However, you did define it as being associated with <$x$> in your first description in post #6 where you said "You can see an exponentially decreasing probability of seeing the particle outside the box - this is quantum tunnelling."

The difficulty is that these are so close, especially if you read the OP's question it is closer to your first definition of tunneling where <$x$> is used.

 

[Demystifier]

After cooling my brain a little bit, I think I understood what was the source of mutual misunderstanding. It seems that two closely related but still different things are both called "tunelling":
1. The split of wave function into two parts, one of which is on the other side of the barrier.
2. The phenomenon that the particle happens to be on the other side of the barrier.

The 1. of course does not need any collapse, observer, decoherence, measurement, Bohmian trajectories, or anything else of that sort.
The 2. does need something (not necessarily everything) of that sort.

E.bar.goum and ZZ, is it clearer now?

 

[DrChinese]

Isn't electron before a measurement just a probability wave? It doesn't have a particular location, since we can't determine the location of a wave. But when interacting with the barrier, the wave collapses (reduction of multiple states to just one state, with the probability of finding it on the other side of the barrier, and inside the barrier). Where am I wrong?

ZapperZ has explained this, but I will add some different words in case that helps.

All electrons are in a state of superposition (and a state of uncertainty) at ALL times. It is IMPOSSIBLE to collapse any quantum particles' non-commuting properties simultaneously. Think about these statements and you will see the assumption you are making.

When any observation is made, you are potentially changing the basis of what is known - ie position or momentum, or potentially even a partial mix of both. Ditto for other non-commuting pairs. But there is always uncertainty. So when you mention collapse, you are really referring to collapse on some basis. And you are either talking about a collapse that changes the basis or one that does not. But you are never talking about "total" collapse because there is not such thing.

 

[Juraj]

ZapperZ has explained this, but I will add some different words in case that helps.

All electrons are in a state of superposition (and a state of uncertainty) at ALL times. It is IMPOSSIBLE to collapse any quantum particles' non-commuting properties simultaneously. Think about these statements and you will see the assumption you are making.

When any observation is made, you are potentially changing the basis of what is known - ie position or momentum, or potentially even a partial mix of both. Ditto for other non-commuting pairs. But there is always uncertainty. So when you mention collapse, you are really referring to collapse on some basis. And you are either talking about a collapse that changes the basis or one that does not. But you are never talking about "total" collapse because there is not such thing.

I am aware that momentum and position both have their separate superposition states and by knowing one, we don't know the other. In my assumption I was specifically referring to position wave-function for which I stated that it collapses when it interacts with the barrier. And also, I'm familiar with the fact that electons are constantly in the superposition states for the reasons I noted above.

After cooling my brain a little bit, I think I understood what was the source of mutual misunderstanding. It seems that two closely related but still different things are both called "tunelling":
1. The split of wave function into two parts, one of which is on the other side of the barrier.
2. The phenomenon that the particle happens to be on the other side of the barrier.

The 1. of course does not need any collapse, observer, decoherence, measurement, Bohmian trajectories, or anything else of that sort.
The 2. does need something (not necessarily everything) of that sort.

Can someone confirm this?

 

[e.bar.goum]

After cooling my brain a little bit, I think I understood what was the source of mutual misunderstanding. It seems that two closely related but still different things are both called "tunelling":
1. The split of wave function into two parts, one of which is on the other side of the barrier.
2. The phenomenon that the particle happens to be on the other side of the barrier.

The 1. of course does not need any collapse, observer, decoherence, measurement, Bohmian trajectories, or anything else of that sort.
The 2. does need something (not necessarily everything) of that sort.

E.bar.goum and ZZ, is it clearer now?

Yes. Finally! That's what we were arguing about.

 

[e.bar.goum]

That's fine as a matter of definition. However, you did define it as being associated with <$x$> in your first description in post #6 where you said "You can see an exponentially decreasing probability of seeing the particle outside the box - this is quantum tunnelling."

The difficulty is that these are so close, especially if you read the OP's question it is closer to your first definition of tunneling where <$x$> is used.

Not really. I was using imprecise language, but I still was talking about $\Psi$(x) - "probability amplitude" would have been more precise.

Anyway, the issue for pages now has been one of terminology, which is the least interesting kind of physics issue.

 

[Nugatory]

In my assumption I was specifically referring to position wave-function for which I stated that it collapses when it interacts with the barrier.

It does not. It collapses (if you're going to use a collapse interpretation at all, and this thread is shaping up to be pretty good evidence that collapse interpretations do not help understanding tunneling) when and if some interaction localizes it to one side or the other of the barrier.

I keep on finding myself returning to the first post in this thread, where you asked about fusion between two nuclei separated by the Coulomb force barrier so that classically they can never collide. I prepare my two nuclei in a state such that the amplitude for them fusing is non-zero; I blink my eyes; and the next time I look they might be fused. There's nothing here that says that we had to have both nuclei located on the same side of the barrier at any point, or that the interaction with the barrier caused any collapse.

 

[DrChinese]

I am aware that momentum and position both have their separate superposition states and by knowing one, we don't know the other. In my assumption I was specifically referring to position wave-function for which I stated that it collapses when it interacts with the barrier.

As Nugatory points out, this is not correct (actually you don't know on what basis there is collapse).

But assuming it was: placing a particle in a more definite position causes its momentum to become progressively more uncertain. That gives it progressively more values it could possibly have... and some of those will take it through the potential barrier. QED.

 

[e.bar.goum]

I am aware that momentum and position both have their separate superposition states and by knowing one, we don't know the other. In my assumption I was specifically referring to position wave-function for which I stated that it collapses when it interacts with the barrier. And also, I'm familiar with the fact that electons are constantly in the superposition states for the reasons I noted above.

In any case, the measurement/collapse that Demystifier was talking about wasn't due to interactions with the barrier.

Can someone confirm this?

Yes, that's what we were arguing about. I would still say that 1. is a more appropriate use of the term, because 2. is just 1+measurement, which doesn't actually help understand tunneling, or elucidate any more physics. But I'm happy at this stage to agree to disagree with Demystifier on this issue . However, again, that is not what your OP referred to in terms of "collapse due to interactions with a barrier".

 

[atyy]

Yes, that's what we were arguing about. I would still say that 1. is a more appropriate use of the term, because 2. is just 1+measurement, which doesn't actually help understand tunneling, or elucidate any more physics. But I'm happy at this stage to agree to disagree with Demystifier on this issue . However, again, that is not what your OP referred to in terms of "collapse due to interactions with a barrier".

To argue about what "is" is, you used the term "phenomenon", so we should also be happy to disagree on which the more appropriate definition of "phenomenon" is

"No phenomenon is a real phenomenon until it is an observed phenomenon." http://izquotes.com/quote/196799

I've become pretty sure that Bohmians are the purest Copenhagenists.

 

[Nugatory]

Oh man... Just when I thought this thread was back on track and responding to the OP...
OK, as long as everyone is using plenty of smileys, I guess it's still OK.

 

[Swamp Thing]

However, again, that is not what your OP referred to in terms of "collapse due to interactions with a barrier".

Maybe a better description of tunnelling would be, "collapse due to interaction, 'beyond' the barrier and despite the existence of a barrier"

 

[e.bar.goum]

Maybe the right description of tunnelling is, "collapse due to interaction, 'beyond' the barrier and despite the existence of a barrier"

I don't really think that sentence makes sense unless you say "measurement of the location of a particle due to interaction, 'beyond' the barrier and despite the existence of a barrier"

But I'd still describe that as "a measurement of tunnelling" rather than "tunnelling" itself. But, as I think we've established, it's all a little moot, and to be honest, I care a little less every time we go around in a circle about this. But perhaps we should stop using the word "tunnelling" at all, since it clearly causes a lot of angst. (The smiley is for Nugatory ).

But I think most agree in this thread that whether or not you require measurement to say that tunnelling has occurred, that measurement has not occurred due to interactions with the barrier, per the OP.

(As an aside: this thread has been an interesting exercise for me: In nuclear physics, we often picture tunnelling very classically. For alpha decay, say, you picture an alpha particle sitting in the potential vibrating around, hitting the barrier, mostly reflecting off, but each time there's some probability for it to get through, and for the parent nucleus to "decay". But to know whether or not that decay has occurred, some de-coherence/measurement needs to occur. So, presumably, a toy universe that contains only one ²⁴¹Am atom would never actually see it decay. Rather, you'd be in a superposition of ²³⁷Np+$\alpha$ and ²⁴¹Am. Which I suppose is always true anyway. And this is different to fusion again, too, since fusion involves coupling to many internal degrees of freedom, so in some sense, there is an interaction (near) the barrier there.)

 

[Demystifier]

But I think most agree in this thread that whether or not you require measurement to say that tunnelling has occurred, that measurement has not occurred due to interactions with the barrier

I definitely agree with that.

 

[Demystifier]

I've become pretty sure that Bohmians are the purest Copenhagenists.

Not in the sense that they accept the Copenhagen interpretation, but yes in the sense that they take it literally in order to clearly distinguish it from the other (especially Bohmian) interpretations.

 

[Jano L.]

After cooling my brain a little bit, I think I understood what was the source of mutual misunderstanding. It seems that two closely related but still different things are both called "tunelling":
1. The split of wave function into two parts, one of which is on the other side of the barrier.
2. The phenomenon that the particle happens to be on the other side of the barrier.

The 1. of course does not need any collapse, observer, decoherence, measurement, Bohmian trajectories, or anything else of that sort.
The 2. does need something (not necessarily everything) of that sort.

E.bar.goum and ZZ, is it clearer now?

I do not thing this is the root of the disagreement. What I see as important is that ZapperZ considers tunneling to be a physical phenomenon which can be described with Schroedinger equation and wave function that gives probabilities, but no measurement of position is done and no idea of collapse is ever needed to describe this phenomenon. This is how quantum theory is often applied successfully; no measurement of positions of electrons are done, natural autonomous processes are being described and calculations are made that compare well with macroscopic measurements (current, ...)

What you were suggesting is another view, less standard, where we want to continue to describe electrons with wave function and the Schroedinger equation even after the phenomenon of interest has happened, i.e. in the ammeter. I think ZapperZ is right that this has nothing to do with description of tunneling. It has everything to do with description of a measurement of current in the ammeter, but that is a different issue. The tunneling can happen even if no current is being measured.

 

[Demystifier]

I do not thing this is the root of the disagreement. What I see as important is that ZapperZ considers tunneling to be a physical phenomenon which can be described with Schroedinger equation and wave function that gives probabilities, but no measurement of position is done and no idea of collapse is ever needed to describe this phenomenon. This is how quantum theory is often applied successfully; no measurement of positions of electrons are done, natural autonomous processes are being described and calculations are made that compare well with macroscopic measurements (current, ...)

What you were suggesting is another view, less standard, where we want to continue to describe electrons with wave function and the Schroedinger equation even after the phenomenon of interest has happened, i.e. in the ammeter. I think ZapperZ is right that this has nothing to do with description of tunneling. It has everything to do with description of a measurement of current in the ammeter, but that is a different issue. The tunneling can happen even if no current is being measured.

The view which you described (which might coincide with the view of ZZ) perhaps makes sense in the case when a large number of electrons tunnels. This, indeed, corresponds to the tunneling phenomena usually seen in laboratories.

But in principle, it is possible to have a situation in which only one electron tunnels. In such a case, the view which you described would not make much sense. I am convinced that, in the one-electron case, my view is quite standard.

 

[Jano L.]

The view which you described (which might coincide with the view of ZZ) perhaps makes sense in the case when a large number of electrons tunnels. This, indeed, corresponds to the tunneling phenomena usually seen in laboratories.

Yes, I believe this is the case considered in textbooks and by ZapperZ.

But in principle, it is possible to have a situation in which only one electron tunnels. In such a case, the view which you described would not make much sense. I am convinced that, in the one-electron case, my view is quite standard.

I do not see why the view that I described would not make much sense. The equation is the same, the interpretation is the same - it gives probability for position of the electron that is non-zero behind the potential wall. We place a detector (screen, photo-multiplier) behind the potential wall and if it registers, we know the electron has tunneled. We can then compare the frequency of clicks to results based on the Schroedinger equation.

It is possible that the measurements will deviate from the equation - after all, it is quite different experimental setup - but there is no collapse required anywhere in the analysis.

 

[ZapperZ]

The view which you described (which might coincide with the view of ZZ) perhaps makes sense in the case when a large number of electrons tunnels. This, indeed, corresponds to the tunneling phenomena usually seen in laboratories.

But in principle, it is possible to have a situation in which only one electron tunnels. In such a case, the view which you described would not make much sense. I am convinced that, in the one-electron case, my view is quite standard.

No, it isn't. The STANDARD tunneling treatment in textbooks is done on ONE electron, because electron-electron interaction is neglected. So you have a many one-body treatment of tunneling.

Why would the detection of a current due to one electron would somehow allows you to (i) claim that the tunneling must occur only when it is detected and (ii) that the barrier allows you to make a determination of the position of the electron. These are, after all, the two issues that the OP has claimed, and which he had used YOUR posts to justify his view.

The problem here, and why I've given up on this thread, is that you never bothered to offer any physics at all to back up your claim. When I directly asked for it, you pointed me to a book, or avoided my question by offering one-line statement or another question. You will note that I've addressed every single point that you've made, even pointed out to you how I would detect the longitudinal position of an electron.

So before I abandon this thread completely and stop following it, here it is once again.

1. The barrier makes NO positional measurement, be it on one electron, or on many electrons.

2. Tunneling can occur without observation of the tunneling process itself (i.e. at the tunnel junction). What is observed is the effect after the fact, i.e AFTER the tunneling process occurred. If you don't know that what you detected was tunneling current, there's nothing that distinguish this from any ordinary current.

And with that, I am done!

Zz.

 

[Demystifier]

What is observed is the effect after the fact, i.e AFTER the tunneling process occurred.

Just tell me one thing, as an experimentalist. How do you know that there was the fact in the first place, that tunneling itself ocurred, if it wasn't observed?

 

[ZapperZ]

Just tell me one thing, as an experimentalist. How do you know that there was the fact in the first place, that tunneling itself ocurred, if it wasn't observed?

I detect electrons AFTER they passed through the junction. I did not have to observe the ACT of tunneling at the tunnel junction. I observe the current AFTERwards.

This is no different than the 2-slit experiment. I detect the pattern on the screen. I don't have to go looking at what happened at the slit!

Zz.

 

[Demystifier]

I detect electrons AFTER they passed through the junction.

You didn't answer my question, so I will repeat it in a rephrased form. Did the electrons passed through the junction even before you detected them? If your answer is yes, then how do you know they did?

 

[ZapperZ]

You didn't answer my question, so I will repeat it in a rephrased form. Did the electrons passed through the junction even before you detected them? If your answer is yes, then how do you know they did?

I did answer! If you think about it, ask yourself the same question about the double slit. If you didn't detect them at the screen, how would you know that the electrons passed through the slit?

If I give you a black box, and all you could measure is a current going through it, do you know what's in there? All you are measuring is what is coming out, without knowing what is in there. It can be a resistor, or a tunnel junction. You have not observed anything different. All you measure is a current!

Now, where in there does it say that I need to actually observed the actual tunneling at the tunnel junction for it to occur?

I've stated this already in this thread. You, on the other hand, have not answered my questions.

Zz.

 

[Demystifier]

We place a detector (screen, photo-multiplier) behind the potential wall and if it registers, we know the electron has tunneled.

Let the wave function behind the wall be |B>. Similarly, let the wave function on the other side, in front of the wall, be |F>. By solving the Schrodinger equation one obtains that the full wave function is a superposition
|B>+|F>
But when you register the electron behind the wall, you know that the wave function is |B>. So somehow, at some point, you must have a transition
|B>+|F> --> |B>
How this transition happens?

but there is no collapse required anywhere in the analysis.

The collapse is a way to describe the transition above. Do you know another way to describe it?

 

[ZapperZ]

Let the wave function behind the wall be |B>. Similarly, let the wave function on the other side, in front of the wall, be |F>. By solving the Schrodinger equation one obtains that the full wave function is a superposition
|B>+|F>
But when you register the electron behind the wall, you know that the wave function is |B>. So somehow, at some point, you must have a transition
|B>+|F> --> |B>
How this transition happens?The collapse is a way to describe the transition above. Do you know another way to describe it?

Did you miss intro QM? I don't understand why you are rehashing it this way, considering that this is done in textbooks!

Can you go back to the undergraduate treatment of a square potential barrier, and tell me where that doesn't fit into your view, especially in the derivation of the tunnel current?

Zz.

 

[Demystifier]

Now, where in there does it say that I need to actually observed the actual tunneling at the tunnel junction for it to occur?

It says so in many books on quantum foundations, usually under the title "quantum contextuality" or something like that.

 

[ZapperZ]

It says so in many books on quantum foundations, usually under the title "quantum contextuality" or something like that.

Then you and those books are claiming that ALL of the tunneling experiments that we have done (NONE of which actually make any observation right at the tunnel junction itself, but rather the tunneling current) are wrong and not the result of a tunneling phenomena. Would you like to write a rebuttal to all those papers and stake your reputation on that?

This is getting sillier by the minute, and you continue to resort to the most dubious arguments to counter it. I can easily tell you to go read E.L. Wolf's classic book on Tunneling. So there!

Zz.

 

[Demystifier]

Can you go back to the undergraduate treatment of a square potential barrier, and tell me where that doesn't fit into your view, especially in the derivation of the tunnel current?

Of course I can. Just tell me what is your favored undergraduate textbook for these matters, and I will tell you what you want by referring to this particular textbook.

 

[ZapperZ]

Of course I can. Just tell me what is your favored undergraduate textbook for these matters, and I will tell you what you want by referring to this particular textbook.

Then why did you resort to the vague setup of your "wavefunctions"? Read one of the links I gave the OP that showed the derivation of the transmission and reflection amplitudes.

Zz.

 

[Demystifier]

Read one of the links I gave the OP that showed the derivation of the transmission and reflection amplitudes.

I cannot find those links. Can you give me the number of your post where I can find them?

 

[ZapperZ]

I cannot find those links. Can you give me the number of your post where I can find them?

Oh brother!

https://www.physicsforums.com/threa...ithout-an-observer.825120/page-3#post-5183291

Zz.

 

[Demystifier]

Oh brother!

https://www.physicsforums.com/threa...ithout-an-observer.825120/page-3#post-5183291

There is nothing wrong at page 2 of it. But what is said there is not the end of the story. Indeed, the whole text assumes that the reader already knows some general principles of QM, so the text does not bother to say everything what can be said about tunneling. In particular, it does not mention collapse, but it does not mean that collapse has no relevance. In general textbooks the role of collapse is explained in a more general context (not specifically in the context of tunneling), and someone who understood the general role of collapse in QM should be able to apply it in the context of tunneling. That's why I asked you to mention one case where the collapse is relevant (in your opinion), because that should help me to better understand how do you think about collapse. Perhaps then I could adjust my explanation to your way of thinking.

To see that I am not the only one who think that collapse is relevant for tunneling, see
http://abyss.uoregon.edu/~js/21st_century_science/lectures/lec13.html
In the Section "Quantum tunneling" it says
"Notice that the only explanation for quantum tunneling is if the position of the electron is truly spread out, not just hidden or unmeasured. It raw uncertainty allows for the wave function to penetrate the barrier. This is genuine indeterminism, not simply an unknown quantity until someone measures it.
It is important to note that the superposition of possibilities only occurs before the entity is observed. Once an observation is made (a position is measured, a mass is determined, a velocity is detected) then the superposition converts to an actual. Or, in quantum language, we say the wave function has collapsed."

 

[ZapperZ]

There is nothing wrong at page 2 of it. But what is said there is not the end of the story. Indeed, the whole text assumes that the reader already knows some general principles of QM, so the text does not bother to say everything what can be said about tunneling. In particular, it does not mention collapse, but it does not mean that collapse has no relevance. In general textbooks the role of collapse is explained in a more general context (not specifically in the context of tunneling), and someone who understood the general role of collapse in QM should be able to apply it in the context of tunneling. That's why I asked you to mention one case where the collapse is relevant (in your opinion), because that should help me to better understand how do you think about collapse. Perhaps then I could adjust my explanation to your way of thinking.

To see that I am not the only one who think that collapse is relevant for tunneling, see
http://abyss.uoregon.edu/~js/21st_century_science/lectures/lec13.html
In the Section "Quantum tunneling" it says
"Notice that the only explanation for quantum tunneling is if the position of the electron is truly spread out, not just hidden or unmeasured. It raw uncertainty allows for the wave function to penetrate the barrier. This is genuine indeterminism, not simply an unknown quantity until someone measures it.
It is important to note that the superposition of possibilities only occurs before the entity is observed. Once an observation is made (a position is measured, a mass is determined, a velocity is detected) then the superposition converts to an actual. Or, in quantum language, we say the wave function has collapsed."

But you are missing the information on WHERE it is detected! I've been saying, multiple times, that we detect the current! I've even made analogy with the double slit! This, somehow, doesn't seem to sink in with you!

Please note that if you look carefully, and based on this quote, you are also contradicting yourself. If I DO put a detector right at the tunnel junction, then I would have COLLAPSED the wavefunction, and it is no longer "genuine indeterminism", i.e. the electron will no longer be spread out! It means that NO TUNNELING WILL OCCUR, because there is no longer a wavefunction where the electron position spread into the barrier! For tunneling to occur, it must still be an "uncollapsed" wavefunction till AFTER it has undergone tunneling! Then you can measure it to your heart's content!

Zz.

 

[Demystifier]

For tunneling to occur, it must still be an "uncollapsed" wavefunction till AFTER it has undergone tunneling!

So I was right in post #73 that we merely disagree on the definition of what we call "tunneling". By tunneling you mean the definition 1, while, by the same word, I mean definition 2. So there is no reason to argue any more. I can adjust myself to adopt the definition 1, in which case everything you said so far about tunneling is correct. (And vice versa, what I said so far is also correct, if one takes the definition 2 instead.) Are we good now?