Naked Science Forum
Non Life Sciences => Physics, Astronomy & Cosmology => Topic started by: sciconoclast on 21/06/2010 16:44:46
-
Why does the double slit pattern in this experiment disappear when the light paths are vertically separated prior to convergence?
If laser light is split into separate paths and then redirected by mirrors to converge at a screen 15m ahead after passing through different 0.75mm slits, separated by 6.00mm, a double slit pattern occurs if the paths are vertically parallel. A barrier shield is place between the two paths prior to the double slits to assure that the paths are separate. If the two paths are not vertically parallel, separated vertically by 10mm at the double slits, but still converge at the screen there is only a single slit pattern instead of a double slit pattern.
Why is the double slit pattern absent when there is still a convergence of multiple probable paths? What is happening here?
Experiment Diagram
[ Invalid Attachment ]
-
The diagrams are not very clear. Could you also be a bit more explicit in your description: what do you mean by vertically parallel - do you mean the beams from the final two mirrors are parallel? This is a problem because noticeable interference will only occur if the angular spread (divergence) in each beam is greater than the converging angle from the final two mirrors. The drawings make this converging angle look rather large although if the screen is 15m away this is not so bad as the drawing suggests. With the spacings you have I would not expect the barrier screen alone to do anything much; is this the case?
-
Thanks for your inquiry.
Your assumption that the barrier screen alone does not do anything much is correct. It is only there to make sure absolutely no light from one path crosses over to pass through the other slit.
When all the light paths are in a horizontal plane there is double slit spacing. The mirrors are focused for the paths to converge 15.15m from the double screen. This is a very narrow angle with the center of the paths only 6.75mm apart when passing through the slits.
If the screen is moved closer to the double slit or father away the spacing is still double slit spacing but there is two center intensities. At the point of focused convergence the patterns overlap or join to form a double slit pattern with a single intensity.
When the light paths do not lie in the same horizontal plane ( they of course are still in the same skewed plane ) the tight horizontal angle to the screen from the final mirrors is maintained. The vertical angle is not as acute but it is still small as the the vertical separation at the double slits is only 10.00mm.
In this case when the screen is moved closer or farther from the double slit there are two broad separate center bans and peripheral bans all at single slit spacing. At the point of focused convergence they overlap or merge to form a single, single slit pattern.
I hope this answers your question.
-
It still isn't very clear, but these things are difficult to describe in words and the diagrams have not quite got enough resolution to view easily. I assume you mean the interference pattern disappears when the slits are arranged:
I
I
rather than
I I
Is that right?
Even though the angles are arranged so that the beams focus at the same vertical level on the screen?
If this were a simple double slit experiment such a loss in the pattern could be because the slit seperation had become too large for the coherent beam or that the diffraction spread from each slit was not overlapping because of the vertical seperation. But because, in this case, of the arrangement of having seperate paths prior to te slits and angling the mirrors you should have overcome these problems.
Have I restated the problem correctly?
A possible reason is that angling one of the beams slightly to counter the lowering of the slit is too much. Each slit also diffracts so that the probability function is spread sideways and overlaps the one from the other slit. Although you can reconverge the two diffraction patterns by angling the mirror there is a limit. Perhaps a way to look at the problem is to imagine looking back from the screen at the two slits. With conventional double slits you would not be able to resolve the two slits with light of the wavelength being used. If you could do so then you could tell from which slit a photon was emerging and then there could be not interference pattern. If the slits are seperated too much the pattern will disappear. If you seperate them vertically too much this also will happen. This may be more severe now because the diffraction does not help in making the direction from which the photon emerges less distinct in this direction.
This is not a very satisfying answer but its nearly midnight and I have not had time to think about it much. I will give it some more thought. Maybe someone else will chip in.
-
Hello Again
I apologize for the poor descriptions and sketch.
Yes you understand the experiment correctly. AS for spread, at 1m before the target screen the horizontal spread between the laser beam centers is 0.44mm and the vertical spread is 0.65mm ( I gave the distance from the double slit to the screen incorrectly; it is 15.36m. Even I have trouble reading my notes. )
I could be wrong but I think the answer lies elsewhere. I am planing to do a variation on this experiment this weekend which may shed some more light on what is happening.
Thanks Again
-
Okay, I am back.
I was intrigued by the suggested possibility that the vertical angle of incidence could be responsible for the disappearance of the double slit pattern. I redesigned the experiment to test for this.
The different vertical paths still converged at the same point. However, I refocused the mirrors so that horizontally they two paths remained separated. Next, instead of a screen at this point of vertical convergence I placed two mirrors.
The additional mirrors focused the laser paths back over and above the bench to converge at a new target screen above and behind the laser. The paths from the additional mirrors to the target screen are now vertically parallel.
The double slit spacing still disappears when the paths through the double slit are vertically separated. The control configuration, or vertically parallel paths at the double slits, still produces double slit spacing.
I hope this helps clarify the experiment.
-
OK, this is interesting. This would imply that the overlapping of the diffraction pattern from each slit has to somehow be near the slits rather than being recombined at a later point in the path. This does not make sense, certainly if it is not possible to determine (at the screen) from which slit a photon emerged. I'm therefore somewhat baffled. If you replaced the slits with small circular holes clearly you would get an interference pattern, independent of whether spacing is horizontal, vertical or diagonal as there would be no distinction, although the pattern is more complicated and you may need a more powerful laser to get the light amplitude practically sufficient. The only difference between such a system and the one you are using is related to the diffraction pattern for each of the slits.
I will think further as time permits though I suspect a practical problem rather than anything fundamental.
-
Interesting comment.
I decided to test your comment ( " This would imply that the overlapping of the diffraction pattern from each slit has to somehow be near the slits rather than being recombined at a later point in the path. " ) even though you stated that; " This does not make sense. " .
I placed a 1.5m long barrier shield between the two paths directly behind the double slit. The wide double slit spacer allowed me to tilt the shield so that any peripheral light reflected off of the shield face would not be directed to the target screen display. With all of the paths parallel and passing through the double slit at the same height and on unobstructed paths to the screen, the double slit spacing disapeares when the shield is in place.
I am setting up a simpler experiment to test this same principle in a slightly different way. I will post the results as soon as I have them.
Of course there may still be a practical problem lurking that will surface eventually.
Thanks again for your input.
-
A much shorter shield might be informative. If the shield is too long, simple geometry may prevent a direct line from each slit from overlapping enough to give any width to the interference pattern. i.e. the right hand side is just light from the diffraction from the RH slit and the LHS is just light from the LH slit with only a very small region near the centre for any overlap. And you can't rely on much diffraction around the end of the shield.
I look forward to your results.
-
Greetings.
I have performed another experiment which is seems to involve the same principle, whatever that is. I have just posted it under the title " What is happening in this experiment? ".
I am sure that if I reduce the shield in length I will find a point where it is no longer effective; but I have not gone through that exercise yet. Whereas it is true that the shield is blocking the peripheral paths and that those paths are responsible for the interference pattern in an ordinary double slit procedure, in this case it is the center intensities that are recombined at the original target screen or the overhead target screen in the extended phase.
It has occurred to me that a better test might be to configure the experiment so that the shield starts well beyond the double slit, leaving the field immediately after the double slits unobstructed, but obstructing only the spreading paths that actually overlap at the screen. I think this is going to be my next experiment.
-
I'm sure it is to do with the overlapping of the diffraction regions. It will be interesting to see the result of moving the shield further away. It must restore the interference pattern at some point: a very short shield by the screen will obviously do nothing as would a very short shield between the slits. Have you checked the geometry? I am still concerned that the shield is blocking a direct line from the slits to the edges of where you are expecting to see the interference pattern. Both slits should be visible when looking back from the screen or you can't get interference.
-
Modifications Completed.
In the experiment posted under, " what is happening in this experiment ", which was inspired by to test some of the questions raised in this experiment, I moved the longitudinal shield farther away from the double slits which restored the double slit pattern. See the other post for details.