Homegrown Oxidizers Part Nine
WSM

The Experimental Cells


The lead dioxide cell with the anode sandwiched between two cathodes.

On the left side of the photo above, a close fitting PVC assembly is shown which is used mainly to protect the Pyrex tank we used. It also added an insulating layer, which wasn’t our intention or particularly wanted.

In the lid of the cell are four fittings: two for the electrodes, one for a Teflon coated thermal sensor and the last one to vent gasses generated during operation of the cell.

In the photos below, showing the running cell, the clear electrolyte appears milky due to the many very fine gas bubbles coming off the electrodes and circulating in a convection-like current within the cell. Many of those bubbles combine and break at the surface of the electrolyte, where the gasses can rise and escape through a vent tube in the lid of the cell. The gasses are primarily hydrogen (H₂), oxygen (O₂) and ozone (O₃).

This detail view shows black Viton rubber insulating barriers on the edges of the anode and clear Teflon tubing supports between the electrodes to rigidly maintain spacing.


The LD test cell two hours after the initial start up.


The LD cell after 48 hours. The voltages in the above two photos are higher than our final reads because they’re from the power supply. For accuracy we later moved the meter connections to the electrode leads.

The first cell run is using the lead dioxide (LD) on titanium mesh anode surrounded by two solid titanium sheet cathodes. The electrolyte is sodium chlorate solution with a concentration of 600 grams per liter. Our intention is to run the cell till the chlorate is down to 100 grams per liter.

As described in part eight of this series, in order to obtain accurate data from these runs we’ve decided to use a small regulated DC power supply capable of constant voltage (CV) or constant current (CC) operation. We elected to run the cells with constant current (where the current remains the same but the voltage varies as needed to maintain the current delivery), for greater ease in calculating the current efficiency (or CE) of the different systems.

In practice the current setting remained “rock solid” with no hint of varying. The voltage in our experiments did vary as the cell chemistry and temperature changed during the runs, but not as much as we thought they might.

Running the LD cell

In the chlorate process several noxious and toxic materials are generated (primarily a strong bleach smell), but in the perchlorate process this doesn’t appear to be the case. That said, we still recommend running the operation in a controlled and well ventilated area, protected from approach by those unfamiliar with or unprepared to deal with the hazards present.
We started with three liters of electrolyte, keeping an additional 400 milliliters in reserve for replenishing the fluid level, should it drop too close to the top of the electrodes during the run. This proved to be a wise decision.

After the first 24 hours of operation, the fluid level dropped about an inch (about 25mm), so the rest of the electrolyte was added to the cell. At 48 hours of operation the electrolyte level only dropped half as much, so it appeared that as the chlorate was consumed (and replaced by perchlorate) the water consumption was reduced. This isn’t certain, but follows our observations.

For the first 48 hours of operation, we noticed the voltage remained consistent (about 4.19 Vdc at the juncture of the power leads and the electrode leads), but after that began to rise very slowly till about 68 hours to about 4.25 Vdc. Our conservative estimates were for 72 hours of operation at the current density we chose (0.1A/cm² anode surface area) to the completion of the run.

Everything else in our setup was running in textbook fashion. The voltage, current and temperature were all within limits that industry says are normal, and with our LD cell running at about 30% to 40% the current density of commercial cells, the ozone generation we noticed was much less than anticipated.

The temperature at the start of the run was at ambient (20°C or 68°F) but slowly rose to between 45° and 50°C (113° and 122°F) and stabilized.

We believe we ran our system a bit too long. After running the cell 69 hours (of our planned 72 hour run), the electrolyte began to have a brown color develop in it. As a result, it’s possible our anode may have some minor damage, as evidenced by the brown tint in the electrolyte.

Fortunately, after the shutdown and sitting for two days, the brown color faded and any residue dropped out to the bottom of the cell. Filtering the electrolyte solution should remove the residue, which we’ll test later for lead (Pb) content.


The brown coloration of the electrolyte appeared after about hour 69. The photo on the left is of the LD cell at shutdown, after roughly 71 hours of operation. The photo on the right is the same cell after sitting two days. The brown material has faded and dropped to the bottom of the cell.

There is some good news though; within a short time of starting the run, we got positive tests for perchlorate. As the run progressed, the positives were immediate and strong. We use methylene blue solution to test for the presence of perchlorate in the electrolyte and a change of color from blue to purple is indicative of perchlorate (even in the presence of chlorate or nitrate).

Methylene blue use is a qualitative test (meaning it tells us whether or not perchlorate is there). What we need is a quantitative test which will tell us how much perchlorate is there.


Methylene blue leaves a blue spot as on the left side, which is the original chlorate solution. The purple spot (showing perchlorate is present) is from the cell electrolyte after running for ten hours in the cell.

The best news is that the qualitative test means we have nearly three liters of raw sodium perchlorate solution (with an unknown amount of sodium chlorate residue) to process from the LD cell.

Running the platinum (Pt) cell


A photo of the Pt cell soon after start up.

A photo of the Pt cell at 72 hours of operation.

Operation of the second cell was with a platinized titanium mesh anode and a pair of titanium cathodes. We ran it with a current density of 0.2A/cm² anode surface area. The anode surface area was roughly one third the area of the LD anode.

Things ran similarly to the LD anode. Due to the higher current density, the side reaction electrolysis of water and the ozone generation was more pronounced than with the LD anode. The voltage values were also higher by half a volt or more, depending on which part of the run was monitored.

With a little off hand calculation we determined to run the Pt test 82 hours, and everything seemed to operate just fine, with no visible damage to the anode. The end point liquor has a light yellow tint to it (as seen in the photo below), but we can’t tell if it’s due to the differences in the two cells because we used sodium chlorate from different sources in each of the two tests.

The Pt cell fluid has a faint yellow tint to it, after running 82 hours.

The test for perchlorate with methylene blue solution in the Pt cell is immediate and positive. We estimate the final fluid level to be about 2.75 liters (roughly 73% of a US Gallon).


Results of the tests

As expected, both setups produced sodium perchlorate. Between the two, it seems the lead dioxide anode shows more promise for longevity in an amateur perchlorate cell, and at a reasonable cost. We were impressed with how close to industrial numbers it ran, but the actual efficiencies of the two cells are still unknown.

Now that we’ve proven we can make sodium perchlorate in a small working model cell, we need to research the purification and conversion of it to more useful perchlorates, such as (sodium-free) pyrotechnics grade potassium perchlorate. A few steps need to be undertaken to get us to that point.

As for the actual efficiency, we don’t know it at this point. The vast majority of the gas generated at the anode was oxygen, and at the cathode was hydrogen, due to the electrolysis of water (the main side reaction). We’re not certain we can accurately, or even roughly, determine the efficiency yet, but we’ll try.

More tests are called for...