Gallium-Aluminum

October 31, 2007 5:18 PM

What are the temperature and pressure needed to make gallium-aluminum alloy?

This is for a science fair project, and we're seeking a good reference.

And if you know the price of low-grade gallium that would be helpful too, thanks.

posted by king walnut to science & nature (15 comments total)

One of my profs mentioned this book in class the other day: Materials Selection in Mechanical Design, M.F. Ashbury 1999, Butterworth-Heinemann. Find it and it may have what you need.
None of my textbooks have anything about a gallium-alu alloy, which is weird as many specific aluminum alloys are covered in great detail in some of my books.
posted by Ctrl_Alt_ep at 5:49 PM on October 31, 2007

Google Books has Materials Selection in Mechanical Design, but a search for gallium returned zero hits.
posted by king walnut at 6:01 PM on October 31, 2007

I'm not sure it's possible to buy "low-grade" gallium; no one would want it, so no one will stock it. The stuff isn't that difficult to purify; there's little producer incentive to leave it "low grade".

This place's current price is 980.00 EUR/kg when buying 5 kg units. Buying smaller units would raise the per-weight price, I would assume.

In recent years the price has fluctuated quite a lot as a function of changing demand. The largest consumer of Gallium right now is the semiconductor industry, and the value and use of GaAs in semiconductors and LEDs has varied a lot.
posted by Steven C. Den Beste at 7:13 PM on October 31, 2007

Search aluminum-gallium alloy, not the other way around.

Materials Handbook, 14th edition says gallium occurs naturally in almost all aluminum ores in the ratio of .11 to .22 lb of Ga to ton of Aluminum. So it's an alloy as an ore. Also, low-grade Ga doesn't make sense: "Commercial Gallium has a purity of 99.9%." It's produced as a by-product of processing other metals. I would say nothing special is required to make this alloy, except melting them down together. Since gallium melts at 30 degrees C, and Al at 660 C, well, I think 660 C, atmospheric pressure is your answer.
posted by Eringatang at 7:24 PM on October 31, 2007

There are at least three ways to make an Al-Ga alloy (and generally any alloy): put pure bulk Al and Ga in contact and wait a long time, combine Al and Ga in the liquid phase (where things mix easily) and freeze, or deposit Al+Ga atom-by-atom in the vapor phase (this is called chemical or physical vapor deposition, depending on whether a bona fide chemical reaction occurs).

I'll assume you're asking about semiconductor/photonics applications of Al_xGa_1-x. To be very general again, you will typically make wafers and substrates by the mixing and freezing approach and make thin layers by the atomic deposition approach.

If you're going to mix and freeze liquids, you want to look at an Al_xGa_1-x phase diagram. A good reference is ASM Handbooks, vol 3: Alloy Phase Diagrams. It tells you, for example, that the addition of gallium to aluminum decreases the melting temperature monotonically from 660°C to 30°C at 1 atm pressure. It also tells you that, above 30°C (about room temperature), if you have have more than 20 wt% gallium in your alloy it will be slushy (part liquid, part solid). For more precise information, you need to specify an alloy composition.

To address the cost question: if indeed you're looking at the semiconductor/photonics industry, the raw material cost of gallium will be negligible compared to the tools and cleanroom conditions needed to purify it and fabricate a useful device.
posted by Mapes at 8:12 PM on October 31, 2007

Thanks for all the responses. Actually this is not for semiconductors, but part of a process of generating hydrogen to power vehicles. See this for example. This researcher, Jerry Woodall, claims that "low-grade" gallium can be used, but I don't know exactly what that means.
posted by king walnut at 8:48 PM on October 31, 2007

"Most people don't realize how energy intensive aluminum is," Woodall said. "For every pound of aluminum you get more than two kilowatt hours of energy in the form of hydrogen combustion and more than two kilowatt hours of heat from the reaction of aluminum with water..."

And since that reaction with water is going to happen below boiling temperature, producing only very low-grade heat, that's half the energy you used to make the aluminium just wasted in the reaction tank.

Running internal combustion engines using non-rechargeable battery chemistry, which is effectively what's being proposed here, strikes me as massively inefficient.

Replacing gasoline with hydrogen for transportation purposes would require the production of huge quantities of hydrogen, and the hydrogen gas would then have to be transported to filling stations.

I don't understand why hydrogen filling stations couldn't just have on-site electrolyzers powered by cheap off-peak electricity. Using metal as fuel strikes me as a poor solution to a non-problem.
posted by flabdablet at 9:14 PM on October 31, 2007

Have you tried to e-mail Jerry Woodall or a student in his lab?
posted by Eringatang at 8:29 AM on November 1, 2007

Oh, I see what's going on. That's pretty clever.

Essentially, the role of gallium is to remove the Al2O3 skin which is always present on a chunk of aluminum. Gallium is a liquid at slightly over room temperature (i.e. outside on a warm day, or held in your hand (in plastic, and wearing a glove!)).

Aluminum oxides quite easily. I would think it would be possible to simply rub gallium around the outside of a chunk of aluminum to remove the oxide. (i.e. a gallium-coated chunk of Al.) As a result, you have a protected Al surface which will stop being protected when you throw it in water. I would think you would probably want very small chunks of Al which will then have a large surface area.

Gallium is available from any chemical supplier (Alfa Aeser, Fisher, etc.) in fairly small chunks. When Jerry Woodall talks about low-grade, he means something other than the 5N+ (99.999%+) gallium used in the semiconductor industry.
posted by JMOZ at 1:09 PM on November 1, 2007

I should mention that you should definitely read all the relevant MSDS (Materials Safety Data Sheets) and use proper precautions when handling these materials. Also, note that gallium is quite corrosive to many metals.
posted by JMOZ at 1:10 PM on November 1, 2007

According to an Al-Ga equilibrium phase diagram the solubility of Ga in Al at room temperature is about 6% (higher than that and you have a mixture of an aluminum phase and a gallium phase). The diagram suggests that a mixture of 6% Ga and 94% Aluminum starts the transition from liquid to solid at about 900 K (627 deg Celsius).
posted by gspm at 3:15 PM on November 1, 2007

don't understand why hydrogen filling stations couldn't just have on-site electrolyzers powered by cheap off-peak electricity.

The numbers don't work.

Just pulling a number out of my ear, let's assume that a "medium" filling station now has about 200 customers a day who buy an average of 10 gallons of gasoline, i.e. 2000 gallons per day. (Some service stations do a lot more than that, of course.)

A gallon of gas is 124,000 BTUs. A BTU is 1055 joules. So 2000 gallons of gasoline is about 262 gigajoules.

Electrolysis isn't very efficient; a lot of the energy goes up as heat. If you're getting 33% you're doing really well; I think it's probably worse than that. But figuring 33% conversion, then to produce 262 gigajoules worth of hydrogen in 8 hours you'd have to draw 27 megawatts. At 110V that's 248,000 amperes.

The power grid isn't designed to deliver that kind of power to small urban customers.
posted by Steven C. Den Beste at 12:48 PM on November 4, 2007

I was wrong. Electrolysis can be as much as 94% efficient.

If you figure 90%, then the service station would need to draw a bit over ten megawatts, which at 110V would about 92,000 amperes. Which is still a real problem.
posted by Steven C. Den Beste at 5:33 PM on November 4, 2007

Steven: Did you take into account the efficiency of the combustion engine vs. the hydrogen fuel cell? Given that the I.C.E. is roughly 30% efficient, you can reduce your amperage by another factor 3 if fuel cells were substantially more efficient.

Also, power supply to a service station would likely be at 440V (or at a minimum, 220V), so now we're talking about ~7700 amps of current (I have assumed all your other numbers are reasonable, since I've not looked), albeit at higher voltage. I'll admit this is still a lot, but it's not unheard of in commercial centers.

Also, I'm not sure that the current usage statistics you assume (2000 gallons a day) translate very well for an electric car. It's probable that usage would differ pretty substantially.

I agree that electrolysis-based stations are not terribly practical for large scale implementation at the present time, but I think they're probably quite a bit less impractical than you've suggested.
posted by JMOZ at 11:31 AM on November 5, 2007

OK, so I'm going to stick with your 200 customers a day number, and pull some more numbers out of my own arse.

The Tesla Roadster's battery pack holds about 56kWh at full charge. That's a battery/electric vehicle, not a fuel cell one, but most of the technology in it is going to be similar to what a FCV would use (a comparable FCV would probably weigh somewhat less).

So, taking that as a typical energy storage requirement for members of a future electric fleet, and assuming a combined electrolyzer and fuel cell efficiency of around 50%, we're looking at supplying each of our 200 fuel-cell-vehicle customers with 120kWh of energy. 120kWh * 200 customers/day / 24 hrs/day = 1000kW = one megawatt, not ten.

But even ten megawatts is not an unreasonable consumption rate for an industrial facility (hell, the Melbourne casino has gas-fired backup generators in its basement that are good for 22 megawatts, and they never get fired up). In any case it's probably easier to run heavy cabling to a service station than it is to run a gas pipeline.

In any reasonable post-fossil-fuel energy economy, the grid is going to need about twice the capacity it has now, simply because electricity will be used to power many of the things that are now powered by fossil fuels. I just don't see that as a terribly serious technological blocker.

Electricity is a really good way to move energy from one spot to another.
posted by flabdablet at 2:55 PM on November 6, 2007