Catch up on stories from the past week (and beyond) at the Slashdot story archive


Forgot your password?
Science Technology

Flat Lens Promises Possible Revolution In Optics ( 60

An anonymous reader shares a BBC report: A flat lens made of paint whitener on a sliver of glass could revolutionize optics, according to its U.S. inventors. Just 2mm across and finer than a human hair, the tiny device can magnify nanoscale objects and gives a sharper focus than top-end microscope lenses. It is the latest example of the power of metamaterials, whose novel properties emerge from their structure. Shapes on the surface of this lens are smaller than the wavelength of light involved: a thousandth of a millimetre. "In my opinion, this technology will be game-changing," said Federico Capasso of Harvard University, the senior author of a report on the new lens which appears in the journal Science. The lens is quite unlike the curved disks of glass familiar from cameras and binoculars. Instead, it is made of a thin layer of transparent quartz coated in millions of tiny pillars, each just tens of nanometres across and hundreds high.PetaPixel has more details.
This discussion has been archived. No new comments can be posted.

Flat Lens Promises Possible Revolution In Optics

Comments Filter:
  • by Anonymous Coward

    From TFA

    ... "Once you have the foundry - you want a 12-inch lens? Feel free, you can make a 12-inch lens. There's no limit." ...

    The talk of building 18-in (450mm) fab has been on and off for the past 10 years or so. Because of the enormous cost involved key players involved (Intel, TSMC, Samsung, et cetera) kept on putting it on the back burner

    If this lens is what it is touted to be and that one can have full maximization of the entire area of the waffle, then an 18-in lens will be so much more better than its 12-in counterparts

    In other words, another solid reason behind the construction of the world's first 18-in fab

    • There is no need though for really large lenses though, at least for imaging. You can actually cluster a bunch of smaller lenses quite easily by etching the correct geometry into each individual lenses.
      • by fyngyrz ( 762201 )

        Light gathering. Definitely a good reason.

        • by WarJolt ( 990309 )

          The diameter of the lens allows you to collect more light, so how does a tiny glitter sized lens collect enough light to be usable? I'm a little bit naive when it comes to the physics, but it seems to me that considering there's less light on the focal plane those sensors will have issues with noise.Isn't that why professional lenses haven't really changed in size decades?

  • by dgatwood ( 11270 ) on Friday June 03, 2016 @05:21PM (#52245503) Homepage Journal

    Sounds like they've taken Canon's Diffractive Optics to a new level. Basically, DO uses Fresnel lenses with smaller ribs. This raises the bar to nanoscale features, which should result in even less distortion and other problems. I, for one, welcome our new superzoom overlords. I'm envisioning a lightweight 16-600 that will outperform everything on the market today by a large margin....

    • by AK Marc ( 707885 )
      Yup. Came here to say that. Someone re-re-invented a 300 year old Fresnel lense.
      • by tlhIngan ( 30335 )

        Except it isn't a Fresnel lens. A Fresnel lens is basically slices of a lens with the bulk removed. This is good for making flat lenses but they suffer from diffraction and are rarely any good for optics as the fringing of the rings affects image quality. That's why you don't see Fresnel lenses used anywhere other than overhead projectors and stage lights where fringing is minimal. After all, if it worked well, then cameras would already be using them.

        It really is a plane of little pillars arranged in stran

        • by AK Marc ( 707885 )
          Cameras have used Fresnel lenses. Manufacturing quality was one of the biggest problems with them, but with nano, it seems they got rid of that problem, so the lense is thin, and optical quality, but still the same basic optics solved hundreds of years ago.
          • by slew ( 2918 ) on Friday June 03, 2016 @07:54PM (#52246279)

            ...but still the same basic optics solved hundreds of years ago.

            Actually, no. More details.

            The short story of what is different is described in their paper (preprint here [])...

            Although visible planar lenses can be realized by diffractive components, high NA and efficiency are not attainable because their constituent structures are of wavelength scale that precludes an accurate phase profile..... To maximize the polarization conversion efficiency, the nanofins should operate as halfwaveplates. This is achieved due to the birefringence arising from the asymmetric cross section of nanofins with appropriately designed height, width, and length

            The new idea (well not new, but meta-material approach) is that for each x,y position on the lens, a nanofin is positioned and rotated so that a localized "half-wave-plate" effect created by birefringence of the nanofin crossection modifies the phase profile of incident circularly polarized light to that which propagation through a spherical lens would have produced: All without a refractive component. A fresnel-like lens uses a small refraction (lens) element instead of a nano-sized half-wave plate to accomplish bending, but refraction requires an interface, and you can't make that interface too small without diffractive effects.

            The catch? Chromatic aberration will be much worse than a traditional lens because the phase profile is only correct for one wavelength (traditional materials also have a index of refraction that changes with wavelength, but it isn't strictly a linearly proportional geometric effect like rotating nanofins). This new technique would work fine for most scientific purposes where you have monochromatic light, but taking a full color picture with this type of lens might take quite a bit of dsp-post-processing to look reasonable.

            • by fyngyrz ( 762201 )

              RGB lenses, then? With recombination, either at the sensor or in post-processing with multiple sensors?

              We've had monochromatic sensors before and used them to make quite high quality color images.

              Most Bayer sensors are monochromatic with filters; for a true RGB sensor, meaning, a cell that responds to all three spectral peaks, Foveon [] has a three-level cell that responds to different spectra at one focal location (uses depth in the cell IIRC.) But separate photosites have been pretty much the goto tech for q

            • by daaxix ( 218354 )

              Yes, this is more correct. Essentially these new structures are modeled just like phased antenna arrays and the like for microwave rf, radio, etc., just at optical frequencies. This makes the manufacture difficult to realize because of the size of the structures required, but it is really not anything different from antenna engineering.

            • by AK Marc ( 707885 )
              The general concept is 100% the same. The physics details change slightly because of the nano scale. "flat lense" is old. "flat lense with nano" is the same thing, but different.
    • And it will be on cheap cell phones in 5 years.

      I wonder what cool applications people will come up with for it?

    • Is it still a Fresnel lens, or is it more like the diffraction pattern you get on an exposed holographic plate?

    • by Loether ( 769074 )

      I'll be excited when I can trade in my bulky, heavy, expensive SLR lenses for one perfect, cheap zoom. For all of the advances in digital sensors, lens tech has been virtually unchanged. Sure they threw on some expensive nanocoatings and made the glass aspherical but at the end of the day it's still the same heavy hunk of glass your grandparents had on there SLRs in the 50s. Never heard of the canon DO... interesting stuff. I am a Nikon shooter like my father before me.

      • by dgatwood ( 11270 )

        They're pretty neat, but Canon hasn't done a lot with them because of the moiré distortion in out-of-focus areas and lens flare problems. Supposedly, their newer designs are better, though I don't know to what degree.

    • It isn't the same as a Fresnel lens.
      The scale of these features are below the wavelength of light. That means that these structures interact with the electromagnetic field in a way that allows for a negative refractive index, which is impossible using conventional optics, rather than bending the light in steps like a Fresnel lens. This is more like an array of phased antennae, but small enough to work with light's high frequency. They have been doing similar things with microwaves for something like a decade or so.
      Google image search for microwave metamaterial. It will show you what sort of thing these pillars are mimicking, but on the order of millimeters instead of nanometers.

      • by joemck ( 809949 )

        No. As I understand it, a Fresnel lens works exactly the same as a normal lens, no smaller-than-wavelength metamaterial structures involved. You merely take lateral slices of a normal lens and collapse them to form concentric rings. A Fresnel lens does not give a better image than an equivalent conventional lens. It just reduces the thickness and mass required.

        You can use a millimeter thick Fresnel lens to do what would normally need a very thick and heavy conventional lens.

        • I didn't say that a Fresnel lens did. My meaning was that the one in the article is NOT a Fresnel lens. It is wrong to compare it to one.

  • by Gravis Zero ( 934156 ) on Friday June 03, 2016 @05:24PM (#52245527)

    one of the limitations of maskless lithography [] is getting a sharp enough lens. an advancement like this could push the limitations from hundreds of nanometers down to what's on par with modern fabrication technologies. being able to cheaply prototype chips before mass production would be a HUGE advancement for the IC development field. it also moves us closer to a tabletop computer manufacturing lab.

  • Hold on a sec.

    How is this new?!?

    Isn't this just a Fresnel lens writ "small"?

  • Monocrhomatic (Score:5, Interesting)

    by Anonymous Coward on Friday June 03, 2016 @06:20PM (#52245827)
    The downside with this type of lens is they are monochromatic. Visual light covers and entire octave (wave length doubles). It is very hard to make a meta material lens that has that big of a range. The actual science paper talks about this and describes how to make a lens in red, blue and green. These lenses need object to be illuminated by lasers. This is amazing stuff, but we are a long ways from cell phone camera lenses, which are talked about in the science "journalism" articles.
  • Telescopes!
  • That's still larger than the wavelength of visible light (about 400 to 800 nm.)

    A thousandth of a micrometre ... yes, that's smaller than visible light.

    The error is in TFS and TFA.

  • Flat Lens Promises Possible Revolution In Optics

    Anyone can promise possible anything...

    These are real, and they’re already generating exceptional results.

    Great. But no pics of these exceptional results?

    • That was my thought as well. Really nice electron micrographs of your new meta-material lens... but no images produced by the lens. Presumably because it isn't ready to actually produce a useable image yet. Because if you had a lens capable of producing better images than the best microscope lenses, you'd take a picture with the best oil immersion lens you can find and then take the same image with your lens. A nice side by side comparison where we'd all go, "Wow, that is better!!"

      So we are at the "look

  • Undetectable hi-res surveillance cameras the size of a flea just got one giant leap closer to reality.

Logic is a pretty flower that smells bad.