Work on Idea #161

An Atomic Filter by Sliding Plane

The idea here is simple. On the bottom, you have a mixture of gases, where the molecules have, in simple terms, different sizes. You are interested in obtaining the smaller molecules of a certain D1 size. 

There is a series of plates that you've prepared with holes, as best as you could in terms of the precision of the machining. These plates are stacked up, and sandwiched between two inclined planes. Both fixing points on top of the inclined plates, represented by black dots, allow them to rotate - these are the pivots. The left inclined plane cannot move horizontally in this pivot, while the right one can adjust its position. These planes can move in the horizontal direction in the bottom of this machine, just over the bottom gas tank. It is not shown in the Figure for clarity but when the left plate is completely vertical it occupy part of the container below (the inclined plain tip actually arc its way into the final position). There are many holes in a single plate but only one is shown on the central Figure.

In order to keep the plates in a really tight position I've added a water tank to the right, that pressures the right inclined plane. You handle the water volume in this reservoir by means of a faucet on top and an exit valve on the bottom, but all that it is necessary is too keep the level good enough to keep the plates in place.

When the holes are aligned, they form a clear path from the bottom tank to the top gas tank. When the plates are misaligned, however, the holes are partially covered and a smaller hole is formed, as a fraction of the original hole. Sounds simple enough. You get a small enough hole and that will filter out larger molecules. When this filter saturates, I mean the holes become covered with larger molecules, you revert some of the flow for a while and clear out the holes. 

But does it work? Can you really achieve hole sizes of a few angstroms (1A = 0.1nm = 1E-10m) with this?

Surprisingly, the calculations indicate so. But it all depends on whether you can manufacture plates with really precise heights and hole dimensions. I'll walk you through the calculations.

Define the height of all the plates as H, the height of an individual plate h and the displacement of the inclined plane from vertical position as Delta_x. The plates have holes of size b, and their size will be reduced to a size b_line. How much each hole will reduce in size is named x. Using simple triangle similarity, we can deduce the final size of the hole based on the parameters. We will consider for that only the last two plates, since to deduce it you only need two vertically neighboring holes. 

The Figure above shows the plates, first in vertical position, then displaced. The holes are the squares in the middle. On the right, there is detail of the holes. 

After deriving this equation, we went to Excel to test some values and gain some insight into its behavior. After a few rounds playing with it, we came across the values:

These values are not far-fetched. A machine as high as half a meter, with 1cm plates, totaling 50 plates, with a hole size of 1cm. The machine is displaced by half a meter. It is not even cumbersome. With this values, b_line = 0 and the hole just closed. 

Now here is the trick:

What happens if you make h = 0.009m? 

Then b_line =  1E-3m = 1mm.

What if you are able to go another notch, and make exactly h = 0.0099m?

Now, b_line = 1E-4 = 0,1mm.  

Then we go further, and make it h = 0.00999m. What happens now?

Now, b_line = 1E-5m = 10um.

You get the picture. Every time we get one precision down on another 9 on next decimal place of h, we narrow the hole by a factor of 10. You can either fix h and change Delta_X, or fix Delta_X and change h, but do note that moving precisely the entire inclined plane is a lot harder than producing a plate with better tolerances on h. 

At first, it doesn't seem like a very astonishing result. You improve your machining skills, and as a reward you can get a smaller hole dimension. One might be tempted to think, "Why not make a hole already if you have such machining skills?". But the truth is that to drill arrays of small holes can be very challenging. Our best attempt at drilling using conventional means seems to be limited to the micrometer range. Though there are equipments that can produce surface features in the nanometric range, they are highly sophisticated, requiring expensive procedures and well-skilled personnel [1].

But the way I perceive it, we are much more advanced right now in the task of producing thin-films, and really what this idea does is to trade the problem of making holes to the problem of making thin-film plates. You tie hole precision to plate height precision. 

There is a caveat in this plan: the holes must also be near perfect. Imperfections in the borders of a hole could turn a nanometric hole into a micrometric one very easily. I'm not the best person to answer if we have a method, right now, capable of machining an array of large (1cm) holes to nanometric perfection, but I do now that at the very least gauge blocks are machined to the micron precision, what leads me to believe that we have the technology, right now, for tolerances up to 1E-6m = 1um. 

Maybe we could, at great cost, produce the plates with the right material using one of the advanced nanotechnology techniques we have today and end up with a filter good for many runs. It is way out of my league, hence why I share this simplistic idea. 

The principle for this filter shouldn't be just for gaseous atoms, but I suppose it could be used in a liquid media, for desalinization purposes, water treatment, filtering out bacteria, sorting proteins, these are just some ideas.

By the way, sorting or rather sieving is achieved in the following method:

You build three machines, tuned to three sizes of hole D1, D2 and D3

You have a gas sample made of molecules of diameters D3 > D2 > D1.

First, you pass the gas sample on the machine D3, and it filters all molecules with D3. Only molecules with diameters D2 and D1 pass through. Your result is a mixture of molecules of diameter D2 and D1

Then, you pass the sample with molecules of diameter D2 and D1 to the machine D2, filtering out the molecules D2. Your result are the gaseous molecules with diameter D1

Doing so, you've separated the gases D1, D2 and D3

Will it ever be possible to just take out, effortlessly, the noble gases from the atmosphere and profit?


[1] Albrecht, T. R., Dovek, M. M., Kirk, M. D., Lang, C. A., Quate, C. F., & Smith, D. P. E. (1989). Nanometer‐scale hole formation on graphite using a scanning tunneling microscope. Applied Physics Letters, 55(17), 1727-1729.  

BANNER IMAGE CREDITS: ESA/Hubble & NASA, A. Filippenko, R. Jansen 

Want to know more about this image? Follow this external link.