Follow Slashdot blog updates by subscribing to our blog RSS feed

 



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

A Greener Chip Manufacturing Process 68

gardenermike writes "A new chip manufacturing process has been developed which uses UV light instead of high temperatures to prepare the silicon. This could lead to cheaper chips and greener factories if it pans out. Apparently the main problem is defects in the material, which are currently 'ironed out' as a side-effect of the extreme temperatures used."
This discussion has been archived. No new comments can be posted.

A Greener Chip Manufacturing Process

Comments Filter:
  • as the defect ratio issue would be negated by the size of the technology being produced.

    for custom fabricating nanocircuitry for nanomachines this seems like an awesome idea.

    so how long till we have a swarm of replicators based on this technology :)
    • so how long till we have a swarm of replicators based on this technology :)

      Well, first this technology would have to produce Reese [wikipedia.org] so we won't have to worry too much...

      In reality, though, Replicator-like situations [wikipedia.org] are infeasible on planets which already have lifeforms(like humans or Asgard). That isn't to say, though, that nanotechnology wouldn't be a very deadly weapon, for example nanobots designed to destroy one's nervous system.

      • Replicator scenarios aren't nearly as implausable as you make them sound. :) in fact, they're a near certainty when the right people with the wrong mindset get access to self reproducing 'viral' technolgies such as home brew nanotech.

        still, I wouldn't worry too much, as self reproducing nanotech still wouldn't realistically be able to wipe out organic life. they could easily consume all the technological resources of a world, but the cost of wiping out organic life would likely be a dead giveaway that eith
        • The scenario itself is possible, but having an effect beyond Peak Everything(which is still quite bad, mind you) would be quite hard, as it would be equivalent to competing with thousands of biological versions of 'goo' that millions if not billions of years more experience.
          • would you believe 100 trillion years? :) i was the last man in the universe, about 80 times.

            and i spent 28 GAME YEARS forging a chaos blade :) i had the entire end of the universe compresed into a single month :) and i know of the virus and the antidote, and the anti antidote :) and have compressed data from the year 100,000,000,000 AD :)
  • You know... chips are supposed to be yellow, regardless of whether they are of the corn or potato variety.
  • And here I was thinking we already had MOS Semiconductor...
  • Side-effect my @$$ (Score:5, Insightful)

    by Nom du Keyboard ( 633989 ) on Thursday June 29, 2006 @07:31PM (#15632697)
    Apparently the main problem is defects in the material, which are currently 'ironed out' as a side-effect of the extreme temperatures used.

    I'd hardly call it a side-effect to have a process that minimizes defects. I'd rather call that an essential-effect.

    • Yes, precisely. They're most likely referring to a process known as "annealing," and its purpose is to repair damage to the silicon lattice that's caused as a normal course of semiconductor manufacturing. As you helped point out, it wouldn't be done if it wasn't essential.
      • Professor Boyd admits that prolonged exposure to other UV wavelengths produces defects, but points out that his technique uses a wavelength of light that has never been used before.
        Mhmmm... a different wavelength.

        Realistically, will that make any difference when it comes to solving the defect situation?

        If it's just a matter of hitting the right frequency, why not get a stack of different UV tubes and test 'em one by one in a scientific manner?
    • >> Apparently the main problem is defects in the material, which are currently 'ironed
      >>out' as a side-effect of the extreme temperatures used.

      >I'd hardly call it a side-effect to have a process that minimizes defects. I'd rather
      > call that an essential-effect.

      Actually, the truth of the matter is that semiconductors NEED defects. Implanting phosphorus atoms (or whatever dopant you choose to use) induces defects - the atoms are injected into the silicon wafers at very high energies,
      • Actually, the truth of the matter is that semiconductors NEED defects. Implanting phosphorus atoms (or whatever dopant you choose to use) induces defects - the atoms are injected into the silicon wafers at very high energies, battering their way into the silicon lattices. These are defects, but they're essential to semiconductor operation - without those defects, there would be no conduction at all (silicon by itself doesn't conduct electricity).

        Typcially dopants are called "impurities," not defects.
      • we need to distinguish between lattice defects and chemical impurities.

        the former is a problem the latter is essential. Unfotunately the most controlled processes for adding impurities also give a lot of lattice defects needing an annealing step to repair the lattice. Annealing however has problems of its own (lots of heat needed and it tends to cause diffusion between regions).
  • by erice ( 13380 ) on Thursday June 29, 2006 @07:36PM (#15632729) Homepage
    Silicon dioxide is the all purpose dielectric in most current chips. It is slowly and painfully being replaced by "low K" materials between wires and "high K" materials under the gate electrodes. The transition to Low K/High K has been pushed out again and again but it is being used in some chips now. If the this new method of growing silicon dioxide is still in research, it seems doomed to reach production shortly after it is no longer needed.
    • by treeves ( 963993 ) on Thursday June 29, 2006 @07:58PM (#15632864) Homepage Journal
      Not so fast.
      SiO2 will still be used in non-critical layers and in less-than-leading-edge technology, which there is a lot of, and will be for a long time to come. Not all chips are CPUs. In fact, most aren't. It's worth a look.
      TFA also said it might allow manufacturing semiconductors on substrates (other than Si) which heretofore wouldn't be possible due to their inability to withstand the high temperatures.
    • The inter-metal-layer oxide that is begin replaced by low-K dielectrics is/was not formed by oxidation but by chemical vapour deposition. This oxide process is targetted at gate or isolation dielectrics. High-K dielectrics are in the roadmap to replace oxide at the gate dielectric, but development is much slower than people thought 5-6 years ago. In fact, it may never happen before we switch to different materials altogether.
  • by tacarat ( 696339 ) on Thursday June 29, 2006 @07:37PM (#15632736) Journal
    Would this process also be useful for making silicon based solar cells? Or is it at a step of silicon processing that's too far towards chip specific manufacturing? If solar cells can be made more cheaply, I wonder what this could make the initial $/watt investment.
    • If solar cells can be made more cheaply, I wonder what this could make the initial $/watt investment.
      You can forget about silicon and use a variety of other materials with the sol-gel process - where manufacturing is just optimising what could be done with a bucket of chemicals to dip yoyur substrate into and a domestic oven.
      • The best hit google gave me on sol-gel solar power was an article from '02 [solgel.com]. At least in this case, isn't this something to use to increase the solar conversion efficiency rather than making something into a solar cell? Do you recommend any links for solgel based solar power generation articles?
        • It's the name for a fabrication technique which among many other things can be used to make multicyrstalline photovoltaic cells - it's really about coating materials with properties you want onto surfaces and doesn't have anything to do with the sun despite the name. The best links I can think of are www.solgel.com and www.isgs.org.

          The reason I mentioned this is because zone refining of silicon to ultimately make large diameter single crystal wafers is an expensive and highly energy intensive process and i

    • Would this process also be useful for making silicon based solar cells?

      Doubtful. Silicon solar cells generally don't use any silicon dioxide layers, so a low-temperature method for forming silicon dioxide isn't much help. Not to mention that most of the cost of a solar cell is in the silicon wafer itself, not anything you do to the silicon after wafer production.

      • Awesome. Just the answer I was looking for.
      • The cost has more than doubled in the last year and a half. Many of the big players (Hemlock Semiconductor in Michigan is the world's biggest) are planning huge expansions, but it will take a few years. Right now they are all making a killing, and most of them are completely sold out for the next year. This is due to both increased demand (and waste) from the IC industry as it switches to 300mm, and because of a massive increase for solar applications.
  • Green Chips (Score:4, Funny)

    by Joebert ( 946227 ) on Thursday June 29, 2006 @07:44PM (#15632763) Homepage
    Good thing this is Slashdot, I usually throw the green chips away.
    • Green chips are supposed to be eaten ala "Soylent Green [imdb.com] .

      Falcon
      • I tried some soymilk once, I learned to stay away from anything that starts with "soy".
        • I used to drink soymilk as well but stopped because it's high in carbohydrates. Tell the truth I prefer real milk anyway.

          Falcon
          • I dunno if I still am, but I was told I was allergic to soy as a kid. There's other compelling reasons not to consume it if you're male, the plant hormones are effectively estrogen. Anyway, have you tried almond milk? As long as it's cold, it's really quite good. If it's warm, it tastes horrible, so drink it cold. I put it on granola, which is really tasty, even if it proves that I'm some kind of post-hippie.
            • If it's warm, it tastes horrible, so drink it cold.

              Yes, I rather liked almond milk, then again I like almonds though I prefer amaretto and pistachios. However I don't, er didn't drink it plain, I used it when making espresso.

              Falcon
              • Yeah, that's the exception to the temperature thing, it makes coffee beverages creamy and nutty. Delicious. I still prefer half and half but I've all but stopped snoring since I all but gave up dairy, so to hell with milk and the like. I've even been eating cashew-based "ice cream" and while it's nowhere near as satisfying compared to ice cream as almond milk is compared to milk, it's still better than having my sinuses filled with mucus.
          • It's beans. SoyMilk Green is made out of beans. They're making our food out of beans. Next thing they'll be breeding them like cattle for food. You've gotta tell them. You've gotta tell them!

            OH MY GOD!!! SoyMilk Green is made of .... BEANS!!!?! IT'S MADE OF BEANS!

            Alright, so it just doesn't have the same impact as "people!", does it?
  • by Pink Tinkletini ( 978889 ) on Thursday June 29, 2006 @07:56PM (#15632847) Homepage
    I posted this to an earlier discussion, where it seems to be eliciting no replies, so I'll ask again here. The Wikipedia entry [wikipedia.org] states: "The successors to 45nm technology will be 32 nm, 22 nm, and then 16 nm technology; it is possible that these numbers are arbitrary, but it is also possible that they reflect fundamental physical limits of some sort." So which is it, arbitrary or fundamental physical limits?
    • They're not exactly arbitrary, but they're not physically imposed (like quantum rules or something) either. They're basically just more-or-less a constant ratio from one down to the next.
      The semiconductor companies get together and publish a roadmap called ITRS [itrs.net] that says we should all try to get to X nm by 20xx, and here are the challenges, etc.

      Now someday we're going to get to one of these technology nodes, as they're called, and find out there really is a fundamental phyiscal limitation that keeps us
    • by erice ( 13380 ) on Thursday June 29, 2006 @08:21PM (#15632995) Homepage
      Semi-arbitrary. Each whole generation is half the feature size of the generation before, starting from 0.650 micron (650nm). In between are "half" generations counting down from 1000micron.

      Half generations: 1000, 500, 250, 130, 65, 32, 16
      Whole generations: 650, 350, 180, 90, 45, 22

      The precise digits are chosen for convenience and actual processes vary a bit up and down for a given technology node. Each node requires new equipment. By moving from node to node together, manufacturers share some of the cost of development. Still, odd ball nodes do exist. DRAM's are often manufactured at intermediate dimentions and 150nm is used by some foundaries.

      Many fabless chip makers will skip half generations. I know a lot of manufacturers went straight from 350 to 180. Still, the choice to skip or not is mostly economic. If a node lands durring a recession, fabless chip makers are likely to hold off until the node that follows. The fabs don't really have a choice. They have to produce each generation in sequence, at least at small scale, or they will not have the technological base to start work on the nodes that follow.
    • The sizes are governed by 2 factors, the wavelength of the UV light used, currently 193nm, transitioning to the 157nm for the 45nm chips, and the diffraction gratings/refraction of immersion fluid/polarization of said fluid, etc. used. Due to the physics behind this (i wont bore you with the long equations, because i dont want to do them again) there's basically certain points at which these effects add up to the greatest possible resolution/intensity/etc. Any more in depth and i'd have to dig up my lithography text, and i dont really want to :)
    • Each is the previous divided by the square root of 2.

      The reason for this is that if you decrease the feature size by the square root of 2 on each side, the feature shrinks to half size (since they are 2D features).

      You can see this by squaring them all

      180^2 = 32,400
      130^2 = 16,900
      90^2 = 8,100
      65^2 = 4,225
      45^2 = 2,025
      32^2 = 1,024
      22^2 = 484
      16^2 = 256

      See how each is half the size of the previous?

      I guess doubling the number of things (transistors) makes sense to humans. It sure makes it easy to calculate how many
      • I can't believe how many comments it took until someone just said, "1 /root 2."

        Basically the goal of each node is for the circuits to be twice as dense, so hopefully each design rule scales by 0.707.

  • Appart from the fact that it might be replace by High K/Low K [slashdot.org] (whatever that may be), saying:
    "could reduce the prices of electronic devices for consumers and, of course, create a positive environmental impact"
    Seems wrong: Isn't production only a small part of actual environmental impact?
    Of course any "greener" pruduction method helps, but when I think about green chips, I think about chips like the efficeon chip from Transmeta [transmeta.com].
    • Not only that it's little compared to the running costs. I wonder how much of the energy required for the manufacturing is actually used for this one step. Let's do some calculations:

      Heat capacity of Si at 25 C: 19.789 J/(mol*K)
      Too lazy to look up the heat capacity at other temperatures right now, so let's estimate it to be twice that much at 1000 C.
      Mass of Si: 28g/mol
      So that's up to 1.4 J/(g*K).
      Heating by 1000 K is an estimated 1400 J/g (max).
      How much does you average CPU die weigh? 1g? 2g? Let's assume 10
  • While the new process may not use as much energy in the manufacture of ICs, chips, the article says nothing about water usage. As it is now ic fabs use prodigious amounts of deionized water.

    Falcon
    • Not only that, but there are a lot of harsh chemicals that are used in manufacturing semiconductors, and this is a concern on many levels (exposure by humans, contamination of things such as water supplies, etc.). When I saw "greener" chip manufacturing process, I was initially thinking and hoping that this was what they were referring to, but lower power usage is obviously a nice benefit as well.
      • Not only that, but there are a lot of harsh chemicals that are used in manufacturing semiconductors, and this is a concern on many levels (exposure by humans, contamination of things such as water supplies, etc.). When I saw "greener" chip manufacturing process, I was initially thinking and hoping that this was what they were referring to, but lower power usage is obviously a nice benefit as well.

        I was hoping for the same thing. One thing I heard years ago was how some chemical processes can use substit

  • by kyc ( 984418 ) on Thursday June 29, 2006 @11:22PM (#15633823)
    Firstly, I have to say that, what they claim to serve as a >novel> technique is not completely novel. Of course, it was known that SiO2 could be formed by other means then sintering ( heating the silicon and letting the oxygen atoms dissolve in silicon ) but the problem has always been the purity.

    Semiconductors, especially devices in nanometer scaling need to be extremely pure. Their lattice structure -hence their electrical effects- can easily be distorted or failed by very little deviations, say, in dopant concentrations random dopant fluctuations. This is shortly called , RDF.

    RDF has become a major concern especially for the newcoming generations because basically when you scale down the channel length, the channel lengths are becoming so narrow (and small) that only about 100 hundred dopant atoms fall inside the channel volume. This , obviously, increases the sensitivity and failure rate of these transistors, let alone their variations (like threshold voltages) in a single die.

    From a mass production point of view, we want to get as uniform parameters as we can from a complete die. The ratio of successful ( uniform and working ) transistors to the total die area divided by a single transistor area ( which means the total number of transistors we wanted to harvest from that die ) gives us the `yield`.

    Now, taking into account the fact that even a failure of a single transistor, could lead to the failure of an entire word line of an SRAM , the yield strongly influences the SRAM or chip reliability.

    And for the companies, it does not matter whether you prepare the chip at room temperature but in a more sloppy way, because ultimately it is going to cost more !

    Of course the need for extreme purity in nanoscale devices is not realized completely. The reason is that we have not produced those chips yet. However, these issues ( especially RDF and process variations- you can google these and see yourself) are very hot topics in LOW POWER VLSI design.

    The people who work in these fields are surely aware of the need for an accurate fabrication and will just ignore this kind of work. There are some papers that try to reduce these effects only to succeed in a relatively low way.

    In modern research, you can easily publish a paper by changing the slightest detail of a published paper or you can slightly vary this known application and claim that you have come up with a totally novel ida.
    This is a draw-back.

    In short, they are not going to make anything green, UNLESS of course, they find a better and reliable method satisfying the needs of the upcoming nanoscale devices.

    Then I would shut up
  • It used to be that "nano" was the latest buzzword in materials research, so that anyone doing nano-stuff got funded because it sounded new and edgy. Maybe "green" is going to be the next one? Really what matters is cost - this could make things cheaper. The semiconductor industry doesn't care about being green. I detect more than a whiff of media-targeted spin in this research.
  • I expect better than this from the Beeb (From TFA):

    Moore's Law, that says the number of transistors on a chip will double every couple of years...

    How about we go with the complexity of integrated circuits, with respect to minimum component cost, doubles every 24 months.

    Bit of a difference. Doesn't have a huge impact on the story, but come on - anyone who reports on tech issues should at least have a better grasp of the basics than your average Digg-er.

  • Reminds me of my mom who used to hang the clothes out to dry in the sun's UV. Now we'll do that for our chips. Clothes line technology will have new life!

"You're a creature of the night, Michael. Wait'll Mom hears about this." -- from the movie "The Lost Boys"

Working...