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

 


Forgot your password?
Close
typodupeerror
×
Science

Negative Refractivity for Optical Computing 125

Posted by michael from the making-the-impossible-possible dept.
zero_offset writes "This article in EE Times details Purdue's efforts to create a material with negative refractivity. One of the important results would be the ability to create optical computers due to the effect's tendency to amplify and focus light at wavelengths larger than the thickness of the nanowires used in the transmission system. Purdue's School of Electrical and Computer Engineering's Vladimir Shalaev says, "Using these plasmonic nanomaterials, we hope to directly manipulate light, guide it around corners with no losses and basically do all the fundamental operations we do with electronic circuits today, but with photons instead." Nanowires, surface plasmon polaritons, optical computers, nanoscale metamaterials, unnatural refractivity -- what's not to like?" We did a story on the first material known to have a negative index of refraction last year.
This discussion has been archived. No new comments can be posted.

Negative Refractivity for Optical Computing More

Negative Refractivity for Optical Computing

Comments Filter:
  • Well, there goes Moore's law out the window.

    I guess it has a bit of life left in it, but with the article talking about 'single molecule' focal points. I geuss we are about to run into a little wall if these ever actually make it to market.

    Of course, with the computational power that will come of this, maybe we will be satisfied for a while. Somebody once said "Nobody will need more than 640 k of RAM" Right?

    • "Well, there goes Moore's law out the window."

      Moore's Law [intel.com] describes an increase in transistor counts.

      I hope you are referring to the idea that traditional microprocessor design would be obsolteted by 'optical computing' thus halting the advancement of traditional microelectronics, thus stopping the advancement of transistor counts as opposed to somehow having transistors being used in 'optical computing.'

      • Moore's Law describes an increase in transistor counts.

        I hope you are referring to the idea that traditional microprocessor design would be obsolteted .....

        Well, yes.

        The role transistors play will not always be performed by transistors. Just like vacume tube technology halted in the 70's and buggy-whip disigns have been stagnate (for the most part) since the 1920's.

        But I think you missed part of my point. That being: weather of not you are talking about a transistor, once a single molecule is used as the focal point of a device, the count will not be rising much from there.

        Don't read too much into this. I am in my early 30's and for IT people my age, Moore's Law has been darned-near the only constant in our professional lives. I was just "thinking out loud" about the possibility if it now being obsolteted.

        kind of a bummer: But at least it will be around for at least a bit longer.

        • Man, that's an awful lot of typos for someone in their 30's !

          (vacuum, designs, stagnant, "whether or" "of it" obsolete)
        • "But I think you missed part of my point. That being: weather of not you are talking about a transistor, once a single molecule is used as the focal point of a device, the count will not be rising much from there."

          OK I get it ... and yes I did miss your point at first. Thank you.

    • Of course, with the computational power that will come of this, maybe we will be satisfied for a while. Somebody once said "Nobody will need more than 640 k of RAM" Right?

      I'm not saying that more power wouldn't have many uses, but it always bothers me when people quote the "640K" line about modern computers. Imagine if Bill Gates, living in three bedroom house, had said "Nobody needs more than three bedrooms." And then now, living in a forty bedroom house, he says "Nobody needs more than forty bedrooms."

      The latter, I think, rings a lot more true than the former. In most endeavors, diminishing returns can kick in after a while. It's the same reason we can't get away from the x86 architecture: There are more important issues than raw performance.
      • and not to mention that the "640k" quote often attributed to bill g is an urband legend [urbanlegends.com].

        • and not to mention that the "640k" quote often attributed to bill g is an urband legend

          yeah... sure it is.... if I had ever said something so monumentally stupid there's no way in hell that I would admit to it... you honestly think that captain doublespeak himself is any different???

      • As I see it, there are two classes of problems: 1) problems for which there's some clever, efficient algorithm (like sorting and serching), i.e. essentially easy problems and 2) problems that are not easy, where the only known option is to try all the solutions and see which work. There is a great divide between these two classes of problems.

        Computers are now fast enough to do the easy problems comfortably, even on quite large datasets. These are the ones people have in mind when they say computers are "fast enough."

        But computers are no where near fast enough to solve large problems of the second kind, and it doesn't seem that they ever will be in the forseeable future. A problem of this type might be, "what lossless compression/decompression algorithm under 1000K in length has the highest average compression on a given sample dataset?" It's easy to write a program to solve this problem, but awfully hard to wait for it to finish.

        I think we're falling into the gap between easy and hard problems. We can do easy problems easily, but have no feasible way to approach the hard problems.

  • Learnt a new word.... (Score:3, Funny)

    by marko123 ( 131635 ) on Wednesday August 28, 2002 @08:33AM (#4155736) Homepage
    I must have been out of the negative refractivity thread of modern physics, but I love this word...

    "They had free drinks that night. Trevor was absolutely PLASMONIC. I mean... shit, man! he almost had a negative refractive index. Lucky we got him in a taxi when we did"
  • Sounds like this technology could drastically improve implementing Quantum Cryptography. Imagine, long distance completely secure connections , that are provably unbreakable. Lets see Carnivore tackle that.

  • Some Claims are absurd (Score:3, Informative)

    by slashnot007 ( 576103 ) on Wednesday August 28, 2002 @08:41AM (#4155776)
    Much of what is claimed in the article and comapnion article is wrong, no doubt distorted through the prism of some "science writer" or attempt to dumb it down. For exampe, you cant focus light to a perfect point or even less that the wavelength of light.
    the ways one can escape these limits in a semantic sense is that you can change the index of refration of the media so the wavelength is shorter than in vaccum, but that's not really accomnlishing the goal. Alternatively, near field or or ther diffraction effects can confine a light field to a region smaller the wavelength, but it cant propagate in vacuum/air that way.

    likewise the claim you could make a perfecly flat focusing lens by combining poistive and negative materials is pretty hilarious too. You can do that right now with conventional positive only materials. (example take two plano confave lenses of high index material, and fill the space between them with water. voila!).

    on the other hand you could do a lot of really interesting stuff with negative index materials that is harder to put in laymans terms. one example, the speed of light might be faster than in vacuum.

    • One interesting application of this might be zero reflectivity materials. Right now the problem with (almost) all materials is that if they have any absorption at all then by defineition thay have a different index of refraction than air and consequently an inescapable reflectivity. (yes, black paint always refelcts some light!). this is bad news if you are a stealth airplane. Some ferrite materials posses an unusualy perimtivity/permiability that lets them actually have absorption yet a matched index but they are too heavy to put on an airplane. this might break open a whole new class.

      • One interesting application of this might be zero reflectivity materials

        now how would that work anyway... if you painted a basketball with "zero reflectivity" paint, you would no longer see the ball, because no light would bounce from it to your eyes, but you would also not be able to see anything behind it... so what would you see.. .or perceive??

        • The ball would simply be perfectly matte black. You would see no features on it, no shadows, no nothing. It would essentially look like a hole into black nothingness from any angle, no matter how much light was on it. You might see heat waves coming off of it, though, if you shine enough light on it.
        • no light would bounce from it to your eyes, but you would also not be able to see anything behind it... so what would you see.. .or perceive??


          It'd look like a perfect black circle was photoshopped into the world. No light comes from it: it's black, only absolute black.
        • Lack of reflectivity is not invisiblity as normally perceived by humans. like one person said, it would look like someone photoshopped a hole in a scene. But for many types of remote sensing (radar, lidar, and low light imaging) you only "see" things that reflect light. In photoshop terms this is like a very empty picture with a black background. Anything that wants to hide merely has to be black. Thus things that dont reflect are never scene in remote sensing. (not unlike the so-called "dark matter" of the universe that has gone undetected to date)
    • one example, the speed of light might be faster than in vacuum.

      No, the speed of light in the medium need only be faster than the speed of light in the surrounding medium--e.g. air.

    • What I find particularly absurd is the reliance on the existence of so-called "negative numbers." Puh-leez.
    • been reading the original negative index of refraction paper from 2000 the following preposterous claim may actually be true. 1) a lens (actually just a flat slab) of negative index of refraction n=-1 would perfectly focus an incident wave IN THE FAR FIELD more tightly than the diffraction limit of light (i.e. to much less than its wavelength). (note I said IN THE FAR FIELD, we all know about nearfield stuff) add to this the recent fabrication of negative index materials and it gets more interesting. first the easy to understand part: n=-1 slab acts a lot like a phase conjugate mirror, except instead of being a mirror it does not reflect the light but rather refocuses it downstream. THE KEY POINT: BUT as we also know they must obey a diffraction limit on how well they can refocus the beam Why cant they do better? well the diffraction limit comes from the fact that not all the fourier components are there. where did they go? well one place they went was the aperature cut off. But suppose we built the worlds largest perfect lens so that the aperature cut-off did not matter. We still could not focus light below the size of wavelength Why? Well there is one other place we lost fourier compontents. imagine the following you have a light source that is smaller than the wavelength of light (e.g. a molecule! or maybe a nearfield light source). what happens? well the small size means there are fourier terms that have k-vercots so large that for any given wavelength they cannot propagate. that is they are evenescant near field light that quickly dies away. the K-vectors that can propagate, the small ones, cant be refocused to an object as small as the source. Thus you supposedly cant ever refocus light in the farfield to smaller than a wavelength of light. But wait! those evanescent waves die exponentially so they do reach the far field, just they are really really weak. suppose you were to selectively amplify them up and then refocus the light with out phase conjugate mirror? well then you could focus light to smaller than a wavelength. Now here is where it gets REALLY SPOOKY!, if you do the math a negative one index of refraction does exactly that, it conjugates the phase and it AMPLIFIES ANY IMAGINARY K-VECTORS!!!!!!!!! that is it AMPLIFIES evenescent waves. But wait "CONSERVATION OF ENERGY YOU SCREAM !!!!". Nope! not a problem, because evanescent waves dont carry energy. "Baloney" you say. well okay, imagine an evenescent wave propagating, where does its energy go? is it absorbed by the air? put it in vacuum and it would still be evenascent. No the poynting vectors conspire to recouple the energy back to the source of the wave. THe author claims this means one can amplify these waves at no energy cost.

  • Negative R.I. for sure? (Score:1, Interesting)

    by Anonymous Coward
    Didn't we have a story [slashdot.org] about how skeptical [physicsweb.org] scientists were about these results.
  • I'm still waiting for the next big breakthrough in quantum computing, but this new optical technology will give way to some really fast stuff. Just imagine having a quantum computer cpu with optical connections to a solid-state hard drive. At least there is something to look forward to in computing nowadays.
  • Damn, that sounds cool!

    Plasmonic nanomaterials

    Plasmonic nanomaterials

    Plasmonic nanomaterials

    Now I'm sorry I went into software. I really, really wish I could tell people that I was into plasmonic nanomaterials.

  • Is it just me or does "surface plasmon polaritons" sound like somebody left the StarTrek technobabble generator on overnight?

    Actually way kuel stuff, SciAm had an article at length about this a few months back, and it was an awesome read... one of the other cool effects of this technology is optical microscopes that are orders of magnitude higher in resolution... visual systems that will let people look at things in optical frequecies that were limited to electron microscopy in the past (means we can see things way up close, that are still alive and kicking... or nanoscopic...)

    "The future's so bright you need shades..."
  • How stable is all this going to be? I have to believe this is only working in vacuum conditions at the moment. I doubt it's going to be hitting the inside of anyone's computer in the near future. As the article says, the first applications will probably be high power microscopes. Not too much else seems feasible in the near term.
  • I thought the index of refraction was defined as:

    n = (speed of light in vacuum)/(speed of light in medium),

    or n = c/cmed

    Now, convenctional wisdom and all modern science says c is always the bigger value, so n is always >= 1, but positive. How the heck does one get a negative refractivity? Niether of these quantities should be signed, let alone oppositely signed, right? What is meant by negative refractivity?

    Tim
    • It happens when the light goes backwards.... :-)

      cmed < 0
    • The use of the term "negative" regarding a material's refractive index is really semantic. Basically, as described by Snell's Law, light bends toward the surface normal as it crosses the boundary of a material. However, materials with a "negative" refractive index have the opposite effect: they cause the light to bend away from the surface normal.

    • Re:Meaning? (Score:3, Informative)

      by guybarr ( 447727 )
      I thought the index of refraction was defined as:

      n = (speed of light in vacuum)/(speed of light in medium),

      another definition, IIRC, is c/sqrt(mu*epsilon)

      mu = permeability
      epsilon = permittivity

      both are coeeficients of the linear response of meterials to the EM field.

      now, if the linear response of a material to EM fields is complex, I guess you can have negative (or imaginary) n.

      imaginary means exponential decay or growth, BTW, but of course in the case of growth the material stops responding linearly at some point, thus changing the dependance.

      IIAC, negative n does not really mean the speed of light reverses .

      Now, convenctional wisdom and all modern science says c is always the bigger value, so n is always >= 1

      AFAIK you're right in saying c is always the bigger value, however there exist superluminal photons [cerncourier.com] , which have phase velocity higher than c.

      This is not, again AFAIK, related to the response medium but to other quantum phenomenas.

      The universe can do some weird, convoluted vodoo ...
      • now, if the linear response of a material to EM fields is complex, I guess you can have negative (or imaginary) n.

        If n=c/sqrt(mu * epsilon) as you suggest, then negative n would simply result from the negative roots of the sqrt. No fancy gymnastics with complex roots necessary.

    • Re:Meaning? (Score:2, Informative)

      by wyldeling ( 471661 )

      Your definitions are correct, but incomplete. The speed of light, c, (and cmed for that matter) are defined as c^2 = e*u (electric permittivity of the substance times the magnetic permeability of the substance). In other words, the speed of light is determined by how well the substance it is travelling through can be influenced by electric and magnetic fields.


      To complete the definition:


      n^2 = c^2/cmed^2 = (e0 * u0)/(e * u).


      (The zeroes indicate that they are in free space.)


      Now to the questions: Negative refractivity should be impossible. Both e and u are positive quantities, and if they weren't, the square would make them at the minimum postive imaginary numbers. The problem is that e and u are only scalars if you are working with the prefered direction of the substance. Otherwise, they are 2nd order tensors (3x3 matrices). (e0 and u0 are always scalars.) I am not sure how this would influence the outcome w/o doing the math. But, it may allow for this type of effect.



    • FYI: The real part of the refractive index for metals is, in general, less than 1.

  • Gheesh. I thought we had enought problems recycling our old CRT's. I wonder what kind of issues these materials are gonna have if they make it mainstream?

    I wonder what country we're gonna pollute this time. Oh! Bad American!

  • Oh, goody! (Score:3, Funny)

    by KC7GR ( 473279 ) on Wednesday August 28, 2002 @09:29AM (#4156047) Homepage Journal
    Ha! New words to play with. Let's see here...

    'Surface plasmonic polaritrons...' Nah, too long. Let's condense it down to something like this...

    "Give your laundry that FRESH, SPARKLING, NEGATIVE REFRACTIVE INDEX with Maytag's NEW SURFACE PLASMONITRON!! Yes, you too can have your clothes looking like they got lost in a physics lab for a month, AND REVERSE THEIR POLARITY, all in three easy cycles!!!"

    (Read all warning labels before use. Not recommended for cashmere, poodle fur, or llama wool. Batteries most definitely NOT included, minor assembly and Ph.d required. This product is not available in Pakistan).

    Ok... who else wants to contribute? ;-)

  • It sounds cool and all, but what do you get with photon logic over electron logic?
    Is it faster? Cheaper? Less heat? More Compact? Some of these but not all of these? What do you lose by switching to photon?

    It does sound like good stuff, but what exactly is the good?

        • If you'd read the article, you'd see that there are many more applications for materials with a negative refraction index than just optical computing. The article discusses perfect lenses which would enable us to see details as small as a single molecule... but you want to know about optical computers. Light travels faster than electricity, and would encounter almost zero resistance, so I don't see anything to limit processor frequency. The heat generated by an optical circuit is negligable. Hell, you could probably put a bit of solar panneling on your computer and unplug it from the wall whenever you have the lights on. Do you lose anything? Mostly just cooking surfaces, if you've been using AMD. Oh, and static pads... they'll be useless. You might have to keep connectors cleaner. All the logic works the same... law of supply & demand: of course it'll be more expensive in the beginning, but if people want more optical than electrical, optical will get as cheap or cheaper than electrical, and optical will receive the majority of
          • Higher clock: True
            Higher switching speed: true

            But optics work on a scale much larger than electrons. Index of refraction requires a medium larger than a few molecules. Microscopic optics would run into size problems long before electronics, requiring whole conduits limited by wavelength considerations. Electronics are approaching the realm of molecular switches. 100 million transistors (typical CPU) would fill a small room with even the smallist optical transistors. Propagation speed advantage of 50% to 100% would pale next to propagation distances of 1000 times more.

            Optical will always be more expensive, you can't lay down optics with lithography or other large pattern duplication techniques. It requires every path to be a fiber.

            So optics have lots of uses, data transmission being the biggy. But logic is not likely to be one of them. I was hoping someone had a good idea of some overriding advantage, but all i get are trolls.

            Flame away. If you know anything it'll show. If you don't, that'll show too.

            • Whos flaming? I'm not. This shit fascinates me.

              Fibres can be grown in place... so where standard lithography fails, we've still got the power to make a crystalline fibre grow where and when we want. Labs are already making self-building chips. Currently, optics DO work on a larger scale than electrons. Much larger. We send billions of photons for every bit sent down a fiber line. I'm talking about optics that work on single photons.

              So even as we are developing monomolecular transistors after decades of improvement, we're still only figuring out how to build optical logic components right; I don't think the comparisson is quite fair. An electron is much more massive (on a handful of orders of magnitude, IIRC) than a photon, and requires more energy to toss around. This implies to me that there is much more room for improvement. And optical doesn't necissarily mean visible. We can (in theory) build optical circuits that work with X-rays with a wavelength under 1nm. Doing a search for molecular transistors, I find that the smallest we've made is still larger than 1nm. IIRC, the optical components have to be around 2nm for them to work on 1nm light. Double the frequency, halve the feature size. Electrons don't "shrink" like that.

              If you want a single electric-circuit-killing uber-feature of optics, you won't find one. Optics is still a very young technology. Think about your statement, "100 million transistors (typical CPU) would fill a small room with even the smallist optical transistors." I seem to recall using a room-sized computer not so many years ago. It ran on electrical transistors.
  • I figure the producers of Enterprise should be able to get five episodes worth of plot points out of the words "plasmonic" and "polariton."

    Stefan

  • Negative refraction could definitely alllow you to have at least some sort of invisibility. If not the ability to make anything disappear like with some sort of "cloaking device", you could make materials that are completely invisible. They would be constructed of this material with negative refraction along with some other material, like fused quartz or something, so that you would have a composite material with a total index of refraction of zero.

    I have thought about this subject before, and I did some research about different materials. Just think of glass. The only reason that pure glass reflects anything is that it has a different index of refraction as the air or vacuum or water that contacts it. If you notice clear ice that is in water, it is almost invisible since the index of refraction of ice is very close to that of liquid water.

    There are many applications of this type of material. Does anybody have any ideas of some applications for a zero-refraction material? Perfectly clear windows? Practical jokes?
    • I don't think that you could ever make a human invisible. Certainly not completely. Only transparent objects could possibly be made invisible by your process.

      And that will only work with some forms of radiation.

      I could imagine that you could construct something so that no photons escaped your body. Then you'd be unrecognisable but people would still know that you were there. A low-tech implementation is a stocking mask!

  • Today on slashdot, we have a nanomaterial that focuses light backwards, and also a nanomaterial that can attach to a flat, clean, dry surface well enough to support 200 lbs with a few square inches (using forces thought to only have effects at microscopic scales). The former is found only in labs and is brand new, and the latter is found in gardens and is older than humanity.

    It's sort of interesting that the article refers to the negative refraction materials as "unnatural". Nature has been doing nanotech for millions of years now. It's pretty likely that, if these materials turn out to be good for anything that occurs in nature, they can be found there.
  • plasmonic nanomaterials
    Heck, that even sounds cool.

  • What negative refraction means (Score:3, Informative)

    by zero_offset ( 200586 ) on Wednesday August 28, 2002 @01:35PM (#4158060) Homepage
    For those of you trying to figure out what "negative refraction" actually implies, the article at the URL below has a pretty easy-to-understand explanation of the key characteristics.

    03/2001 photonics.com article [photonics.com]

  • ...then it's exactly what you need in order to make this [yahoo.com] work. [Stop- HAMR time.]
  • Quoth the intro: "we hope to directly manipulate light, guide it around corners with no losses"

    Stupid scientists, always reinventing the wrong wheel. Rather than blow billion$ and years of research trying to make light turn corners, just get a fricking ruler and make those pathways straight! Do dragstrips run around in circles ? No, they're straight. Straight = fast. I don't care if my optical CPU core is 20 inches long and 2mm thick with a big protective slab of iron wrapped around it, as long as it puts out 20ghz of pixel-twaddling goodness.
  • When reading this story, I couldn't help noticing that it sounded like something out of Star Trek.

    For those who don't know why, it's because ST shows use terminology like 'plasmonic' as buzzwords.

Slashdot Top Deals

"Pok pok pok, P'kok!" -- Superchicken

Close