For cell body I used glass, a 250ml jar with screw-lid where the electrodes and gas exhaust were installed.
I used OTC (well, not in germany, but in France at that time- nowadays very hard to find even there) sodium chlorate solution (45% strength). It is completely chloride- free as a test with AgNO3 solution showed, and this is also important for minimizing anode erosion.
2-4 g/L of potassium dichromate must be added to this in order to prevent reduction of chlorate at the cathode, which would produce chloride and cause anode erosion.
The cell liquor must be magnetically stirred, as proper mixing is very important.
The power supply must be of the type with adjustable current limiting, otherwise you won't be able to keep a constant current.
And 15 A are WAY too much. For a 250ml cell, 2 A are good and 4 A the absolute maximum when used with a cooling bath.
A 0,5mm Pt wire is also not to be used with more than 4 A, or the solder connection will de-solder itself.
The maximum current that can be run through the cell depends firstly on the anode and secondly on the cooling and therefore on the surface of the cell body that touches the cooling bath, and also on the thickness on the glass. Do not let the temperature rise above 40°C.
The progress can be nicely observed by watching the anode.
At the beginning, there will be only very slow oxygen evolution, as that oxygen goes into oxidation of the chlorate.
When there is only little chlorate left to oxidise, the oxygen production increases, and an ozone smell can be noticed, which gets more and more intense as the chlorate is used up.
When nearly all chlorate is gone, there will be strong oxygen/ozone evolution and all energy now goes into splitting of the water.
I connect the exhaust of my cell (connection must be airtight, of course) to a wash bottle filled with water which serves as a bubble counter.
At the beginning, only hydrogen flows through it and the gas evolution is measured by counting the bubbles in a time span of, say, 30 seconds. When the chlorate is used up, the gas evolution will have increased by 50% and the gas consists of a perfect hydrogen/oxygen mix which can be collected in an inverted test tube and ignited FAR AWAY from the cell. It will explode with a loud bang, opposed to the very light puff that the pure hydrogen at the beginning of the electrolysis makes.
To oxidise 1 mol of chlorate into perchlorate, 2 mol of electrons are needed. 1 mol of electrons is 26,8 Ah. Calculate the required ampere hours and run at least the double amount of this through the cell.
When the cell is done, I have a ca. 50% NaClO4 solution (NaClO4 is very soluble and cannot be crystallized easily, so it is a batch- wise process).
I destroy residual chlorate by acidifying the solution, adding sodium disulfite and boiling. The solution must give off SO2, otherwise more disulfite must be added. A SO2 smell from the solution is the sign that all chlorate has been reduced.
This also reduces the added dichromate to trivalent chromium, which is then precipitated as the hydroxide by adding NaOH solution and filtering.
From this I can make KClO4 by adding KCl solution (precipitate is very fine and difficult to filter and must always be recrystallized from boiling water to make larger and purer crystals), or ammonium perchlorate (saturated NH4NO3 solution is needed for this, and NH4Cl is not soluble enough so that the yields will be low if you use this). For ammonium perchlorate, strong cooling is necessary since it is quite soluble at ambient temperature.
Yield of KClO4 calculated from NaClO4 is often as high as 90- 95 %, even after recrystallization (at 0°C it is realy very sparingly soluble which explains the good yields).
For ammonium perchlorate yields are about 50%.
A perchlorate cell is much cleaner to operate than a chlorate cell, but the parameters temperature, current density at the anode and composition of the electrolyte must be closely controlled for it to work. If this is done, then a perchlorate cell is often much more efficient than a chlorate cell.
Next post: From Garage Chemist:
The lid was the ordinary plastic coated metal lid it came with (it originally contained tomato sauce ). The connections and exhaust were fitted into the lid by drilling holes and using silicone. This worked well.
Make yourself some potassium dichromate, it's very easily made from Cr2O3 (available at art supply stores). Go to frogfot's site for a detailed synthesis.
And the hydrogen evolved will NOT result in satisfactory stirring. Use a magnetic stirrer, it can be as simple as a magnet on a motor under the cell and a steel bar coated with plastic or sealed into a glass ampoule inside it.
Otherwise acidic layers will develop around the anode and attack it.
You cannot precipitate NaClO4 by adding anything, it is far more soluble than NaClO3! Even boiling down the cell liquor to half its volume and cooling won't make any solid NaClO4.
I use my NaClO4 as the 50% aqueous solution I get by destroying the chlorate and precipitating the chromium, the only impurities are sodium sulfate, sulfite and chloride when HCl is used for acidification.
For all reactions using it it has to be dissolved anyway, so it would be a waste of time to isolate the solid salt, which is a real pain. I speak from experience (boiling down the solution will make a thick sludge from which no crystals can be obtained).
NaClO4 cannot be obtained in a continous process, as it can't be separated from the solution. It is produced as the 50% solution in water which can be used for all desired purposes.
If the cell is operated correctly, the Pt wire is not visibly attacked.
The only time it got attacked was when I ran the cell without dichromate.
After this, the cell also contained some chloride due to reduction of chlorate at the cathode.
Persulfate will not prevent this. For platinum anode, dichromate has to be used.
It forms a diaphragm of hydrous chromium oxide around the cathode, preventing the contact of nascent hydrogen with the electrolyte (CORRECTION: cathode).
Next post (Garage Chemist)
Of course I meant reduction at the cathode. what a stupid typo!
I haven't treid persulfate, but I don't want to attack the platinum another time, so I keep on using dichromate from which I know it works.