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

 



Forgot your password?
typodupeerror
Check out the new SourceForge HTML5 internet speed test! No Flash necessary and runs on all devices. ×

Silicon Superconductors 141

Diana writes "Physicists at CNRS have demonstrated superconductivity in silicon, the element long known for its semiconducting properties. High doping is the key — by substituting 9% of the silicon atoms with boron atoms, it was found that the resistance of the material drops sharply when cooled below 0.35 K. A small increase in the transition temperature is likely with further work."
This discussion has been archived. No new comments can be posted.

Silicon Superconductors

Comments Filter:
  • by multiplexo ( 27356 ) on Wednesday November 22, 2006 @11:08PM (#16960918) Journal
    But you can make pretty much anything superconductive if you get it down below .5 Kelvin. I mean really, go much lower and you can make Twinkies superconductive much less boron doped silicon.
  • So..... (Score:4, Funny)

    by PHAEDRU5 ( 213667 ) <instascreed@gmai[ ]om ['l.c' in gap]> on Wednesday November 22, 2006 @11:08PM (#16960922) Homepage
    Superconductivity in non-superconductive materials, except where they've been doped to be superconductive.

    Makes me want to get back to the pub.
    • by substituting 9% of the silicon atoms with boron atoms

      That makes me wonder if it is still legitimately considered silicon. I mean, replacing nearly 10% of it with another element must mean that it falls into another classification. I don't think it could be considered a compound since the atoms are not bonding in the traditional sense, they are simply occupying places in a crystalline structure. Perhaps it is more appropriate to call it a "silicon-based material"

      • Re:So..... (Score:5, Informative)

        by Dunbal ( 464142 ) on Wednesday November 22, 2006 @11:49PM (#16961220)
        replacing nearly 10% of it with another element must mean that it falls into another classification.

              An alloy, if you will?
        • No, not really an alloy, since at least one of the constituents of an alloy needs to be a metal. Silicon and boron are both metalloids, and are therefore weak conductors. The change in conductivity is created by the effect the doping has on silicon's conduction band.
      • Re: (Score:2, Informative)

        The doping is a small enough percentage that it still retains the crystal structure of silicon, so on a macroscopic level, it still looks and behaves like silicon. The crystal structure forces the boron to act like silicon, which is key because boron has two fewer valence electrons than silicon, resulting in stable "electron holes" in the conduction band that raise the overall mobility of electrons, and therefore raise the conductivity of the material.

        So no, it's not pure elemental silicon, but it's still
        • So no, it's not pure elemental silicon, but it's still silicon. It's like saying that even if my tap water contains 10% impurities, it's still water.


          Bullshit. Dump 10% Kool-aid powder in there and get Kool-aid. Stick a teabag in there and get tea. Run it through some beans and get coffee.
    • Re: (Score:2, Funny)

      by gt_mattex ( 1016103 )

      Superconductivity in non-superconductive materials, except where they've been doped to be superconductive.

      Makes me want to get back to the pub.

      After your good and 'doped' up do we throw you in the freezer and run a current through you?

  • How useful is this? (Score:3, Interesting)

    by TubeSteak ( 669689 ) on Wednesday November 22, 2006 @11:08PM (#16960928) Journal
    0.35 K? as in... barely above absolute zero?

    Etienne says that they will probably be able to increase the transition temperature a bit further, although the material will be unlikely to have any applications in consumer devices.
    What non-consumer applications will it have? Getting something down to .35K isn't exactly trivial...

    IIRC, anything that doesn't superconduct at the temp of liquid nitrogen is a pain in the ass to use.
    • by swordgeek ( 112599 ) on Wednesday November 22, 2006 @11:45PM (#16961186) Journal
      First of all, it's called RESEARCH! It's a very new and different bit of science--who knows where it could lead us?

      Secondly, just because things are a pain in the ass doesn't mean they don't have useful applications. NMR/MRI have been dependent on low-temp superconductors (i.e. liquid He or even colder) for decades, and they're immensely important for research and medicine.
    • by wass ( 72082 )
      It's more of interest for physics-level study of electron interactions. This announcement would be one of many similar research papers to hit the physics journals each week, nearly all of which don't have immediate applications in mind. But that's the point of physics, to study what the hell happens in systems, and to explain why we see what we think is weird behavior, or to try to predict other effects.

      I'm sure most of the people studying electron band structure of p and n doped silicon could never have
    • by Fred_A ( 10934 ) <fred@freds h o m e.org> on Thursday November 23, 2006 @02:52AM (#16962140) Homepage
      What non-consumer applications will it have? Getting something down to .35K isn't exactly trivial...
      Agreed, I know I definitely am going to need bigger fans to get to that level of cooling in my machine. Much bigger fans.
    • by ccp ( 127147 )
      Getting something down to .35K isn't exactly trivial...

      In space it is.

      Cheers,
      CC
  • whatever that means!
    • by NanoProf ( 245372 ) on Thursday November 23, 2006 @12:43AM (#16961506)
      It means quite a bit. Strontium titanate was the first superconducting semiconductor, predicted to be so by Marvin Cohen (my theses advisor :-) and later confirmed by experiment. The general idea is that a semiconductor with multiple valleys in the conduction band into which to place dopant electrons can rapidly develop low-energy electronic states under doping, and these are the states that couple to lattice vibrations and so generate superconductivity. If you don't have a problem with the term "doped semiconductor," (which is a material that actually conducts- how do you think those electrons get through transistors on computer chips :-), then you should be ok with "superconducting semiconductor".
    • Silicon is a semiconductor (at room temperature, undoped). If doped and at 0.35K, it becomes superconductive. Saying "superconductive semiconductor" is quite misleading, like saying, "liquid ice". It's the same material, but it has different properties at different Temperatures.
    • The best name for this would actually be "semisuperconductor". Since a semiconductor is sometimes an insulator and sometimes a conductor, it follows that something that is sometimes a superconductor would be called a semisuperconductor.
  • by ryanisflyboy ( 202507 ) on Wednesday November 22, 2006 @11:22PM (#16961044) Homepage Journal
    And just to make the article more clear: Let's substitute "boron" with Tom (hey, what guy wouldn't want more boron?), and "silicon" with Suzie (hey, what girl woudln't want more, eh, yeah.).

    "Because it has one fewer electron than Suzie available for bonding with neighbouring atoms, Tom incorporated into Suzie leaves a positively-charged "hole" at each site where Tom's "missing" electron would be paired with one of Suzie's."

    Well they did do it in France, you know.
  • by physicsphairy ( 720718 ) on Wednesday November 22, 2006 @11:29PM (#16961084) Homepage
    If we can turn semiconductors into superconductors, then we can probably turn my band conductor into a semi-conductor, which would at the very least mean less thrown chairs during parent teacher conferences, and less thrown chairs can only be good for Linux!

    (Yes, that happened; and yes, he is still in band director.)
  • Boron (Score:5, Insightful)

    by sankyuu ( 847178 ) on Wednesday November 22, 2006 @11:33PM (#16961100) Journal
    Gah!
    How is this useful? What are applications? Blahblah

    Since when did science have to have applications?
    (This isn't sarcasm; science is about discovery. Applications of those discoveries are mostly accident. You can't automatically "succeed" at science. Failing to find a room-temperature superconductor isn't failing per se; it means succeeding to eliminate another coulda been material. Finding dead ends is part of the quest. And this result might not yet be a dead end.)

    So far, most of the comments have been posted by boring morons.

    -A bored moron
    • Re: (Score:3, Informative)

      by espressojim ( 224775 )
      Eh, just because you haven't found a material that is a superconductor at room temperature doesn't mean that there aren't any. It's easy to say "X can happen, because we have example Y", but you can't say "X can't happen, because we have example Y". All you can do is state all the places you've seen that it doesn't work. Sometimes, you can generalize those results (water from the atlantic ocean is not made of cheese, and thus we have no reason to believe that water from the pacific is either.)

      Negative re
      • by Sique ( 173459 )
        This is called the inductive principle (we assume, that the future is not much different from the past), but in fact it has not much value in science as such. It is one of the possible hypothesis generators, but it is neither the only one nor does it warrant a "better" hypothesis than others.
        Each hypothesis, independent from it's origin, still has to undergo evaluation, until it finally dies in an experiment which proves, that the hypothesis is wrong, or it goes thrown out because it is only a restatement o
        • I wouldn't ever say that an alternative hypothesis is 'better' than the original one you're testing without evidence. However, when you say A->B, and you can disprove a bunch of alternatives, it makes the original more attractive, because you've shown that the alternatives are unlikely.

          For backround: I'm in the bioinformatics field. One of the last papers we published was on the effects of selection on conserved non-coding sequences. There were a number of hypothesis for the effect we were observing w
          • by Sique ( 173459 )
            But you describe the Homesian deduction principle: "How often have I said to you that when you have eliminated the impossible, whatever remains, however improbable, must be the truth?" (Arthur Conan Doyle, The Sign Of Four), where the hypotheses, that remain after a bunch of them are eliminated due to contradictions with observations, are the ones that are considered correct until more evidence shows up. That's quite different from induction, where you collect evidence until you spot a pattern and then use
            • But you describe the Homesian deduction principle: "How often have I said to you that when you have eliminated the impossible, whatever remains, however improbable, must be the truth?" (Arthur Conan Doyle, The Sign Of Four), where the hypotheses, that remain after a bunch of them are eliminated due to contradictions with observations, are the ones that are considered correct until more evidence shows up. That's quite different from induction, where you collect evidence until you spot a pattern and then use

    • I've been watching too much of the original Battlestar Galactica. With their "centons", "sectons", "furlons", "crawlons", and of course "Cylons", when I saw the term boron, my first thought was that it was some sort of unit of boredom. Then I read the article, and realized I was right.
  • by LM741N ( 258038 ) on Wednesday November 22, 2006 @11:47PM (#16961206)
    Boron was made superconducting by doping it with 90% silicon.
  • 0.35K is rather cold (Score:5, Informative)

    by NixieBunny ( 859050 ) on Wednesday November 22, 2006 @11:49PM (#16961214) Homepage
    I realize that this is just a laboratory curiosity at this point, and no one would try to use such a compound commercially. Still, a brief description of what it's like to make 0.35K is in order.

    I work on a radiotelescope that uses receivers cooled to 4K. These use a helium refrigerator that works just like the Freon thing in your car but using helium instead of Freon as the phase-change medium. It takes three stages of cooling (with compressors and heat exchangers) to get to the 4K point. It also takes 10 kW of electrical power to cool one watt of load to 4K.

    We until recently had one receiver, a bolometer, that was cooled to 0.4K using the 3He isotope of helium that has a lower boiling point. The refrigerator for this is a fist-sized gadget that uses a charcoal trap, a heater resistor and some plumbing to make a refrigerator that can be cycled to produce 0.4K for a day or so at a time. It makes many microwatts of 0.4K coldness from less than one watt of 4K coldness.

    Unfortunately, the 3He leaked out and the gizmo is currently a paperweight since it was made by a very clever French guy who's no longer in the business.

    You can still buy 3He refrigerators from other manufacturers, but they are two feet long. The 3He is available for several thousand dollars a bottle.

    • by wass ( 72082 ) on Thursday November 23, 2006 @02:01AM (#16961896)
      Getting to 4K is relatively easy, you get a dewar of helium (this is the relatively abundant He4 isotope) at roughly $4 per liter. You can cool to 1K relatively easily too by pumping on the vapor over the helium, evaporatively cooling ot down to 1K. It's inefficient to do this, though, people tend to build a 1K pot into their cryostat to only pump on a small volume of helium to cool their system to 1K, not the whole dewar.

      Regarding the Helium 3 Fridge, that's actually doing the EXACT same thing as the 1K pot above, you're evaporatively pumping He3 with the charcoal sorb. Since He3 is rare and expensive, this is done in a closed system and recycled.

      I know your pain, though, our He3 fridge has a leak, luckily not on the He3 system (He3 is super expensive), and it's been a pain in the ass to try to fix. To fix your system, you probably don't need that French dude to fix it, get a leak checker (find some experimental condensed matter guys that do vacuum sputtering or evaporation work, they'll have a leak checker), track down the leak on your He3 system, plug the leak (silver solder if possible w/ your machine shop), then pay some $$$ to inject some He3 back in when you're damn sure you've got no more leaks left.
    • Janis Research makes a closed-cycle adiabatic demagnetization refrigerator system [janis.com] that can cool things down to 0.1 K, requires no maintenance, and is tabletop-sized (well, at the limit of the definition: the two boxes, wessel + compressor, weight over 200 kg when added together).

      Their site seems to be down right now, but it must be a temporary glitch.
  • by ebers ( 816511 ) on Wednesday November 22, 2006 @11:52PM (#16961230)
    Yes, 0.35 K is really cold. Refridgeration methods that reach this temperature cost ~ $100,000 and use the helium-3 isotope as the working fluid, which costs several hundred dollars per gaseous liter at STP. But this may still be useful because there is lots of established technology for making very small things out of silicon, and lots of fundemental physics that can only be done at very small length scales and in very cold environments.
    • by joto ( 134244 )
      Thank you. You are the first to give a plausible explanation as to why anyone would be interested in another superconductor, when it needs such a low temperature. However, if people are interested in doing low-temperature nanoscale physics experiments, I'm all for that!
      • I've got another one, related to the first: Super-conducting sensors are several orders of magnitude more responsive than normal sensors. The problem is to pass that data to the non-superconducting circuits that record the data, and in designing those sensors.

        We already have designs for sensors made out of silicon, and I bet these superconductors integrate fairly well with normal computer components...
  • Doping (Score:1, Redundant)

    This goes to show that even silicon when doped get high!

    --
    Now, that's a sig!
  • by ElephanTS ( 624421 ) on Thursday November 23, 2006 @12:05AM (#16961300)
    I tried super-doping myself but it got boron after a while.

  • A small increase in transition temperature was the last thing we needed at this point. What little luck we had left was spent during the gravitational slingshot and none of us needed a glance at the holo-panel to know the ion shields had only minutes before a collapsed inner hull went from worry to worse.

    Without saying it aloud, we all knew the survival of the ship...our survival...was totally dependent on staying out of sensor range for just a bit longer. The sub-orbital alerting buoys, with their grid-
  • substituting 9% of the silicon atoms with boron atoms, ...
    That means doping boron, not silicon.
  • So I guess Pamela Anderson is more useful than we thought! Other than the obvious, of course. ;-)
  • Things (Score:3, Funny)

    by mqduck ( 232646 ) <mqduck@mqdu[ ]net ['ck.' in gap]> on Thursday November 23, 2006 @04:51AM (#16962596)
    I can do things I normally can't when I'm doped, too.
  • Damn, I guess this water cooling system isn't going to cut it for my next upgrade.

  • Some years ago, evidence for superconductivity was found in a gallium arsenide epitaxial device. The work was duly published, and only some time later was it realised that the superconductivity was occurring in metallic indium on the back of the device - the indium had been used as a good thermal conductor for mounting the GaAs substrate in the epitaxial growth chamber, and had not been completely removed.

    If these guys have done their work carefully, they will have gone to great lengths to ensure that they
  • We've been able to create chips with Niobium implanted in silicon for quite a while, and Niobium superconducts at similar temperatures

A budget is just a method of worrying before you spend money, as well as afterward.

Working...