Unconventional electrolysis

  • Update 1: After just 10 minutes, the electrolyte has turned opaque black and the current is up over 6 amperes. Should I dilute the electrolyte more?

    It depends on how slowly you intended the experiment to progress. Typically, once I start adding HCl, eventually the deposition process accelerates rapidly, and current increases to around 8-10A (at 12V) unless the electrodes are in a short-circuiting state. At this point the resonating sound would start, presumably generated by micro-explosions in the gap portion above the electrolyte level.

    Interesting that with the mild steel electrodes the process seems much faster than last time. It makes sense if the other ones were made of stainless steel, though. That also probably means that with SS you could use a narrower gap than what I observed to be manageable with no direct width control means.

  • Is clacker motion also fully programmable so that it can be varied depending on external dynamically changing variables? (current, etc)

    It's manual control only unfortunately. The function generator and amplifier module are cheap ones from Amazon and have just a typical volume control for amplitude and push buttons for setting waveform and frequency. Full automation with a stepper motor would be nice but probably overkill at this stage of development.

    The current rapidly rose to over6 amperes, which will heat the electrolyte pretty fast. Ive started a video stream now and see what the clapper does.


  • Looks like there's enough rust built up to prevent metal-to-metal contact. The clacker only seems to increase the electrolytic current slightly, possibly by dislodging the bubbles forming on the surfaces. I will let it run for some time to see if that changes.

  • magicsound

    If there are oxides on the anode preventing electrical contact, it's possible that solution pH is still not sufficiently low and/or that gap width is such that nascent hydrogen from the cathode isn't recombining with oxygen at the anode at a rate high enough preventing them to form. Can you measure solution pH safely?

    As a side note, I think in my most of my tests water was at boiling temperatures at least inside the gap, although I've never measured its temperature directly.

  • If there are oxides on the anode preventing electrical contact, it's possible that solution pH is still not sufficiently low and/or that gap width is such that nascent hydrogen from the cathode isn't recombining with oxygen at the anode at a rate high enough preventing them to form. Can you measure solution pH safely?

    Well, running the clapper would certainly do that, even with a small gap. Lots of turbulence and pumping action would disperse the gas rapidly out of the gap. I think cleaning the electrodes before the next run is needed.

  • magicsound

    Yep, ideally the electrodes would be constrained in a precisely tight location as to limit gas dispersion from them moving apart, but that would probably be overengineering something that was originally intended to work around equipment and materials limitations in the first place. The graph I previously posted might have shown that such oxide layer was progressively breaking down on the other hand (although water conductivity also increases similarly with temperature), and perhaps metal-to-metal contact could have been possible after a few more minutes of operation.

    As for electrode cleaning, I have used, not necessarily all at the same time:

    • Scraping them with a plastic brush;
    • Scraping them against each other;
    • Scraping with 180-grit sandpaper;
    • Washing with diluted HCl;
    • Washing with warm water;
    • Washing with water and soap

    And they don't seem to affect their capability of reproducing the resonating sound, but it's possible that soap/detergent impurities can promote foaming afterwards.

    Letting the deposition layer dry after using the electrodes will usually make it much more difficult to remove, even if it is relatively loose just after bringing them out of the solution (at least the HCl concentrations used in my tests).

    EDIT: by the way, I tried manually sampling GMC-320+LN7317 Geiger data from yesterday's test run, although probably it was too short for any trend change to be significant. EDIT: I added also ROI CPM from the gamma spectrometer.

    EDIT2: added spreadsheet (.ods LibreOffice format).

  • can thanks for the details on cleaning the electrodes. Your comments on gas recombination between the closely spaced surfaces is a good insight worth considering, but hard to prove without collecting and analyzing the remaining gas in the head space.

    I've been delayed by other technical issues, like checking for ground loop issues on the radiation detection and DAQ circuitry. I'll get the cell apart and cleaned later today, and hopefully be ready for another test run first thing tomorrow.

  • magicsound

    The main reason why I think that it's happening is that the anode (+ electrode) in my narrow-gap electrolytic tests never formed a thick oxide layer and just washing it under water and rubbing it lightly with a nitrile glove would leave a shiny surface, often more so than the cathode. This was especially noticeable with KOH electrolyte (see this photo I posted a good while back. The anode is the left electrode; the other is the cathode, with a very thin deposition layer), but it can be argued that experimental conditions are different with that.

    When you'll take the cell apart later today for cleaning, try checking out if that's also what's happening in your case to some extent, but don't let the electrodes dry in the air before that, as they will likely oxidize quickly.

    Given the apparent (and strangely opposite) correlation of the Geiger and gamma spectrometer to (probably) current it's worth making sure that everything is ok on that regard. On a related note, did you have the AM radio running at the time and noticed broadband noise increasing as current also increased?

  • did you have the AM radio running at the time and noticed broadband noise increasing as current also increased?

    Unfortunately my lab environment is pretty noisy for AM radio. The noise comes from outside my property, probably from the 22 kV power lines that run down my street. I've struggled with it during years of short-wave radio listening, and haven't found a solution.

  • Here's how the electrodes look. Not surprising that there was no contact!

    The cathode on the left is almost invisible under the crud. It's made from a piece of 1/8 inch mild steel strap.

    The anode made from a thicker piece stayed surprisingly clean by comparison.

  • magicsound

    Possibly, by introducing more HCl and/or allowing current to further increase, that layer could have got reduced and become conductive and more easily displaced. Water exceeding boiling temperature inside the gap too likely has an effect on the processes involved.

    EDIT: In my tests with HCl I never got into a condition of too low conductivity, but it sometimes happened, with the electrode gap filling up similarly to the above photo, with an alkaline electrolyte. With HCl in higher amounts as customary in my tests, the particles form a low viscosity slurry in the gap as the photo below from an older test hopefully shows.

  • I left the electrodes immersed in tap water overnight, to prevent drying of the residue that ended yesterday's test.

    This morning scraped the residue out with a putty knife. It was completely soft and came out easily. The residue is mostly black in color, with only a few traces of red rust. After a thorough rinse, the electrodes now conduct well and don't need to be disassembled.

    The residue is also strongly magnetic, so probably contains mostly Iron particles. That also explains why it concentrated in the gap, drawn there by the field of the Nd button magnet at the top of the cathode. So next test the magnet will not be used. I'll set the gap a bit closer, maybe 1 mm. The clacker mechanism will not be activated until the first sign of arcing is seen. At that point it will be driven at low frequency by a reverse sawtooth, pulling the electrodes apart periodically to (hopefully) prevent latch-up.

    After that I'll try connecting the clapper driver input directly to the 0 to1V signal from the current shunt resistor. That will essentially create a buzzer mechanism, should be interesting....

  • magicsound

    Basically that's what I also observed in my tests. The black particles produced from narrow-gap electrolysis are ferromagnetic and can be easily moved around and reoriented with a small Nd magnet, even outside the jar (which is where I often put it in earlier tests. In newer ones it was omitted). With a magnet placed that way the particles would form spikes and so on on the electrodes, but I don't have a photo of that from tests with HCl electrolyte, so here's one with KOH.

    A more recent photo with HCl was made with partially dried residues scraped from the electrodes and a slightly magnetized cutter.

    Perhaps applying a magnet directly on the electrodes could be attempted only after arcing starts. Disappointing that potentially the last test didn't work because of this.

    Looking forward to seeing what will happen in your next experiment.

  • A while back I linked this source (specifically, DOE-HDBK-1015/1-93, page 117 of the PDF document or CH-02 Page 15) on the corrosion of iron in water at various pH levels. Going along figure 7, it might have taken a slightly lower pH to accelerate the process to a much quicker rate.

    It started at around pH=1.5 but the electrolytic current was too high at 15 amperes. So I diluted it to the present level, which gave 3-4 amps. For the next run, I've removed about 1/3 of the liquid to give 5 cm of immersion. That will also reduce the current, so a bit more acid may be needed to set the current again.

  • magicsound

    I never measured this precisely but in the last tests I performed the electrodes were immersed in the liquid by visually about 1.5-2.0 cm. The actual electrolyte level, for the sake of reproducing the resonating noise, didn't seem to be critical as long as it was low; the narrow-gap electrode portion outside the electrolyte might have been what mattered the most, and in your case it seems it would end up being much larger.

    The steam and gases produced typically managed to push the wetted iron particles (and to keep them wet) up to the top of the unimmersed portion of the electrode gap, and electrolysis continued to some extent there as well, as some of the videos I made with the electrodes out of the solution showed.

    Other that in my tests water heated up much faster and that I didn't make any effort preventing it from evaporating (which it quickly did), there don't seem to be other large differences. All that remains should be setting a lower pH level.