In a research lab at U of Michigan. Just for the sake of continued anonymity I'll refrain from stating which one (since there are several here). We grow quite a lot of GaAs and other III-V combinations. I should mention that we do not grow any GaN, which I think is what has primarily been proposed for white LED lighting.
Just to help clarify, GaAs as it is used in devices is a solid and a pretty stable solid at that. In LEDs and other devices you'll only see it as a thin film or possibly as a wafer substrate, and while it does cleave/break up easily, you'd actually have to grind it into a dust to make it an inhalation risk. I'm in a lab that uses GaAs wafers and grows GaAs thin films regularly, and we don't worry so much about the GaAs as the pure As or As containing gases used to produce it. So as Miseph said, don't grind up your LEDS and eat them and you'll be fine. As for GaAs in land fills, again the stuff is a stable solid. It won't be leaching into the water supplies like lead in electronics is feared to do. Besides, isn't LED lighting supposed to last so much longer than CCFL's or incandescents that we should see less material making it to the land fill?
Spot on. Data is data, but as soon as we start trying to analyze and contextualize it, it becomes something else. Even if pure mathematics are used to analyze a data set, the results must be interpreted, and that process will always possess some assumptions and a degree of subjectivity. Particularly in scientific experimentation, all data is the product of how it was collected.
Not to say that the type of data collection and analysis proposed in the article won't be immensely useful, but it will hardly spell the end of the scientific method. If we are to invent and innovate then we need laws, principles and models to explain the world around us.
There's also the nasty problem that if you heat a metallic glass up too much (thinking high temp aerospace application here) you may very well drive crystallization, eliminating your property benefit and producing a microstructure you have no control over.
If they do figure out a way to cast really big, complex parts of metallic glass, it will still be really cool.
I'll have to agree with anmida here. The glass transition is poorly understood, whether in inorganics or polymers. The scientific community has yet to reach a final definition on what a solid is in all cases, exactly when melting occurs, when a non-crystalline solid becomes a viscous liquid, etc.
I think you've confused your use of the word crystallize as well. Amorphous solids, though they solidify, do not crystallize. No long-range order (lattice geometry as you've put it) means the material is non-crystalline.
Right. Glass just refers to an amorphous solid, inorganic in the cases discussed here. Metallic glasses are just metals that were cooled so quickly (or under otherwise extreme conditions) that they were unable to crystallize into their normal equilibrium structure. The stuff we commonly call glass is silica (SiOx) and some other elements in an amorphous form. When it does manage to crystallize we get SiO2, quartz.
You are quite right. Window glass is a solid and non-crystalline. Crystallinity is defined by long-range atomic order. Glasses are amorphous, meaning they have no or only short-range atomic order.
On your second point however, you are a little off. Bulk pure silicon is crystalline (google the "diamond cubic" crystal structure to see silicon's structure). Amorphous silicon is used for high speed transistors, but there it is a thin film. The primary component of window glass is SiOx, silica.
I'm kind of surprised metallic glasses are being touted here as a "new" advancement. Research into metallic glasses has been ongoing for quite some time.
Slashdot Mirror
In a research lab at U of Michigan. Just for the sake of continued anonymity I'll refrain from stating which one (since there are several here). We grow quite a lot of GaAs and other III-V combinations. I should mention that we do not grow any GaN, which I think is what has primarily been proposed for white LED lighting.
Just to help clarify, GaAs as it is used in devices is a solid and a pretty stable solid at that. In LEDs and other devices you'll only see it as a thin film or possibly as a wafer substrate, and while it does cleave/break up easily, you'd actually have to grind it into a dust to make it an inhalation risk. I'm in a lab that uses GaAs wafers and grows GaAs thin films regularly, and we don't worry so much about the GaAs as the pure As or As containing gases used to produce it. So as Miseph said, don't grind up your LEDS and eat them and you'll be fine. As for GaAs in land fills, again the stuff is a stable solid. It won't be leaching into the water supplies like lead in electronics is feared to do. Besides, isn't LED lighting supposed to last so much longer than CCFL's or incandescents that we should see less material making it to the land fill?
Spot on. Data is data, but as soon as we start trying to analyze and contextualize it, it becomes something else. Even if pure mathematics are used to analyze a data set, the results must be interpreted, and that process will always possess some assumptions and a degree of subjectivity. Particularly in scientific experimentation, all data is the product of how it was collected. Not to say that the type of data collection and analysis proposed in the article won't be immensely useful, but it will hardly spell the end of the scientific method. If we are to invent and innovate then we need laws, principles and models to explain the world around us.
There's also the nasty problem that if you heat a metallic glass up too much (thinking high temp aerospace application here) you may very well drive crystallization, eliminating your property benefit and producing a microstructure you have no control over. If they do figure out a way to cast really big, complex parts of metallic glass, it will still be really cool.
I'll have to agree with anmida here. The glass transition is poorly understood, whether in inorganics or polymers. The scientific community has yet to reach a final definition on what a solid is in all cases, exactly when melting occurs, when a non-crystalline solid becomes a viscous liquid, etc. I think you've confused your use of the word crystallize as well. Amorphous solids, though they solidify, do not crystallize. No long-range order (lattice geometry as you've put it) means the material is non-crystalline.
Right. Glass just refers to an amorphous solid, inorganic in the cases discussed here. Metallic glasses are just metals that were cooled so quickly (or under otherwise extreme conditions) that they were unable to crystallize into their normal equilibrium structure. The stuff we commonly call glass is silica (SiOx) and some other elements in an amorphous form. When it does manage to crystallize we get SiO2, quartz.
You are quite right. Window glass is a solid and non-crystalline. Crystallinity is defined by long-range atomic order. Glasses are amorphous, meaning they have no or only short-range atomic order. On your second point however, you are a little off. Bulk pure silicon is crystalline (google the "diamond cubic" crystal structure to see silicon's structure). Amorphous silicon is used for high speed transistors, but there it is a thin film. The primary component of window glass is SiOx, silica. I'm kind of surprised metallic glasses are being touted here as a "new" advancement. Research into metallic glasses has been ongoing for quite some time.