Decent cell build....

I finally got fed up with the halfbaked solutions and contraptions that I've been using as an excuse for a cell regarding electrosynthesis of chlorate and perchlorate.
It is time to create a worthy vessel to contain the process better and not stink up the room or cover it in salt residue. Not to mention the zombie outbreak of corrosion that follows...

So.....a "huge" beaker shaped glass vase is chosen as the main vessel for the electrolyte. It is cheap, functional and replaceable.
Now we begin to machine the flanges that shall encase the vase from bottom and top, containing all the goodies inside.
Titanium rod is machined to make the electrode stems....drilled...tapped. Oh it is fun to tap a blind hole in Ti !




It is taking shape. The flanges are made from PP plastic and grooved to fit the lip of the vase against a gasket and also the bottom is sitting in recess for correct centering and stability:





Into the top flange three ports are installed as threaded inserts. For leading away the formed gases....for sampling of electrolyte.....and for holding a glass thermometer in place. The thermometer port has an adjustable ptfe gasket that gently grabs and seals the outer perimeter of the glass tube.
Also proper fasteners are machined and oxide coated to tighten the flanges on the treaded rods:





Almost operational:



The cell was fitted with a slightly but passionately used set of electrodes (MMO/ Ti) for chlorate synthesis. Just some details to take care of before we can fill it up and turn the current on.
Gotta direct the gas flow to outside and find/make a gasket to seal the upper flange against the lip of the vase reliably.
That should prevent the gasses and salt spray from exiting anywhere but through the off gas port and make the process much more pleasant for the surroundings.

We are in business....and the air is clean



I must fit a plastic vessel under the cell to contain the fluid if the vase should break. I can already see it happening....

The contraption has been operating with NaCl solution (1kg charge in the vessel) for about 48 hours now....seems to work quite stable, clean and smooth:








The temperature hovers around 70-75C with no extra cooling or heating and the current is at about 27-30A with 4V applied across the electrodes. Considerable amount of ClO3- has been generated, assessed by the amount of crystals forming when a liquor sample is mixed with KCl solution. There is minimal salt creep from the connections...I trust there would be virtually none if the gas exhaust would not have become blocked during the first 12h of operation. That generated pressure in the cell and forced the liquid through the fittings. I used a too small exhaust tube and tried to lead it straight out into the cold winter weather.....of course the salt spray and water vapour collected into the tube, were forced out to the cold and there they assumed it would be a good idea to block the tube and move no further. I kind of thought that would happen, now it's has been proven. So I just chucked the tube into a bottle of water beside the reactor and called it good. There is virtually no smell at all...just a steady stream of H2 bubbles coming out of the exhaust tube. That can of course be changed into a radical chlorine stench if I should decide to mess with pH control
But naaah....too much hassle, I just let it run it's course and do nothing of the sort.

Finished the first stage of the conversion from Cl- to ClO3- in five days. I suspect it was pretty much done on the fourth day already. Current started dropping off slightly and temps went up to almost 80C.

Cell was dismantled cleaned from salt creep and fitted with a homemade Pt coated Ti anode for the next step of conversion from ClO3- to ClO4-

The anode was made by coating a previously sanded and cleaned Ti sheet with a thin layer of hexachloroplatinic acid solution which was thermally decomposed in a kiln at 350C to yield a coating of metallic Pt. I suspect the decomposition temperature was a bit on the high side, but it worked nicely.



It's ticking along with a quiet fizzing and the temp holds stable at 50C.

Nice work

I do not like the hazard of many amps-worth of hydrogen gas being continuously discharged,
but I do not have a simple solution to offer.

Generally, electrolysis tends to produce various valuable waste gasses,
(hydrogen, oxygen, chlorine ..)
my puny electrolysis experiments are more for learning than producing,
but you may find a good use for by-products ?
(hydrogenate, oxidise, chlorinate .. something, hydrogen balloon ?)
just a thought.

P.S. Keep a look out for a spare/replacement glass vessel

[Edited on 24-2-2019 by Sulaiman]

Quote: Originally posted by Sulaiman

Nice work I do not like the hazard of many amps-worth of hydrogen gas being continuously discharged, but I do not have a simple solution to offer. Generally, electrolysis tends to produce various valuable waste gasses, (hydrogen, oxygen, chlorine ..) my puny electrolysis experiments are more for learning than producing, but you may find a good use for by-products ? (hydrogenate, oxidise, chlorinate .. something, hydrogen balloon ?) just a thought. P.S. Keep a look out for a spare/replacement glass vessel [Edited on 24-2-2019 by Sulaiman]

30A worth of hydrogen shall hardly contribute to a rise in hazard level. Besides I do lead it outside....or at least that was the original intention. But winter conditions got in the way and since it is more of a trickle than an outpour, I just let it rise to the ceiling and out the cracks.

Gathering and storing it is not a viable practice.....there is not so much, it is contaminated by chlorine, oxygen and probably several more things. So of little worth in most instances. And it likes to creep out of containment anyways.

According to my practice the glass vessels last a very long time.....that is if one does not break them
But even in that case they are like 3$ a pop in local "walmart" so not to worry. And there are even specimens about three times the volume of this one available for asocial prices too.

excellent - all covered.

I'm still only at the dreaming stage re. making suitable anodes

Aaaargh.....jesus, what a cluster***k!





The whole thing went to hell on a greased rail. From the beginning of the perchlorate stage I felt that the cell was not acting quite the way it should have. Anode was just holding up too good and I saw no signs of passivation or current dropping within the first five days. Which in all honesty is not a bad thing, had it not been accompanied by absolutely zero ClO4- forming. So I started to ramp up the voltage to about 5V and all that happened was a quickly passivating anode which caused a massive polarisation that was high enough to start corroding the anode current collector. Seems it was composed of a Ti alloy that does not hold up to high anodic potential. The mess of white percipitate on the bottom of the cell is originating from the current collector.
Also the solution in the cell retained stubbornly the faint green hypochlorite coloration during the whole run time. It is noncharacteristic and shows that there is no perchlorate being produced. In the stage where perchlorate is starting to form, the cell liquor becomes absolutely crystal clear and remains so until the end.


Like thusly:



Platinum plated Ti anode coupled with 316 stainless cathode in a tiny test cell on the way to ClO4- completion. The crystals on the cell bottom are sodium chlorate that formed during first stage of the run starting from chloride and dropped out due to supersaturation. By the end of the run they were dissolved and turned to perchlorate.

And a bit bigger setup I tried in a 500ml vessel also starting from sodium chloride solution and going happily way up to perch in a single run with Pt/Ti anode and SS cathode:





I managed to convert in excess of 200g of NaCl directly to perch in the 500ml cell. The solution was supersaturated and the chloride remained on the bottom until it became converted to chlorate and gradually dissolved. So direct conversion of Cl- > ClO4- can be done in a single run with Pt.


The "ubercell" although failed completely on the perchlorate stage and I can not figure out wether it was the previous conversion to ClO3- by MMO that poisoned the process....or the unfortunate lesser grade Ti on the anode stem....or the Pt deposition process.......or just plain bad luck.

Anyways I shall recoat the anode with a new Pt layer and perhaps try the direct approach again starting with PT/Ti and a chloride solution. That gig has worked repatedly, so why not this time.

Well....the anode died officially during yesterday. With zero ClO4- having formed in the cell:






I swapped it out to a new one which I deemed to have failed in the coating process. The surface had been poslished and could not be wetted by the hexachloroplatinic acid. It formed just some droplets, but I chucked it into the kiln anyway:






Into the cell it goes:






And a miracle happens!! Lo' and behold the visual formation of the heavy perchlorate solution around the anode at the lowest voltage setting of 4,2-4,3V:





Can you see it? The haze of heavy perchlorate solution distorting the light around the anode surface and sinking down into the cell.....this is so cool!!

Well at least I hope it is the desired end product bending the light there. And we have definite ozone development going on, it can clearly be detected in the off gas flow.

I discontinued the endless agony of waiting for the endproduct to form. There is clearly some kind of a pollutant in the current cell liqour batch that effectively prevents anything above ClO3- from forming.

The brief successful formation of heavy product solution that happened when I switched the anode, disappeared gradually after half hour of operation and everything went back to mere electrolysis of water as much as I could notice. There was no more ozone generation and no density changes around anode.

I tried to switch off the cell and let it cool down to room temperature to see if it was perhaps the lower temp that triggered the brief formation of perhchlorate. The cell did cool down to 20C during the previous anode change. Well, no success.
I tried to dial up and down the applied voltage to determine if there was a critical polarisation level which the anode has to reach before it can start forming pechlorate (well obviously there is ) but to no avail. Even applying 10V across the cell would not change things.

I just gave up on that run and removed the cell contents which have a clearly greenish blue discoloration to them. Also there is a strong presence of free chlorine and hypochlorite, which should both be pretty much gone at the stage where perchlorate formation begins. At least previous successful small scale experiments have shown it to be this way.
Most of the discoloration surely originates from the anode stem that could not resist the urge to dissolve, but I'm not completely sure it is the only culprit.
I also suspected that perhaps the anode surface texture was unsuitable. Older german literature states that only smooth Pt is suitable for electrosynthesis of perchlorate and plated or deposited Pt fails to do so, as it does not reach the critical polarisation level due to the uneven surface allowing too much charge transfer.
This may be the case, but anodes that I used in previous small scale experiments were pepared in the same way.....from the same source of Pt ....on exacly the same Ti substrate material. And they worked pretty darn good for a disposable superthin layer unit. In fact the first anode that died in this cell was deposited on the same piece of Ti I used to drive the 500ml cell to a complete "endsieg".

The only difference was that first stage conversion was performed by MMO this time and the anode stem in perhlorate stage was made from inferior Ti grade.

I'm kind of reluctant to accuse the MMO, so this leaves the contamination originating from the inferior anode stem. Mostly it should be Ti, Fe, Al, and V compounds that entered the electrolyte and perhaps found a way to fight my cause?

I'm preparing a small test cell to find out if my previous problem was electrolyte contamination related or there is some other factor at play:






Two NaClO₃ solutions side by side:



The yellow one being the last problematic batch with metallic contaminants that originated from the low grade Ti shank.


Some Pb0₂ anodes from previous experimentation. They do shed their skin quickly in a perchlorate cell, but the Pt coat underneath shall prevail for quite some time. So I'll try one of these in the test cell:




Oh yeah.....for those who, like myself, think that constant current mode for a perchlorate cell sounds like a good idea:



That's what's gonna happen when the power converter forces a constant current through a cell with a gradually failing anode....everything overheats, Grade 1 Ti starts to dissolve (take a look at that anode carcass on the right), cell boils dry and if one is especially lucky, bursts into flames.




[Edited on 10-3-2019 by markx]

Yay!! I have definite and massive perchlorate formation going on in the small test cell. Again the formation was delayed roughly by 36 hours. During that timeframe I could not detect any perchlorate type of deposition (sudden and ultrafine) with interaction upon KCl solution. But today I get the characteristic fine deposition of cristals and there is quite a lot.

I used one of the Ti/Pt substrate lead dioxide anodes and NaClO₃ solution from the clean container without the contaminants. This NaClO₃ batch was also formed by the same MMO anode, so I can rule out the chlorate conversion stage of being the source of problems for further conversion.
The cell operates at about 5V and between 1-1,5A of current. The active anode surface area is roughly 8cm², so the current density hovers around below 2kA/m². This be on the low side for top efficacy according to literature, but I hope to stretch more life out of the poor thing this way. It has deteriorated rather healthily already.....the PbO₂ skin fell off during the first hours, after which I filtered the solution to extract the lead dioxide particles. Now I can see also the Pt coating giving up in areas of greater current density. It does not so much passivate, but just dissolves away from the surface. It remains in the solution bound up by some form of a colorless complex. Letting the perchlorate solution stand for weeks after synthesis gives slow rise to a greyish slimy precipitate from the clear solution. I suspect this blob contains the platinum or at least part of it. Might try a regeneration with some aqua regia if I succeed to extract the precipitate at some point. It is really fine and tends to seep through filters.

This type of electrode deterioration is to be expected, but really I see that as a minor inconvenience. The small stock of hexachloroplatinic solution I have got is more than enough to cover a lifetime worth of Ti substrate. One literally needs a single drop to create an anode of the size I have in the small cell now.

About the CC operation mode.....yes the meltdown was caused by lack of failsafes. There sure are power supplies that have the option to preprogram all kinds of limitations, in fact I think the SEPIC converter I used "sort of" has this option available intrinsically, but I found no convenient way to check where the limitation regarding voltage up conversion kicks in at the time, so I just let er rip until the bitter end.




Side by side differences in the formation of KClO₃ and KClO₄ upon the action of KCl solution on cell liquors:



The right side "vial" contains a sample from the electrolyte of the small perchlorate cell and it displays the characteristic formation of KClO₄ in a dense and fine layer of cristals forming instantaneously on the liquid surface where the KCl solution drop hits.
The left side "vial" contains the starting solution of NaClO₃ and it displays the formation of KClO₃ in a slower reaction that produces scattered coarse cristals in the solution. Just a simple qualitative method to assess if an appreciable amount of pechlorate has arisen in the cell...





[Edited on 12-3-2019 by markx]

The small 150ml cell has been ticking along for a week now and as far as I can see it is about as good as it will get. The current flow effectively and sharply drops to a few dozen milliamps if I one lowers the cell voltage to 3,5V or below.





The solution has obtained the characteristic "clarity" associated with last stages of perchlorate conversion from NaClO3 and there already is visible percipitation of the product at the bottom of the cell where temperature is lower. Interestingly enough I observed basically no ozone during the whole conversion run. And the anode is still very much alive. A rather different display of events compared to what I have observed so far....successful though.








On the right one can observe the amount of KClO₄ that percipitated from 1ml sample of the cell liqour. The supernatant liquid (about 7ml) is water. It was added to loosen up the crystals and dissolve any chlorate or chloride that may have been present.

Quote: Originally posted by markx

30A worth of hydrogen shall hardly contribute to a rise in hazard level.

If the cubicle one performs the synthesis in is "tighter than a duck's butt" then for sure it would not be a good idea to play around with fire in there

A rather more serious hazard which actually tickles the hair at the back of my neck a bit is the gas mix contained in the top part of the cell above the liquid level. If under unfavorable conditions the oxygen evolution on anode is increased, then a rather nicely detonable mix shall be contained in that volume. Even more startling is the ozone evolution stage that tends to happen at the end of perchlorate conversion stages. The current collectors run through that volume of gas. If they are undersized for the amount of current that they have to commute or there is a bad connection in there.....overheating.....bang.
I have not actually heard of such a thing happening regarding electrosynthesis of the compounds covered here, but the possibility is there. And I know from first hand experience that even moderate amounts of hydrogen mixtures with that composition can do a lot of bad things if confined in a reactor.

I'm performing a second run with the small cell and a set of anodes I made at different Pt deposition temperatures (pyrolysis temperatures more precisely).

The cell has been running for 6 days now and there are massive sodium chlorate crystals on the bottom. They tend to drop out in an intermediate stage of the conversion where both perchlorate and chlorate are present in a healthy concentration. At the end of the conversion they dissolve again.




The set of small anodes I made to investigate the effect of Pt deposition temperature on the longevity of the anode:




I varied the temperature from 250C - 400C and also varied the number of coats applied : either 1 or 4 coats of Pt.

The procedure that constituted one coating of Pt applied, was as follows:

Grade 1 Ti strips that were previously sanded clean with 400 grit wet dry paper and stored under ethyl alcohol were dried and coated with hexachloroplatinic acid solution. They were just wiped with a drop of the solution and as much as remained on the surface in a uniform manner was left there. Excess was removed. The electrode was then placed on a fixture and into a preheated temperature controlled laboratoory kiln for 10 minutes. After the 10 minutes were up, the electrode was removed from kiln and cooled down. It was flushed with water and wiped with ethyl alcohol to remove any loose Pt particles. After that a "coating" was deemed to have been applied. The procedure was repeated at the same temperature to apply multiple coats.

The set I prepared was as follows:
250C 1 coat
250C 4 coats
300C 1 coat
300C 4 coats
350C 4 coats
400C 1 coat

The cell was loaded with 150ml of NaClO₃ solution and electrolysis was commenced at 4,3-4,6V and a current of 0,7-0,8A. The anodes have a surface area of about 8cm². So the current density is in the range of 100mA/cm². Abit on the low side, but otherwise the cell temperature creeps up too far.
I started off the conversion run with 250C 1 coat anode: it was not durable and perished wihin the first 3 hours of operation due to passivation.
After that the 400C 1 coat anode was put to use and it showed markedly better resistance against passivation. It persisted for four days in the cell, before noticeable passivation occurred. For practical purposes I could have stretched one more day out that one, but I opted to change it to 350C 4 coats specimen.

That one has been bubbling away for 2 days now and for all intentive purposes does not yet show noticeable passivation.

Seems that it makes sense to try even higher Pt deposition temperatures....say in the range of up to 800C. I was afraid that higher temperatures would increase the formation of oxide coating underneath the Pt layer and render the anode less effective, but current results show a reversed trend. At least up to a certain point I guess.

Quote: Originally posted by MrHomeScientist

This is fantastic work; really great to read. Your cell construction looks very nice, and makes me yearn even more for a shop with those sorts of tools. Out of curiosity, what is the end use for all this perchlorate?

Thanks for the kind words! Really, there is no "end use" of any particular purpose for this project. Just doing it to expand the extent of my knowledge sphere. Perhaps I'll try some visco compositions based on perchlorate to yield a clean burning fuse without much soot and dross. As a matter of fact the accumulation of product from the "ubercell" would be a real problem.

Been tinkering with the small cell and higher deposition temperatures involved in the anode making while I wait for my grade 1 Ti rod to arrive, so I can revive the ubercell.

I went ut to 600C for the Pt deposition temperature and prepared a set to test out how the coats perform.







Cleaning of spent anode substrates and storage under alcohol prior to recoat






Heating up the kiln with the stand to hang the anodes from in it.





Applying a thin layer of hexachloroplatinic solution on substrate. Into the kiln for 10min and out they come again...






Final set of 500C 1 coat, 500C 4 coat and the same for 600C.


I did start the testing with the 600C 1 coat anode to see how it comapres to the 400C counterpart that endured for four days before giving up the ghost.
All conditions being reasonably repeateable the 600C specimen endured for 6 days and some exta hours before passivation set in. Not bad....the 400C counterpart stretched 4 days of life out of it. So a day for every 100C raise in deposition temperature




Passivation...only a few spots are seen to bubble on the anode. But the PT coating seems to be mostly there, just it has a TiO2 coat underneath that prevents it from conducting.






600C 4 coats anode goes in and the surface conducts evenly. In fact I had to dial down the voltage to 4,5V as the cell heated up and current creeped above 1,2A at 4,8V. This is too much for the little plastic pot and it shall boil up. Now the 600C 4coat has been ticking along for 3 days at 0,8-0,9A and everything is stable.

The 350C 4 coat anode that I used to finalize the previous conversion run was working without any signs of passivation for 8 days until the chlorate became compeltely depleted and then it abruptly and totally passivated. It happened within one hour of operation and the deterioration was devastatingly complete. It further cements my conclusion that most of the damage to the anode shall happen at the last stages of conversion.

But for the life of me I have not noticed even a hint of real ozone production from the current small cell....that is strange. Because I remember the small cell that I ran with similar anodes (made with not precicely controlled deposition temperature) previously, produced so much ozone that it one could smell it two rooms away.

Quote: Originally posted by hissingnoise

Your amber liquid looks like my filtered cell-solution after a 30 hour run and a half-eaten cheap oiled graphite rod...

Totally different animal though!

Come to think of it.....years ago when I saved this dangerous amber beauty from disposal I thought that I shall likely never find a reasonable use for it. In fact I mostly took it for the fancy bottle.....how wrong I was.

Some update on progress.....I tried to kill the 600C 4 coat anode by letting it run the cell to full conversion and beyond as far as could be perceived. The poor thing was ticking along in the cell at 4,5-4,6V and 0,7-0,8A for three days, then chlorate depletion set in and current dropped to 0,3A. I let the torture continue for 7 more days in hopes of reaching the end for this Pt deposit. Aaaaaand, it does not die, nor wear, nor passivate!
I gave up on that one, dumped the cell contents to collecting flask and charged a fresh chlorate solution into the cell. At 4,4V I get 0,85-0,9A current flowing, depending on how good the alligator clips manage to connect to the electrodes and the whole surface is working like brand new. This is amazing....it means that the anode survived a full conversion and depleted cell conditions without a hint of passivation. That is a first, usually I go through two anodes in a conversion run. The first one tends to die close to 3/4 th conversion and the second one ususally at the very end if chlorate becomes depleted. But the 600C 4 coat survived without a damage and is currently processing a second conversion. I think this might be very close to a sweet spot regarding the Pt deposition parameters. Let's see how far we can stretch this specimen

Methylene blue has arrived.....surely a lifetime supply for pennies:





Strong presence of perchlorate is indicated in the second conversion run of the 600C 4 coat anode



In fact the second conversion run is nearing a rather end stage and the anode shows no signs of degradation. It is running strong at 4,7V and 1A. Defiantely this specimen is the most succesful anode attempt that I've managed to conjure so far.

The 600C 4 coat anode died at last......after accomplishing three full conversions of a total volume of 500ml concentrated sodium chlorate stock solution. Likely even a bit more than 500ml as I added stock to the cell to replenish the losses during runs and it amounts definately more to 50ml excess in total. Did not register the total amount of the additional chlorate stock, but then again I was just trying to find out how long the electrode would last. The anode has been working constantly from the 5th of April until today, 6th of May. Minus the time it was off during recharging the cell inbetween conversion batches, but that sums to a total of few hours at best.

A grand total of roughly 700h of operation in rather loosely controlled perchlorate cell conditions before It gave up the ghost. I think it is a remarkable result for something with such a fragile and thin Pt coating. I must note that the first signs of degradation could be noticed at about midway through the last conversion, so the specimen has been outstandingly rugged compared to previous ones that were deposited at a lower kiln temperature.

I think it is time to execute the same deposition conditions on a larger substrate and try this with a scaled up version in the bigger cell. Shame the Ti rod for the current collector has not yet arrived and I can not fabricate an operational setup, but the active surfaces can be prepared in advance

I shall percipitate the converted solution stock with KCl and measure the total yield. Or try with a smaller sample from the combined stock and extrapolate for the total volume/mass. I really have no immediate need for KClO₄, so might as well keep it in the solution form for any further ideas I might come up with.

The Ti rods arrived, so I proceeded to fabricate a new anode stem for the "ubercell":



It should be at least grade 2 Ti by the looks and feels of it. Machines easily, is soft and pretty much ok to tap. I suspect the previous rod that I used as anode stem was from grade 5 (alloyed Ti) and that material was tough as a coffin nail. Near to impossible for any tapping operations. By extreme carefulness and a lot of prayer I succeeded, but it's not an experience that I would love to repeat. Well, let's see if it holds up in the cell under anodic potential....

As for the active part of the anode I took a spent substrate from last time and cleaned it by sanding. Then deposited 5 coatings of Pt to it at 600C in the lab kiln. Inbetween every coating the anode was wiped clean with a moist paper and then with alcohol until no more loose Pt could be detected on the paper towel. The first coating had the most content of loose Pt dust, but from the third one forward there was literally nothing that could be wiped off:





The Ti plate was attached to the slit end of the stem by compression fitting. It is pretty "goodentoit", but removable upon request and against my intrinsic suspicions this type of connection seemed to work without passivation in the cell. At least it did so with the previous anode stem.


Charged into "ubercell" (the main conversion run with 600C 5 coat anode begins 08.05.19):



[Edited on 8-5-2019 by markx]

Zooo....the contrapulation has been fizzing away on my tabletop for about 5 days now. The anode stem is not corroding, the anode has not passivated and conducts very nicely and we have indication of perchlorate in the cell liqour. Both tests, methylene blue solution and percipitation by KCl show the presence of perchlorate:




I did upgrade to larger cathodes as the stumps I had in there did hinder the process to such degree that I could only get 4,5A through the cell at best. This will not do in terms of time expenditure....it would take forever to convert that load in the cell. So in went the large Ti cathode plates I use for chlorate runs. I also made special Ti "bolts" to secure the plates to stems and a key to fasten them:





That's more like it....10 amps and temperature around 50C. I try to keep it around this mark as it is a safe middle ground. No excessive heating, but still reasonable productivity. The liqour is quite clean as can be seen, so nothing of contaminative character seems to have dissolved into the electrolyte. Seems to work out this time...

It is coming along very nicely. ClO₄- content is growing steadily as recognized from precipitation test results. Ozone can be detected in off gas flow, not much but it is clearly noticeable.


The titianium bolts securing cathode plates to the stems:



Anode edges can be seen to have lost conductivity (the darker rim around anode bottom part):



Not too bad though, the overall conductivity is very good:




The electrolyte is very clear and translucent. It has lost the greenish hypochlorite tinge, a telltale sign that the convesion is going well and has reached a stage where perchlorate content has become prevalent.

There is offensive salt creep down the sides of the cell, clearly the seal at the top of the cell needs improvement, but this is more of a cosmetic issue.

[Edited on 15-5-2019 by markx]

I did measure the KClO₄ yield in the collected liqour from the small 150ml cell conversion runs and it came close to 0,9g/ml.....pretty thick stuff

2 glass vials were cleaned, dried and weighed to establish tare mass. Then 2ml of collected cell liqour was inserted into one vial and weighed to 0,01g accuracy.

Density of the cell liqour was established as 1,5ml/cm³

6ml of saturated KCl soln. was added to the 2ml cell liqour sample to precipitate the product. 8ml of distilled water was added into the vial to loosen up the cristalline layer and to enable it to separate from the liquid phase.

The liquid phase was removed with a syringe and needle to the maximum possible extent. Some of the precipitate was sucked along with the liquid. This was injected into second vial an let settle. From second vial the liquid phase was again removed after the layer settled and then discarded. The main precipitate was once again washed with 2ml of distilled water that was again removed with the syringe.
The vials were carefully microwaved to dry the precipitate and then weighed again.
The collective mass of preciptate from both vials summed as 1,8g.

I also measured the collective volume of cell liqour from the 150ml runs and it sums up to 610ml total weighing in at 913g d=1,49g/ml

The total expected yield when converted to KClO₄ from this amount of liquor would be around 550g. Depending on how much losses occur at purification.
That is the result of a month worth of electrolysis with the small cell running at an average guestimate of 0,6A.

If my very ragged calculations hold any water then this would amount to about 50,smt% current efficacy?

[Edited on 16-5-2019 by markx]

[Edited on 16-5-2019 by markx]

She's clearly cooked and done (08.05.19-24.05.19):








The cell liquid had obtained a deep clarity with almost a vitreous bluish tinge. High refractive index can be observed, characteristic of a concentrated solution. Further torture is meaningless and shall only manifest itselt in electrolytic discharge of water and bringing the anode closer to death.

I carefully suctioned the contents of the cell into a holding canister and discarded about 50ml that contained the slimy haze at the cell bottom.

The contents of the cell weigh in at 3672g by a volume of 2,5l. I visually assess that about 75-80% of anode surface is still active, so it should be good for one further conversion of same magnitude.

Lets find out:

The second conversion is done.....so is the anode. I let it cook until it passivated completely at the maximum of 5,4V from the power supply. Also the thermometer coupling disintegrated where the threading used to be for mounting it to the cell cover. The plastic electrode couplings also seem to turn into slime. Nothing short of PTFE or PE is going to hold up very long in them fumes. But it finished 2 "ubercell" conversions with one anode, totaling guesstimatively above 4kg of KClO4 and that is not too shabby at all!
It is time for an overhaul regarding the couplings and this time I aim for full PTFE set. Also a more effective seal for the cell cover is needed. The makeshift plumber tape wound contraption is not really tight and has considerable salt creep+ some of the gas leaks out. Luckily it does seal against the mist, but evident improvement is needed.

The second conversion was done with the contaminated sodium chlorate batch from first attempt that did not want to form any perchlorate with the old (grade 5?) Ti anode stem. In fact it had formed some perchlorate after all as it turned out. It was faintly indicated with the methylene blue test, but not enough to be indicated by the precipitation test with KCl. By the time I was running that batch, I had no methylene blue, so I could not detect the faint presence of product. I filtered the goop out from the stored solution and gave it a try with the new anode setup. Worked beautifully the second time and it was quicker than the first conversion.
So it seems my suspicions about the metallic contaminants from the dissolved anode stem hindering the conversion were unfounded. I guess the problem was more likely of electrochemical nature (dissolution of the anode stem pulling down the anode potential enough to keep perchlorate formation at a minimum?)....or the chlorate conversion was incomplete to begin with and it just took way longer than I had the patience to endure.

So far, so horrible.....but what the hell am I going to use the product for now?


I replaced the tired plastic couplings with a set I machined from PTFE stock.......should be way more durable:




The second conversion batch weighs in at 4111g at about 2,75l capacity:




A sample of 205g of the NaClO₄ liqour was taken ans subjected to percipitation with saturated KCl solution:




The product was filtered on a glass frit :





[Edited on 3-6-2019 by markx]

[Edited on 3-6-2019 by markx]

The yield from the 205g (135ml) stock solution when converted into KClO₄:

I did spin some visco based on the KClO₄:






I am utterly unimpressed so far.....for something that required so much effort and fun to produce on my tabletop from scratch, it really performs no wonders on it's own. But I guess that has more to do with the preparation (or the lack of) of the fuse composition.

I took a very simple approach: 75/25 KClO₄/C (charcoal)
Ingredients were milled in electrical coffee grinder (obviously in separated turns) and mixed together in a bowl. 5-7% EtOH was added to the mix and it was screened twice through a fine sieve (0,5mm) to yield a slightly granulated powder. It was allowed to dry completely in the sun for 2 hours and then spun into fuse. The fuse was coated twice with diluted contact glue. A standard method I use also to coat BP based fuse. It yields a durable flexible and waterproof product.

Well it does burn.....rather consistently, but :
really hard to light
not very agressive
still a lot of dross (not nearly overoxydised enough to burn away the threads)

I mean sure it was a halfbaked approach, but I expected at least some kind of wow effect
At least it does not stink of sulfur....

Quote: Originally posted by yobbo II

Perchlorate is much more stable than chlorate

I finally managed to devise a way to make the perchlorate work effectively in fuse formulations: the key is in preparation of the mix. It absolutely needs a very intimate contact with the fuel, just like black powder based formulations.
There are two major ways to accomplish this in an amateur setting:
a) (ball) milling the mix- not viable for perchlorate formulations (at least my selfpreservation instinct does not allow it)
b) impregnating fuel by a perchlorate rich solution to form crystals in and around the fuel particles

I chose the second option. I made a small 2g sample of 75/25 KClO4/C, but this time I changed the preparation process to try out a "vaccum assisted preparation" idea:



Ingredients were weighed into a small glass dish and 0,5ml of azeotropic EtOH +1ml water was added. The alcohol serves the purpose of lowering surface tension to yield better wetting effect of carbon.

The ingredients were mixed with a spoon until uniform (a few minutes) and compacted into a solid cake. The consistency of the mix was like damp dirt so it stuck together nicely.

After that the sample was placed into a vaccum chamber and evacuated for 30 minutes at room temperature (20C). Then the sample was taken out, the cake broken up into loose rubble and put back into the chamber. Another session of 30 minutes under reduced pressure and slight warming at 40C yielded a semi dry granulated product. It was still damp, but good enough to test for burning speed.

A small lump of the prepared mix was collected onto a stainless spoon and ignited with a butane torch: the sample burned with satisfactory violence and was energetic enough to flying the burning lump off the spoon. Definitely the uniformity and vigor of the burn has been improved by the vaccum assisted preparation process.

The slightly damp mix was granulated by forcing through a fine sieve and let completely dry at room temperature and atmospheric pressure (24h).

The sample was spun into a lenght of fuse after it had been dried:







It works just beautifully now! A nice hot, energetic burn that proceeds at an even pace with a sharp long flame jet and a powerful hiss. I'm very satisfied with the outcome.





[Edited on 20-7-2019 by markx]

Quote: Originally posted by underground

Markx i got an MMO electrode and i want to spot weld to it a titanium bar for delivering the current. I have not any spot welder tho and i was thinking to use some electrolytic caps in parallel, charge them, and spot weld it. What do you think, is it going to work ? I have seen some videos that some doing it with a super-capacitor. [Edited on 7-8-2019 by underground]