Anode discolouration

[Martin]

Here’s a rather pretty image of the discolouration from the anode that @Bingqiao returned to me:

At the bottom, above the shadow, we see the brown discolouration that we found on the cathode-facing side of the anode. Above this you see a transition to rainbow discolouration. At the very top you can see the natural platinum colour.

My initial fears were that this was caused by loss of platinum, as titanium is well known for taking on rainbow oxidation patterns. There are two things that give me hope that this isn’t the case:

Firstly, in this image of the top of the electrode:

The left of shows the un-plated titanium, with plated on the right. The discoloured section appears to have the same platinum plated texture. It could be that the platinum is porous and the titanium is oxidising beneath it, but to me it looks like the rainbow is relatively continuous and on the surface of the textured platinum.

The second is thanks to @gerrit stepping in to help. The depicted electrode arrived the day I went on leave, and I’ve been off work since with man-flu that I went down with on my first day back. I’ve been using PHREEQC to determine the theoretical conductivity of the solution we were working with - unfortunately I wasn’t taking notes during the training sessions so don’t have the measured conductivity or even temperature.

My plan was to run electrolysis at the same conductivity but with a single salt. Gerrit gave this a go while I was still running calculations. I understand he used excess sodium bicarbonate and ran at 10% intensity (measuring 13 mA and ~4.9V) for over 20h without seeing any discolouration.

These results combined made me wonder if what we are seeing is ferric (phosphate?) deposition at visible thicknesses in the brown area and wavelength thicknesses in the rainbow area.

I haven’t been well enough to go out for strong acid, but swabbing the second quarter from the left in this image with distilled barley malt vinegar:

…visibly transferred some brown to the swab. That suggests deposition. I’m a little concerned that the rainbow wouldn’t lift there, but hope that it will with a strong acid.

So, I’m hopeful that we are not irreparably impacting the platinum plating, but if this is ferric phosphate deposition, it will reduce our effective anode area and increase required overpotential. Chris French mentioned switching away from phosphate previously. This may be another reason to try that.

[Bingqiao]

Hi! we have some updates now. I utilised citric acid to wipe the anodes days before, but nothing happened. Then, I applied 0.05M and 0.1M sulfuric acid to wipe, and the discolouration became weaker but still visible (Figure 1 (before) and Figure 2 (after)). The electrodes are with a metallic lustre (figure 3). When I rub the black area with unglazed porcelain, there is a black dot on the porcelain (figure 4).

figure 1, before wiping with sulfuric acid

figure 2, after wiping with sulfuric acid

Figure 3, electrodes with metallic lustre

Figure 4, the black dots rubbed by electrodes on the unglazed porcelain squared in red.

Another interesting thing is about the autoclave. I sent vials with electrodes to be autoclaved after wiping with sulphuric acid. When I received that, the black area is much more black than I sent them (figure 5).

Figure 5, the electrodes received from autoclaving.

[Bingqiao]

I think the black colour and metallic lustre are from different reasons. The black colour can be removed by wiping with sulfuric acid and the unglazed porcelain, but metallic cannot be.

[Martin]

Really useful observations!

When you mentioned it going black during autoclaving, I was thinking it might either be the iron phosphate somehow turning to a black compound, or a carbonised biofilm. Gemini suggests black iron compounds include Fe3O4 (magnetite), FeS, Fe(OH)2, or Fe2O3 (hematite).

Of these you would only normally expect magnetite or hematite to have a metallic lustre, or rarely carbonised biofilm.

To tell the difference between the two minerals, I understand geologists perform a streak test: with magnetite the streak would be black, with hematite it would be reddish-brown. I imagine a carbonised biofilm’s streak would be also be black, but it this would be softer rather than gritty magnetite.

A final test would be that magnetite is magnetic, whereas the others are not. But I don’t imagine you’ll have enough to be able to determine if it’s magnetic unless you have loose particles of it and a neodymium magnet.