In my chlorate adventures so far, all has gone relatively smoothly with one exception... not knowing when to halt the chlorate cell production. When using KCl as the electrolyte, the less soluble KClO₃ begins to fall out of the solution once it has become saturated. The KCl is consumed, and as the chloride ion concentration drops, the rate of production rapidly falls. Worse, the electrodes, specifically the anode, begins to take a real beating. Per Wouter, as the chloride concentration approaches 10%, perchlorate begins to form. That's great if you have an anode capable of perchlorate production, but my cell has a MMO (Mixed Metal Oxide) anode that is unhappy producing perchlorate. Thus, knowing the concentration of the remaining chloride ion is important to maximize yield.

There are examples available on the web using Silver Nitrate titration methods to determine quantitative chloride in the electrolyte, but when Miech posted that chloride test strips were available, I jumped all over that. The ones I found are made by Hach.

If you go for these, be sure to get the high-range strips, which test between 0.2g to 6.0 grams / liter. The first step, for me at least, was to verify that the strips are reasonably accurate, and do in fact work. At less than a buck a strip, I don't mind "wasting" a couple to get the hang of their use.

I decided first to prepare a mid-range salt solution. My beaker was 40 ml; 0.12 grams of NaCl dissolved in 40 ml should create a 3g/l standard solution.

The strip is inserted and allowed to remain in the test solution. The column is silver nitrate based; some sort of brown powder or gel. At the top of the test strip is what appears to be a simple yellow nylon string which is soaked in some sort of indicator. Immediately, a much lighter tan color begins to creep up the strip. Hach calls it "white" - regardless, the color change is very obvious. When the moisture reaches the yellow string, it rapidly changes to a brown color, and the test is complete at that point.

As another test, I noted the level of the tan color, but left the strip in the solution to see if it would continue to rise. It did not, which is pretty interesting to me. There is plenty of chloride ion remaining in the beaker, and I would have supposed that the color change would have continued until all of the silver nitrate indicator had been converted to silver chloride. The fact that the color only went so far, and no further, is a sign of a good, reliable strip titration mechanism.

The standard solution yielded a peak of 5.2 on the test strip, and that correlates to 1787 mg/L chloride ion. Initially my thought was "whoa, that's not even close. The stock solution is 3000mg/L" But that is 3 grams of SALT, not chloride ion. Chloride is 61% of NaCl, so 3.0 X (0.61) = 1830 mg/L.

Stock solution: 1830 ppm

Strip: 1787 ppm

YES!! Very nice and well within the errors of this simple test.

On to the used electrolyte from the chlorate cell. I knew the concentration of the liquor was well beyond the range of these test strips. The simple answer is accurate dilution of the test sample. 10 ml of the liquor was diluted to 100 ml using distilled H2O.

The dilted test solution was placed in a small beaker, and once again, a test strip was inserted. The titration was rapid, and I was unsure if the dilution was adequate. When the yellow string turned brown, the white band peaked at 8.9, which was above the high-end of the strip. Even though a bit of brown remained, the label on the strip bottle stops at 8.0, which indicates anything above that is probably not valid.

Another sample was dilted 20:1, and this sample peaked at 7.8, still a bit high. I wanted the peak to be a bit lower. The sample was again cut in half, now to 40:1. This one worked perfectly. Here are all of the tests, from left to right, the NaCl standard, and dilutions 10:1, 20:1, and 40:1

The scale is NOT linear, thus the 40:1 sample is not 1/2 of the 20:1 sample. The strip bottle reveals these Cl- ion concentrations:

20:1 = 7.8 = 5512 mg/L

40:1 = 5.8 = 2288 mg/L

Again, 7.8 is too high on the test strip, and enters a region of inaccuracy. The 40:1 dilution was right in the heart of the test range, and with test strips like this, as opposed to a true titration, you want to be in the middle, and not at the extremes.

Now, for the conclusion, and the primary reason I'm doing this, what is the chloride ion concentration in the used electrolyte from my cell? Remember, we want the chloride ion concentration to be 10% or greater in a chlorate cell...

2288 mg/L X 40 = 91.5 g/L for the used electrolyte. So I was just a bit lower than 10%. it's a good thing I halted the chlorate production where I did.

The solubilities of salts of interest are...

Salt: 0 deg - 100 deg

KCl: 0238 - 0567

KClO₃: 0071 - 0570

KClO₄: 0008 - 0218

NaCl: 0357 - 0391

NaClO₃: 0790 - 2300

NaClO₄: 2090 - 2840

A saturated KCl solution on a hot summer day, at maybe 40 degrees celcius, is probably about 350 grams per liter, of which slightly less than half is chloride, call it 47.5% chloride ion by weight, so a saturated KCl starting electrolyte would be 166 grams/L chloride.

Whew! We process, then, a starting electrolyte of 166 g/L chloride, and continue until we reach 100 g/L, or a 10% chloride ion concentration.

Theoretical yield of KClO₃, which has a MW of 122.6: The chloride ion is 29% by weight of the KClO₃ molecule. Therefore, 66 g/L choride ion should convert to a harvest of 227 grams per liter of starting KCl electrolyte. With my cell holding 3.5 liters, a full harvest should be about 795 grams. And interestingly, my first harvest was slightly less than that, while the second was about 500 grams. It looks like I'm doing OK.

Future use of the test strips: These strips serve two purposes. The first use is to monitor the chloride concentration in a cell. With 100 g/L being the minimum chloride ion concentration, a 40:1 dilution of the electrolyte should not be allowed to fall below 2500 mg/L, which is 6.0 on the strip. The second use, to come later, is to test dissolved samples of the final product to determine remaining chloride, which, while not dangerous, could wash out colors.

If you've read this far, congratulations, I welcome you to Dork City, my new home. It's all good!

Standard disclaimers apply: Chlorates can be dangerous. Be smart, don't make illegal fireworks, be a good neighbor, and peace will reign.

One final note: The Ring Lord is a great source for commercially pure titanium and niobium wire. I ordered a pound of 1/16" commercially pure (CP) Ti wire, and it arrived in two days! I plan on sput welding a pure titanium electrode "cage" which will sorround my anodes in future cells.