This page expands on the last cell I posted about. It's still going good (as of 7/8/07 -- three months!), though the graphite took a bit hit of wear (maybe 1/8" off the thickness) when I let it run too long. Of interest, I upgraded the power supply. I now have 6VDC available at 60A load. That's a bit much for a cell, so I connected a ballast resistor -- a chunk of steel wire -- in series, which drops cell voltage to a more suitable 4V. I also installed an ammeter, which probably came from a battery charger. It isn't very accurate, but it shows around 60 amperes to be a typical current, and I have no reason to doubt that. The cell now reaches 65°C, which is hot enough for better efficiency, or something. Fed up with bad connections, I also copper plated the top four inches of the graphite anode. (That reminds me, I need to write a page on that.) It is corroding, due to some particularly malignant pores going through this bar of graphite, but that's not much of an issue as in a few months I can merely peel off what's left and replate it, at the expense of a few days downtime.
I've been running pretty well constantly for the last few weeks, and have come up with a few more pounds of sodium chlorate. I estimate I am producing 2 pounds per week. How much is that? Well, 2 pounds is about one kilogram. NaClO3 is 23 + 35 + 3*16 = 106 g/mol molecular weight, so I have about 9 moles per week. It takes 6 mol electrons to produce 1 mol of the stuff, so, at 100% efficiency, I must be using 54 moles of electrons per week. But 60 amperes for one week is 420 A-days, or about 10 kAh. (Which, at a typical 7V supply (that includes losses), is 70 kWh, or at $0.10/kWh, some $3.50 per pound...if I were paying the electric bill.) Since one faraday is 96485 C/mol, or 26.8 Ah/mol, 10 kAh therefore is 376 mol electrons, or 63 mol NaClO3. But I'm only making 9 mol/wk. Holy crap, that's 100 * 9/63 = 14% efficiency! I've never heard of anything so atrocious.
To remedy this problem, I think I'm going to get rid of the nasty liquor I've been using. It's kind of foamy and yellow, a lot like well, bleachy-smelling urine, or beer. I suppose there's not much difference between those analogies... It would be effective to dry it out, then wash with hot water to recrystallize the NaClO3, leaving the NaCl and impurities behind. Twiddling the pH may also have some effect, though transition metals (I would suspect Fe and Cr) tend to precipitate in high pH, which is what this is, which makes absolutely no sense to begin with. There must be some complex at work, but chloride has been up and down lots of times with no change, and though the chlorate ion has a lone pair (like ammonia and other ligands), it's tightly held by the three other oxygens on the molecule. The pH isn't sooo high that something like chromite could be present. And that's just those...who knows what else is present, sapping precious efficiency.
Much like London rebuilt with stone and brick after the great fire, in the process removing the environment in which black plague spread, I think after this realization I'm going to improve my structure as well.
Up to now, I've been using the following process: after a few days' run, I boil the liquor (in the cell, as I've said before), then dump it into a bucket to crystallize. The solid is broken up and freed from the graphite sludge as best can be done (still, the product is black). The liquor is then filtered from the graphite, ran through a salt column to resalinate it, then it finally finds its way back to the tank, where it awaits another cycle.
In my new plan, I will more or less follow the above diagram. Assuming much better efficiency (I'll try to keep the pH reasonable), after a few days' run in the cell (1.44 kAh = 9 mol, or about a kilogram at 100% efficiency), I'll change the electrolyte. (I could hot filter it, but this seems to be working well enough.) On cooling for 24 hours, I'll seperate the NaClO3 from the liquor. The crude sodium chlorate goes into a bucket for storage, while the liquor is filtered, boiled with salt and sent back to the tank. There, it will cool, precipitating some more NaClO3, while being a much stronger chloride solution than the cold salination I have been using, will allow more run time -- and yield -- before the anode gets into trouble. Perhaps weekly, I'll recrystallize the crude product, by boiling it in water, hot filtering and crystallizing. Some I will mix with similarly hot and concentrated potassium chloride solution, giving a crop of crude KClO3. This may also be recrystallized to give exceptionally pure KClO3, which it already appears will be very popular with ameteur pyrotechnicians. All the leftover liquor and wash water will be recycled by hot salination, supplying the necessary salt and water feedstocks for the process.
This came about since, in my hurry to keep up with the increased rate, I've been skimping on filtering (trying to use the salt column instead) and skimping on crystallizing time, trying to cut off a day here and there. (It really takes 24 to 48 hours at room temperature (~70°F) for all the stuff to finish crystallizing. Any shorter and you'll end up with a fine dispersion of crystals when you start working with it, or worse.) This has left me with a salt column clogged with graphite crud and sodium chlorate, as well as very low chloride concentration at the tank. Last week, after three days' run (which in my experience should've been about complete conversion of the liquor), I discovered the anode was missing a big layer of thickness, instead having a mushy surface of graphite oxide! Clearly, the chloride ran out and it moved on to perchlorate, something it does a very bad job at, oxidizing the graphite instead of the chlorate. Ouch. To attempt to fix this, I've dumped my salt column and boiled the gravelly mass with clean water, removing graphite and solubles. NaClO3 crystallizes on cooling, while the brine remains, ready for the cell after a filtering. I haven't gotten to drying the nasty electrolyte yet; I have a few other solutions pending evaporation that I should do first.