I misread and thought you wrote "H2O2" instead of "2H20". I now see that you actually meant that H2O2 may possibly also be formed (in lieu of 2H2O), but not in an energetically favorable way. What I tried searching on Wikipedia was if H2O2 formation from OH (in any form) is a known reaction.
H2O2 is produced electrolytically for sure, but that needs to be done under very controlled conditions otherwise the yield is very low.
Conditions are of course far from being controlled here. Then, OH- is probably not directly involved in the observed reactions as I previously assumed.
Perhaps [at least the explosions] might be something as "simple" as H2–O2 from water decomposed by the hot electrode and/or the plasma. Tadahiko Mizuno actually discussed years ago about the anomalous (non-faradaic) decomposition of water from cathodic plasma electrolysis. In some cases it could apparently greatly exceed conventional electrolysis efficiency.
Other authors have observed this effect as well, it turns out after looking on Google Scholar if/when that same paper has been cited.
From the above list from Google Scholar, this paper (and possibly other works from the same authors) could be relevant. It seems that the effect should be able to occur at both the anode and cathode: https://www.researchgate.net/p…faradaic_chemical_effectsQuote
Abstract: Normal electrolysis (NE), at sufficiently high voltages, breaks down and undergoes a transition to a phenomenon called contact glow discharge electrolysis (CGDE) in which a sheath of glow discharge plasma encapsulates one of the electrodes, the anode or the cathode. The chemical effects of CGDE are highly non-faradaic e.g. a mixture of H2 and H2O2 plus O2 each in excess of the Faraday law value is liberated at the glow discharge plasma electrode from an aqueous electrolyte solution. [...]
Some interesting tables from the paper:
This other one is a quite recent paper (published July 2020) from different authors, even: https://iopscience.iop.org/art…149/1945-7111/aba15c/meta
If those bubbles were a stoichiometric mix of (2H2) + O2 then they would be very very explosive when thet see even a sniff of flame. Not a pop, but a bang.
While I do get loud bangs under certain conditions from electrode operation immersed in the electrolyte, the large and persistent bubbles themselves did not seem to be readily explosive.
I couldn't make a test with a matchstick, but an unsatisfactory test with a long lighter (unfortunately out of fuel) appeared to show non-explosive combustion of such bubbles.
EDIT: I made another test with an enclosed cathode. I do not get persistent bubbles with it (they do not accumulate in large amounts and quickly burst), but I have similar problems likely caused by the weak power supply which cannot maintain a constant voltage at less than very light loads. In the video I ramp voltage up, then down until the reaction disappears. Above a certain voltage the gas pockets formed around the cathode detonate immediately and make the reaction look apparently more continuous and stronger, but from RF measurements it becomes more intermittent.
A better/stronger power supply for proper testing at high voltages, where supposedly the reaction can become much more efficient than normal electrolysis, would be very useful.
A possible major advance in how these experiments can be run.
Following suggestions I've read in some mainstream publications, I tried adding a separator between the anode and the cathode. It only has a small hole on the bottom for ion exchange to continue. This significantly increases inter-electrode resistance and stabilizes the plasma.
As a result I could substantially increase input voltage in plasma electrolysis experiments (up to 750–800V) without annoying explosions and/or loud spark discharges.
On top of this, I tried adding at the cathode a few filaments of stranded tin-plated copper wire. The idea was to have a thinner cathode that would be hopefully more efficient in producing the [glow] plasma discharge.
Copper is an easily melted material. It did melt fairly quickly in the tests, but surprisingly it also formed semi-molten balls in the process.
Plasma electrolysis for some reason appears to occur easily with these balls despite the large surface area, at least on top of the water level. Occasionally the formed copper ball can be observed heating up very quickly.
After finishing the test, the melted copper material appeared to be shiny, meaning that hydrogen had indeed been evolved and the material did not oxidize with intense heat.
I've made a video of the process.
- [0:00] Initial test starting with copper filaments
- [2:15] Second long test with the previously melted copper ball
I was trying to form a surface oxide layer on an aluminium anode (in a hope to make it perform better as a cathode. Under certain conditions Al2O3 can glow quite intensely bright as a cathode) when I noticed an unusual luminescence effect. At the same time, a large amount of gas or vapor would be evolved from the electrolyte surface.
The electrolyte was 0.12M K2CO3 in 100 ml tap water; open-circuit voltage was 780V, but decreased to very unstable 240-280V under load in the conditions shown in the the video; this is due to power supply limitations.
A plastic separator placed between both electrodes acts as a barrier to increase inter-electrode resistance.
The video below is only 23 seconds long.
EDIT: here's another, 1:17. It's a thick and already partially oxidized Al foil folded multiple times.
EDIT: below is a gif of the effect, here only transiently occurring, when there is an Al2O3 layer on the electrode as the cathode. I'm not sure if it's just helping the Al burning.
Aluminium does burn very white- like magnesium.
The same reaction does not occur just with clean Al, or at least not as easily (it tends to form spheres as with other low-melting point materials, or slowly oxidize from the heat and fragment into small pieces depending on several variables). My hypothesis was that the oxide would heat up to very high temperatures—possibly in excess of 2000 °C—before the metal starts burning at a fast rate.
EDIT: Here's a photo of one with a melted tip.
Below is a more extreme case of "ball-forming" reaction using copper, although this is probably not strictly plasma electrolysis anymore. Still, hydrogen is evolved in the process and with a sufficiently powerful power supply the cathode (-760V relative to the anode) could be probably immersed completely in the electrolyte (0.05M KOH in 150ml tap water).
I have already showed this type of reaction in a completely different thread some time ago. The cathode ball heats up quickly and cools slowly, and the reaction can be easily maintained as long as the material is hot. If anything, it quickly 'runs away' and the reaction must be stopped to not make the material further recede from melting.
Here is a detail of the smooth sphere formed after the above reaction:
The above ball was made using 2x0.25 mm copper filaments. After more tests I believe that the melting point of copper is significantly exceeded and the glowing ball stays on the wire as a liquid drop, while apparently spinning at the same time. This was earlier on (different thread) likened by Alan Smith to a sort of homopolar motor. I find that as soon as the material melts, it stays hot longer than in solid form, even after power is removed (of course, not indefinitely).
I am considering getting molybdenum (m.p. 2623 °C) or tungsten (m.p. 3422 °C) thin wire for more of the same tests with potentially better materials at higher temperatures, but the latter tends to be very expensive and of the former I have no knowledge about for these applications. Suggestions?
Typically tungsten welding electrodes would be used in these experiments, but even 1.0 mm diameter ones might not perform well in my case due to power supply limitations.
Thin Molybdenum wire is cheap and easy to obtain. It is used for cutting cheese, but more recently for separating the screens of broken mobile phones.
Get a linear tungsten-halogen lamp (they are really cheap these days) smash it. The filament is tungsten ( the lower the lamp voltage,
the thicker the wire) The lead in and filament support wires are molybdenum.
I just ordered some from another vendor, 0.10 mm thickness.
I should have thought of that a few weeks ago when I replaced a small 12V one at home, although probably the filament was too thin.
I'm starting to suspect that there are other non-intuitive factors involved, in any case. I just tried thin guitar string steel wire; it seems as if copper still heats up faster and cools down slower, especially once it melts. If I add a copper filament to the tip of a thin steel wire, it becomes easier to bring to high temperature, as long as there is copper that can easily melt (and causes a green plasma). High conductivity could be involved, or perhaps there is some effect contributed by the molten metal.
EDIT: perhaps the effect slow cooling of copper observed under these conditions (which are not exactly plasma electrolysis anyway but more like heating with atmospheric glow plasma over a liquid anode) is due to recalescence: https://en.wikipedia.org/wiki/Recalescence
From this paper https://doi.org/10.1088/1009-0630/10/1/07 on electrolytic glow plasma discharges the authors have these current-voltage graphs:
Regarding Fig.2a they point out that between points B-C current readings fluctuated significantly, and in D-E greater chemical yields than Faraday's law may be observed. From this it might be inferred that in conventional experiments just using a simple variac+bridge rectifier such region might not be reached. A higher electrolyte conductivity shifts the curve to lower voltages. Point B is roughly where breakdown starts occurring.
I think the current instability in the B-C region is associated with intense RF emission. I could observe in my tests this by dropping the cathode further into the electrolyte, which increased load, causing my small HV power supply to limit current by decreasing voltage. The plasma then turned from metal-ion colored (I was using one single 0.25 mm copper filament, which produced a mostly green color) to a weaker yellow color, became unstable and broadband RF noise increased substantially, even at zero gain setting at the RF receiver.
Although it has been sometimes suggested that such "negative resistance" region is important, if greater (or much greater as reported in some cases) than Faradaic efficiency is achieved at higher voltages, would that really be important?
As a side note: the previously discussed "ball-forming" effect is best achieved at low electrolyte concentrations (I was using 0.012M KOH), which has sometimes been described as "atmospheric glow plasma" in the literature. At higher concentrations (e.g. 0.05M KOH), "true" plasma electrolysis becomes easier and bringing in contact the electrode with the water level causes sharp bangs. Higher concentrations (I tried up to 0.25M KOH) make plasma electrolysis more difficult to control at high voltage and cause rapid melting of the cathode material.
With cathodic plasma sometimes I get fast melting, with the molten material "eating" through much of the cathode wire. In the case below I had two 0.25mm copper filaments heated to high temperature wrapped around a 0.35mm iron (steel) core. I also only used tap water (with some KOH impurities) and 780V (open-circuit voltage; decreases to about 400V with a cathodic plasma without further separation between the electrodes).
Steel should have a melting point of about 1400 °C depending on the alloy, but copper melts at 1084 °C (Wikipedia). Apparently copper was heated sufficiently enough to melt steel.
I wonder there it was some sort of thermite reaction involving Cu instead of Al, and the slight oxide layer on the surface of the steel wire above the Cu filaments. However with "copper thermite" I mostly get search results about Al+CuO thermite, which I also observed sometimes (so far without consequences).
EDIT: a more complete video here.
That sure looks like a thermite type reaction. A table in this reference shows the ignition temperature of Cu in O atmosphere as 1338K, a bit below the melting point. But copper doesn't ignite in normal air, so there must be something else at work in your example.
Slightly off-topic: CuFe2O4 is a synthesized alloy used for organic reduction catalysis. Sound familiar?
Possibly the atmosphere inside the open jar had a higher concentration of H2 and O2 than a normal atmosphere, although the above gif was not made much after the test was started (check out the video I uploaded on Youtube for more context).
There are also suggestions that both H2 and O2 may be produced at the electrode where an electrolytic plasma occurs, although the example above was not strictly one. E.g. from Mizuno:Quote
Abstract: [...] Thus plasma electrolysis may be a better alternative, it is not only easier to achieve than direct heating, but also appears to produce more hydrogen than ordinary electrolysis, as predicted by Faraday's laws, which is indirect evidence that it produces very high temperatures. We also observed large amounts of free oxygen generated at the cathode, which is further evidence of direct decomposition, rather than electrolytic decomposition. [...]
The steel wire onto which the copper filaments were wrapped around is from a "nickel-wound" guitar string set. It should be just high-carbon steel.
I am not familiar with CuFe2O4 catalysts unfortunately. In general it seems that copper heats up quicker than other materials in these tests, so it does not seem an alloy-dependent effect, but I haven't tried a very large material selection yet.
It could possibly also be in part related to this effect:
The rate of cooling of the glowing hot sphere occasionally appears to slow down or almost briefly stop. Sometimes it even briefly heats up again. I think this is called "recalescence" and is a known effect with cooling metals, although I wasn't aware that copper could be affected as well.
After filling the electrolyte (just tap water with some impurities from the anode) with graphite powder and decreasing the amount of water to a very low level as to cause it to rapidly heat up, I could finally obtain a nice and typical cathodic plasma electrolysis reaction using a 1mm-thick steel wire even with my fairly weak HV power supply. Once the reaction started, I could slowly submerge it for 4–5 millimeters.
The reaction appears to be strongly favored by the boiling water. When the water at least around the cathode is fully boiling, a transition occurs and the plasma becomes yellow-colored, probably from Fe ions. RF emission increases at this point. Emissions further increase as the electrode is immersed deeper, to the point where the reaction becomes unstable.
Several regularly-spaced dot patterns also become visible on the cathode surface. It was difficult to take a good photo, but they should be visible in the animation here:
The video below is a combination of two different tests under different camera shutter speeds. Under these conditions it appeared that the reaction could be ran indefinitely without significant wear or melting.
0:00 First test
1:24 Second test with a faster camera shutter speed.
EDIT: a couple photos of the slightly worn tip. The worn area is about 2 mm long.
A few more after cleaning: