Sciencemadness Discussion Board

Gallium-based aluminum reducing reagents, free to USA members who do a writeup

Melgar - 17-7-2017 at 19:18

I've been experimenting with Al/Ga/In/St alloys here, as a less hazardous version of the popular Al/Hg reduction. It can work very well for some reactions, but results are often inconsistent, and depend very much on the aluminum alloy that's used. I said that I'd give a member here some salts to use instead of mercury, and so while I was at it, I prepared some extra, figuring that others here might want to try using this method.

I normally use the alloy, since it's easier to prepare from recycled material, but to compare directly to Al/Hg, you'd need to add the salts to aluminum. This CAN work, as long as salts are added to the aluminum in water at a temperature well above gallium's melting point, and as long as the gallium is added first. As you can probably imagine, there are lots of other variables that come into play here, and I hope anyone attempting this can figure out a few more of them too.

Anyway, I prepared three times what I was was planning to send to the member I mentioned, and I have enough additional materials to give two more members similar packages. Rules are, you have to be US-based (Canadian members will be considered, but no promises) and you have to have an extensive post history with at least one writeup of a reaction that you've done (link to it, if possible). Experience with Al/Hg is a plus, since I'd like the reactions to be compared side-by-side at some point, once this variant is optimized.

You'll be getting gallium and indium trichloride, probably an aqueous solution. Three grams of gallium, and one of indium, not counting the mass of the water and the chloride. A few grams of stannous chloride, just because it's cheap, and because varying the amount of it can have very large effects on the reaction. 30 grams of 3040 aluminum (from the disposable serving trays that I use, just to make sure my results can be replicated) as a 90% alloy with galinstan. And also 30 grams of aluminum serving tray metal chopped up, because you really need thicker metal for the Al/Ga/In/St variant.

The one thing I haven't been able to demonstrate so far, is the mercury-free variant and Al/Hg side by side, but I figured with all the Al/Hg experts here, someone would take me up on that.x

[Edited on 7/18/17 by Melgar]

notoxicshit - 19-7-2017 at 04:24

Very interesting and Id be glad to help but I am european.

Planned to do further experimentation, already have tried an Al/Ga, just with traces of wanted reduction product.
I already bought the indium.
What I wanted to try was a Clemmensen reaction without mercury.
But I'd also try reductions of oximes, later, I think.

Where do the salts come into play?

And if youve just failed at Al/Hg, you say, then what have you achieved so far?

Melgar - 21-7-2017 at 00:09

The point is, like 99% of my experience has been with this stuff, I don't really even know how to use Al/Hg. And its more that I don't have proper lab facilities for determining yields side-by-side.

I actually developed a method of reduction that's a very similar process to the typically Al/Hg. You take crumpled up aluminum, and put it in water. Heat it to 50-70˚C, then add gallium chloride solution while stirring quickly. Keep stirring for five minutes or so, then pour off the water. Add 70% ethanol or so, then add indium salts, at about 1/3 the mass of indium to gallium. Then add a small amount of tin. It should barely be bubbling, if it is at all. But when you add 95% ethanol and some acetic acid to lower the pH, it runs almost identically to Al/Hg.

For reducing oximes, sodium in ethanol works great. Lithium sort of works, but its tendency to float doesn't help.

samm - 28-7-2017 at 13:55

Don't mean to derail but does anyone know if this alloy would reduce the same types of bonds as Stannous Chloride; eg: Nitrobenzene >> Aniline?

Melgar - 28-7-2017 at 15:17

Certainly possible, seeing as it contains tin, and that tin would get recycled. There would be some chloride ions from adding the salts initially, although I'm not actually sure how important the chloride ions are for the reduction.

Still, I can't imagine it wouldn't reduce a nitrobenzene, seeing as that group is one of the easiest possible groups to reduce.

samm - 29-7-2017 at 13:10

IIRC Vogel's recommends ~2x w/w SnCl2/substrate for the nitro reductions.

How does this alloy compare?

Melgar - 29-7-2017 at 22:39

I don't have good experimental numbers to compare this to different reduction methods, but my guess would be that it's most similar to Al/Hg, but with more variables to take into account. I have definitely reduced aliphatic nitro compounds to amines with this, quite easily too.