Directional photon detector to detect which path?

[Colin Mitch]

I know this is not possible but I want to know why. Why cannot a directional photon (or electron) detector placed on the screen of a 2 slit experiment detect which slit an individual photon comes from without destroying the fringes? Fire the photons one at a time. They can be detected building up on the screen over time into the interferance fringe pattern. Why cannot I build a directional detector into the screen positioned at one of the bright fringes aimed at one of the slits so I can tell at a particular firing of a photon which slit it came through? After all, the photons land individually on the screen anyway. I can see where they land. All I am doing is putting a directional detector set either on or behind the (semi-transparent?) screen at one spot.
I am imagining the detector as a long tube with a small hole at the front and the photomultiplier tube at the back, aimed so that the field of view is one of the slits only.

 

[jewbinson]

I think you can observe which slit the electron goes through and exactly where it ends up on the screen. However, observing an electron as it goes through one of the slits changes the direction of the electron, and so you will not get an interference pattern. You will get the same pattern as if you were firing a machine gun/marbles at the slits.

You can only get an interference pattern by not observing the electrons between the source and the screen.

 

[Colin Mitch]

Yes, I realize QM would entail that the interference pattern would be destroyed. But I am not trying to observe an electron as it goes through one of the slits. My detector is at the screen itself. It detects the electron as it arrives at the screen (or after if the detector is behind the screen). My detector is not at the slit, so it should not interfere with the electron on the way through the slit. It passively detects the arrival of a photon at the screen. It is aimed to point at one of the slits. (I guess I would have to assume the electron traveled in a straight line from the slit). Since I am not interfering with the electron at the slit why would the pattern be destroyed?

 

[Cthugha]

It passively detects the arrival of a photon at the screen.

There is no - as you call it - passive detection for single shot measurements. Every measurement detecting the arrival of a single photon is "active" and therefore destroying the interference pattern. There are things like weak measurements that give you something similar to what you want to have, but they only give info about quantities averaged over many measurement runs.

How could you detect a photon without interacting with it? This is only possible for classical macroscopic objects.

 

[San K]

Yes, I realize QM would entail that the interference pattern would be destroyed. But I am not trying to observe an electron as it goes through one of the slits. My detector is at the screen itself. It detects the electron as it arrives at the screen (or after if the detector is behind the screen). My detector is not at the slit, so it should not interfere with the electron on the way through the slit. It passively detects the arrival of a photon at the screen. It is aimed to point at one of the slits. (I guess I would have to assume the electron traveled in a straight line from the slit). Since I am not interfering with the electron at the slit why would the pattern be destroyed?

A detector behind the screen would do nothing, i guess, because the arriving photons will have already stuck the screen and made an interference pattern.

Also I am not sure if a detector behind the screen would be able to give which-way info.