Sunday, 31 August 2014

10:41 – Barbara is ironing and cleaning house while I do kit stuff. Right now, I’m dehydrating some magnesium sulfate heptahydrate to anhydrous form. The hydrated form, AKA Epsom salts, is cheap, a couple bucks a kilo in USP/FCC form, which is as pure as reagent grade. Buying the anhydrate from a chemical supplier runs $80 to $100 per kilo, which is outrageous.

So I just spread about a kilo of Epsom salts in a large casserole dish and stuck it in the oven at 500F (260C) for an hour. That removes most of the water of hydration and forms a thin glassy layer of magnesium sulfate. I break that into chunks and toss it into a blender that I reserve for such work. I then blend it on high to break up the chunks into mostly powder, run it through a flour sifter, repeat as necessary until all the chunks are broken up, and then put the powder back in the oven for another hour at 500F to finish drying it out. I do this while I’m doing other stuff, so the whole process requires maybe ten minutes of my time. Add the cost of my time to maybe two or three bucks in materials cost and electricity, and I end up with half a kilo of magnesium sulfate anhydrate for less than $20, even billing myself at $100/hour.

The price of many chemicals has gotten ridiculously high. For example, we use copper(II) acetate to make up Barfoed’s reagent, which is essentially a 0.5 molar solution of copper(II) acetate with 10 mL of glacial acetic acid added per liter. I was about to order some copper(II) acetate, but found my regular supplier wanted $120/kilo. Geez.

So, the next time I need to make up Barfoed’s reagent I’ll do it from scratch on the fly. I generally make up four liters at a time, so I’ll start with two clean 2-liter Coke bottles. I’ll transfer two moles of copper(II) sulfate to one bottle. That copper(II) sulfate is from Home Depot, which sells a 2-pound (907 g) bottle of the stuff for about $10 under the name of Root Kill. The assay on the bottle says it’s 99% copper(II) sulfate, which I’ve verified gravimetrically. The remaining <1% by mass is mostly insoluble copper oxide. The molar mass of copper(II) sulfate is 249.68 g/mole. Dividing that by 0.99 gives 250.22 g/mole, so I'll transfer 500.44 g of the Root Kill to the two liter bottle and dissolve it in hot water. (It dissolves quickly in hot water; in room temperature water it can take literally a week to dissolve.) I'll then filter the resulting two liters of pretty blue solution into the second bottle, rinse out the first bottle, and divide the solution with one liter in each of the two bottles. So far, I'll have used up maybe five minutes of actual working time and about $5 worth of the Root Kill.

I’ll then add either sodium carbonate or sodium bicarbonate, both of which are cheap, to precipitate the copper ions as insoluble copper(II) carbonate. Once the precipitate settles, I’ll decant off the supernatant liquid, which contains mainly sodium sulfate with a small amount of the excess sodium carbonate or bicarbonate in solution. If I decant 90% of the supernatant liquid and refill the bottle with tap water, I’ve diluted the original level of soluble contaminants to 10% of what they were. Repeating that process a few times, ending with a distilled water wash, reduces the soluble contaminants to 1%, 0.1%, 0.01%, and finally 0.001%, which is better than good enough.

I don’t even need to filter out the copper(II) carbonate and dry it. I can simply wet it with a liter or so of distilled water and add glacial acetic acid stoichimetrically to convert the copper(II) carbonate to copper(II) acetate in situ, add an extra 20 mL of the glacial acetic acid, and then bottle the resulting Barfoed’s reagent.

And don’t get me started on ammonium metavanadate. The last time I bought it, maybe three or four years ago, I paid something like $15 for a 25 gram bottle. I thought $0.60/gram was pretty high then, but that’s now tripled to nearly $2/gram, and that doesn’t even include the required poison-pack container and hazardous shipping surcharge. Geez. I can synthesize the stuff from scratch here for something like $0.05/gram, and it’s no more difficult than the copper(II) acetate synthesis.