Real-time Single-molecule Imaging Nature Article., Is there a flaw in this new experiment?

[unquantum]

I am trying to understand this new Nature Nanotechnology letter where they diffract dye molecules: "Real-time single-molecule imaging of quantum interference". Just search the title to find the original article. They show the fringes are not parallel and explain it with a gravity effect as a function of particle velocity. The problem is that the beam height at the sensor plane is determined by a projection from the slits which are 100 microns tall. So velocity resolution at the image plane should certainly be smeared out larger than 100 microns. Yet they show velocity resolution near 10 microns. I emailed the author and they wrote:
". If you do the full calculations, you will find, that the measured velocity distributions as a function of height on the detection screen are in agreement with theory."
Really? How? So what is going on here? The low particle density does not suggest a self focusing effect, yet it acts like it is self focusing. There is a major problem with this experiment unless it is made clear how the beam reconstructs in the vertical direction to give velocity resolution. If the beam reconstructs vertically it can do so horizontally. But that would void the whole experiment and the way fringe spacing fits the deBroglie equation.

[unquantum]

Recent correspondence with author:

Dear Mr. Reiter,
concerning your considerations:
1) The equations are of course right, but our source emits molecules in
all directions. Thus a flight parabola is defined by three source, the
grating (which is only written onto a 100µm high window) and the height on
the detection plane. Thus it is wrong to simply enter the distance
source-detection plane into the calculations, since in the plane of the
grating all molecules pass at the same height.

2) Your observatin is right. The high intensity of the higher interference
orders is due to the van der Waals interaction between the molecules and
the grating wall. This is mentioned several times in our paper.

3) Please don't forget, that also the grating is only 100µm high and that,
especially for the slow molecule, the projection is a non valid
approximation.

4)I don't agree. Regarding the high transversal coherence in our
experiment the shape of the fringes is in agreement with the theoretical
predictins.


Best regards,
Thomas Juffmann



On Di, 22.05.2012, 01:54, Eric Reiter wrote:
> Dear Dr Juffmann
>
>
> Regarding your recent article, "Real-time single-molecule imaging of
> quantum interference," I have performed calculations on your data that do
> not make sense to me.
>
> 1) Let's calculate the fall of a particle. We can use (1/2)gt^2, where t
> = time = distance/velocity. For a fast particle Hfast =
> (9.8/2)(2m/340m/s)^2 = 169x10^-6 meters. For a slow particle Hslow =
> (9.8/2)( 2m/140m/s)^2= 1x10^-3 meters. Hslow - Hfast = 830 micrometers.
> But you show only 240 micrometers. Therefore the difference in falls
> should be 3.4 times larger than you show.
>
>
> 2) I used a multiple slit diffraction simulation tool to test what the
> intensity profiles should be. I found your first order fringes were a few
> times brighter than they should be for the given wavelength/slit-width and
> wavelength/slit-spacing ratios. The the tool I used is
> http://wyant.optics.arizona.edu/multipleSl...ltipleSlits.htm. Though
> this tool has fewer slits than yours, I found this did not change the
> intensity ratios.
>
>
> 3) Given the dimensions of your instrument, the velocity resolution should
> cover 0.43 of the sensor plane by the following calculation: The slit
> height is 100 micrometers, and the projection to the sensor plane should
> make this 2/(2 - 0.56) larger, that is 138 micrometers at the sensor
> plane. But the sensor plane is 320 micrometers high. Since 138/320 =
> 0.43, a particle of any given velocity could land anywhere in a vertical
> segment of height that is 0.43 of the screen height. So the first order
> fringes should have been very noticably widened as the fringes descend, by
> this apparently poor velocity resolution.
>
>
> 4) In the published movies of the detector plane, the intensity profiles
> of the fringes have edges that seem to rise and fall too abruptly. Also,
> the intensity profile of each fringe, especially the central fringe, in
> the movie looks flat. Fringes should have peak-like profiles. The fact
> that the peaks appear in fig 4c is irrelevant since they are a result of
> integrating offset overlapping square shaped fringes between the dashed
> yellow lines.
>
> Unless I have made several silly errors, there is something going on other
> than quantum interference. Please consider a control test to eliminate
> the possibility that you are looking at a shadow pattern that has been
> magnified by a charge deflection effect at the slits. It would be very
> easy for the slits to become charged to deflect dye particles in a manner
> similar to a cylindrical lens. A simple test would be to introduce a
> voltage control wire to the slits. An even simpler test would be to shade
> half of the slit array to see if a half side of the fringe pattern
> disappears. Whether or not a focus effect was like a positive or negative
> lens, half of the fringe pattern would disappear. A focused shadow would
> explain the anomalies I point out.
>
> Thank you for your consideration and I hope to hear from you.
> Eric S Reiter
> Unquantum Laboratory