Optical Antenna Distortion Free Flat Lenses

[wuliheron]

http://www.sciencedaily.com/releases/20 ... 093523.htm

This the stuff of far out, freaky, and funky science fiction. A flat perfectly distortion free optical lens only 60nm thick. If I understand the technology correctly it uses surface plasmons to provide the same phase delay thick glass might in an ordinary lens and is made using standard semiconducting fabrication technology. That's the only reason I post it here is because from what little I can tell the technology looks too good to be true and could easily be mass produced within a few years. There are obvious advantages a few years down the road to using them in an Oculus Rift, but the long term potential uses for plasmonics like this go straight down the rabbit hole into LaLa land. The ability to squeeze light down much smaller than traditional optics allow means in several decades the Oculus Rift might have little resemblance to the current model.

[brantlew]

That's really incredible!

[Chriky]

Unfortunately it seems like it doesn't work with visible light at the moment, however its not too far off (near infrared)

---

EDIT; Actually, scrap that, I think they mean it goes from almost infrared all the way the other way. It's a bit confusing because they cite Hz value for wavelength instead of frequency.

[cybereality]

This could be really great for making small form-factor high-FOV HMDs.

Wonder how many years we will wait for this to reach the consumer market.

[wuliheron]

Unfortunately it seems like it doesn't work with visible light at the moment, however its not too far off (near infrared)

---

EDIT; Actually, scrap that, I think they mean it goes from almost infrared all the way the other way. It's a bit confusing because they cite Hz value for wavelength instead of frequency.

No, you are correct this first lens does not work in the visible spectrum. It's more in the range of far microwave to near infrared. However, it's inventors say it should be possible to apply the technology to almost any optical system and mention microscopes in particular. Theoretical studies have shown a broadband version of a Luneburg lens spanning the visible light spectrum is entirely possible and people have been racing to make it happen. This kind of "fisheye" glass marble-like lens can focus light from all directions equally well and a superlens version would have no distortion whatsoever. Perfect for a microscope and, I assume, something like the Oculus Rift.

[wuliheron]

This could be really great for making small form-factor high-FOV HMDs.

Wonder how many years we will wait for this to reach the consumer market.

This particular version of the lens is being developed for fiber optic use, and a broadband version could take some time. However, the fact that the fabrication process is so easy suggests we won't have to wait long before they develop one for the entire visible spectrum and then mass produce them. If we're lucky my guess is about five years.

P.S.- After reading over this I realized I was a bit cryptic. Scientists have been struggling to create a superlens for the visible spectrum for over a decade now only to run into endless engineering problems with every design they've come up with. They can make superlenses that can have limited bandwidth, but making them complex enough to do the entire visible spectrum has proven frustrating. This is the first really simple and completely flat design anyone has come up with which means they can easily add as many layers as they want to give it more bandwidth. Not to mention this first design has enormous applications in fiber optics including possibly making it cheap and easy to finally connect fiber optics to your home computer and your Oculus Rift providing all the bandwidth you might need for full HD or better.

[wuliheron]

Just a quick update, but according to Wired magazine the website of the creators of the new lens received so many hits it almost went down. They also confirm that the design should make it possible to create a lens covering the entire visible spectrum hopefully within a few years.

http://www.wired.com/rawfile/2012/08/ne ... know-them/

It's essentially a new metamaterial, but one of the simplest-yet-amazingly-useful ones devised to date. There is also a new optical microprocessor with a similar design of 3D gold V shaped antennas that allow light to be manipulated on the chip itself and my first thought in seeing this flat lens was it must be based on the same research.

For those who don't know what plasmons are, they are a kind of quasi-particle created by photons. Without going into the quantum mechanics of quasi-particles you can think of them as light being converted into oscillating electron densities on surfaces like metals. That way you get much smaller than the wavelength of the light, and then convert the plasmons back into light because electrons both absorb and emit light.

[android78]

This is fascinating. So often we forget that light is really just very high frequency electromagnetic waves, and have all the same properties. I assume the current IR end of the spectrum is because they have much longer wavelengths, so the nano antenna can be larger and it's just a case of miniaturization further to create for other wavelengths, but i wonder how this would work with multiple wavelengths. For the visible spectrum, may need to have a composite lens where each segment would be for a different wavelength.

[zalo]

I wonder if they did the graduated steps on purpose. I doubt it will be much of a problem but I wonder what manufacturing limitation forced them to striate it like this (as opposed to a smooth gradient) for one of their first prototypes.

Also, I think a uni-directional lens would be really cool that just shifted all the light in one direction.

[wuliheron]

I wonder if they did the graduated steps on purpose. I doubt it will be much of a problem but I wonder what manufacturing limitation forced them to striate it like this (as opposed to a smooth gradient) for one of their first prototypes.

Also, I think a uni-directional lens would be really cool that just shifted all the light in one direction.

I believe it's a Luneburg lens and theoretically you can't get the same property of it being completely distortion free using any other design. They may have to play around with different arrangements of these nano antennas to find the best compromises for other types of lenses, but this is the ideal arrangement that gives them benchmarks for how good future lenses can be and what the tradeoffs might be. I don't know if a unidirectional lens is possible using a 2D design, but I think there are 3D ones being investigated for invisibility cloaking.

[brantlew]

This is wild speculation because I don't understand the technology fully - but I wonder if this technique could be done dynamically. In other words, I wonder if either you could create a similar nano-structure that could be electrically activated either continuously "tuning" the atomic surface to a desired refraction level or maybe just having a dense interspersed pattern of structures all tuned to different levels that could be selectively turned on or off. The end result being that you could have a flat lens that could be instantly refocused electrically. With pupil tracking you could then dynamically adjust the focus to compensate for the cornea and have a variable focusing HMD.

[android78]

I wonder if they did the graduated steps on purpose. I doubt it will be much of a problem but I wonder what manufacturing limitation forced them to striate it like this (as opposed to a smooth gradient) for one of their first prototypes.

Also, I think a uni-directional lens would be really cool that just shifted all the light in one direction.

It may be due to the fact that different antenna lengths will have different absorption/transmission properties. For instance, a quarter wave antenna will have similar proprieties to 1/2 and 3/4 antenna, but will be totally different to 1/3 or 2/3. So to do a 1/4, 1/2, 3/2 or full wave delay, then it would be matching. It looks from the diagram that they are using 8 different distinct phase shifts, which may be related to re-polarizing, to get an additional shift in the phase.

[wuliheron]

This is wild speculation because I don't understand the technology fully - but I wonder if this technique could be done dynamically. In other words, I wonder if either you could create a similar nano-structure that could be electrically activated either continuously "tuning" the atomic surface to a desired refraction level or maybe just having a dense interspersed pattern of structures all tuned to different levels that could be selectively turned on or off. The end result being that you could have a flat lens that could be instantly refocused electrically. With pupil tracking you could then dynamically adjust the focus to compensate for the cornea and have a variable focusing HMD.

There are already tunable metamaterials in existence and it isn't nearly as far out on a limb as you might think. Exactly how that might be done in this case is anyone's guess, but certainly there is no theoretical reason it can't be done. MEMs or microelectromechanical systems are a distinct possibility that come to mind. Instead of changing the material you might simply move the gold antennas around on the lens to change their overall optical properties. What is even more exciting in the long run is the possibility of combing such lenses with plasmonic circuitry.

This is going down the rabbit hole into LaLa land like I mentioned in the original post, but imagine something as thin as a piece of paper covered with "lenses" almost as small as photons. That may sound strange to some, but visible light is actually quite large compared to things like the individual features on modern computer chips. Each "lens" could convert individual photons into smaller plasmons that could be manipulated by computer circuitry inside the paper. That also might sound strange to some, but computer chips are typically millimeters thick. The computer circuitry could then collate and manipulate the photonic information the plasmons represent in countless ways. If you want, the other side of the paper might even be able to spit out the finished image as photons again. At any rate, this "paper" could contain thousands or millions or billions of cameras that could cover an entire wall or whatever and would be capable of reproducing virtually any image your little heart desires from the photons it absorbs.

It would essentially make almost all glass optics worthless for any kind of serious camera, telescope, or microscope because instead of merely manipulating the light you are actually capturing all the information the light contains using a cheap and easily mass produced piece of "paper" that can cover anything from an Oculus device to an 8 story building and manipulating the information it represents in an almost infinite number of ways. Need a telephoto lens? Simply zoom in on the image wherever you want because all the information is there already. Need a microscope? Simply zoom in on the image because all the information is already there. Need a false color image or 3D, no problem because all the information is there and you have the ability to manipulate it in an almost infinite number of ways. That's the ideal and we'll see a lot of intermediate steps taken in that direction in the next couple of decades, but these easily mass produced lenses are an exciting first step.