I have been pondering anode materials for the FFC Cambridge Process. Graphite or other forms of carbon are not suitable as they are entirely consumed
with release of CO or CO2 as opposed to oxygen. At the process temperature of 900°C, probably no metal would hold up for long.
A document, here, suggests a tin oxide based material. I am somewhat confused as to how this operates, as SnO will readily oxidize to SnO2, but SnO2 would
tend to decompose at the operating temperature if power is not applied. The transition between the oxides would seem likely to ruin the structure the
anode. I am not sure entirely how such an anode would be constructed, but it would possibly involve an indium oxide base layer as is used in some
commercial electrodes. This could be electroplated with tin and then oxidized by anodizing. Another possibility is to electroplate tin oxide by
anodizing in tin nitrate, similar to the method for making lead dioxide anodes. Either way, a base layer would likely be necessary to protect the
base metal from oxidation by SnO2.
I am unsure on any alternatives.not_important - 3-11-2008 at 06:52
Hard to tell what they are using, and there's no experimental detail
Tin dioxide is the standard resistive heating element for doing precision glass melts; it's used much like silicon carbide, sintered into rod form
with electric contacts brazed or pressed on at the ends. It's likely the "tin-oxide" of the document sited actually refers to
SnO<sub>2</sub>.dann2 - 3-11-2008 at 15:37
Quote:
Originally posted by kilowatt
Another possibility is to electroplate tin oxide by anodizing in tin nitrate, similar to the method for making lead dioxide anodes.
Hello,
Just wondering how that is done. Tin Nitrate does not really exist as such.
Cheers,
DANN2kilowatt - 4-11-2008 at 01:20
I think it does in very concentrated nitric acid. It precipitates as SnO2 on contact with water. Probably not very practical though.
It might not be too difficult to make a sintered SnO2 electrode from fine clay-like powder. Its melting point is low for an oxide; most sources say
1127°C.
[Edited on 4-11-2008 by kilowatt]dann2 - 4-11-2008 at 01:43
There is also a pile of SnO2 anode related stuff in the 'More on PbO2 anodes' thread.
SnO2 is usually made by heating Stannic Chloride. This if for thin film anodes (ie. coated on Ti or other substrate).
I though Tin Oxide was an extreme refractory.
Dann2watson.fawkes - 4-11-2008 at 04:16
SnO<sub>2</sub> used as electrodes is typically doped with Sb or As to enhance its semiconducting properties. Sintering aids are CuO
and/or ZnO. Not the easiest small-lab manufacturing project. Rather than build, buy. not_important posted a couple of links to manufacturers. My
information is from <i>Electroceramics</i>, by Moulson and Herbert.watson.fawkes - 4-11-2008 at 04:36
Frustratingly, I can't seem to find consistent information on the melting point of SnO<sub>2</sub>. The WebElements page lists one at 1630°C. Certainly the Wikipedia entry of 1127° C is probably wrong, given that this material is used
at temperatures significantly above that. (That figure also appears in MSDS pages, which is where, I'm assuming, the Wikipedia value came from.)ycheff - 4-11-2008 at 07:19
Chapman and Hall data base also gives 1630 for mp and 1800-1900 for bp.kilowatt - 4-11-2008 at 09:04
Quote:
Certainly the Wikipedia entry of 1127° C is probably wrong
At least 4 different MSDS's state 1127° C, but I have seen 1630°C a few times too. I just went with the one that was seen more.
About buying the anodes: in theory the undoubtedly high cost might not be such an issue for an already expensive and high productivity cell. What I
don't like is the low surface area that all the commercial anodes seem to have, possibly not suitable for the high current (several kA) of even a
mediocre sized FFC Cambridge cell. I wonder they are very porous, as this would increase the surface area.
The film deposition methods using SnCl4*5H2O and SbCl3 look like the most straightforward, but I am confused on one point about these. When heating
the coating to high temperature to oxidize SnCl4 to SnO2, how is it that the majority of the SnCl4 does not boil away? It seems unlikely to me that
it would remain well enough to form a coating on anything by carrying out the process in an open system.
[Edited on 4-11-2008 by kilowatt]dann2 - 4-11-2008 at 15:06
Hello,
Perhaps some of the Stannic Chloride goes off as vapour before it converts to Oxide but the procedure definitely works.
I have no idea if this type of anode is suitable for what you want to do.
The prodecure below if from a guy who made lots of anodes at college.
(leave out the bits regarding PbO2)
This is my procedure to prepare Ti/PbO2 anode with SnO2+SbOx(Sb2O5+Sb2O3) inter layer.
The Ti plates (5 × 1 × 0.1 cm) were used as substrates. Prior to using as substrates, they were polished with 320-grit sandpaper, degreased in 40
wt% NaOH solution at 80C for 2h, etched in a boiling 10 % oxalic acid solution during 1 h, after that, they were rinsed with distilled water and dried
naturally. (If you don't have oxalic acid solution,sometime concentrated hydrochloric acid also works, you should see a purple solution during the
boiling Ti plate, this step is very important, only via this you can remove the TiO2 insulting layer,Remember: do not use raw water in this step, a
distilled water is preferred.
The SnO2+Sb interlayer was prepared following a standard thermal decomposition method and the procedures were as follows:
A solution containing SnCl4:5H2O and SbCl3 was dissolved in a mixture of n-butanol(or ethonal)+HCl.(eg. 4g SnCl4:5H2O + 0.2g SbCl3 + 9ml n-butanol +
1ml HCl solution, you can made about 5 anodes, the best mol ratio is SnO2: Sb=100:4~8). The precursor solution was distributed onto the pretreated Ti
plates by brushing. The solvent was dried in air or a oven(90C,a hair dryer also ok), and the electrode was introduced into an oven at 450C for 10 min
for the decomposition of the salt and the formation of the metal oxides. This process was repeated for 15-20 times. A final annealing of the electrode
was performed at about 500C for 1h.
Anodes like this are sometimes used for wastewater treatment.
Maybe for SnO2-electrodepositing the sulfate might be used instead of the nitrate (really not existing ?),
==> since that would be an analogy to the PbO2-manufacturing within H2SO4; only that SnSO4 is water-soluble, while PbSO4 is not; thats why with
PbO2 the nitrate has to be used ... ?
Also: Usually the SnO2-electrodes are sprayed, but this is because it is applied as one of the few transparent conducting materials, at
solar-cell-surfaces etc. . Since the substrates are not conducting, the SnO2 must be applied another way (not anodizing).
But anyhow a good description of making SnO2-layers might be interesting, because I since a while like the idea of trying some solar-cells (CuO is
said to work at 2 % efficiency,; with In one could reach 10 % by CuInS2; maybe other materials can be used too, need to be semiconducting (like many
sulfides are))
[Edited on 4-11-2008 by chief]bereal511 - 5-11-2008 at 11:41
kilowatt, I want to know what spurred your interest in the FFC Cambridge Process.
I could only imagine that a SnO2 based electrode would not hold up in the molten chloride bath for a long protracted time of cell operation (perhaps a
eutectic would be necessary to retard erosion at lower temperatures). I constructed a rudimentary cell based on the FFC Cambridge Process and I have
to say that although graphite erodes, it definitely works and the cost of replacing graphite certainly doesn't exceed the energy, time and cost of
making SnO2 electrodes that may or may not be "permanent" in the sense that they would never have to be replaced.
What are you hoping to produce from the process, just curious?kilowatt - 5-11-2008 at 12:46
I had read about the FFC Cambridge process for titanium production, then a suggestion by not_important in the magnesium thread got me thinking about
the potential it has for refining all manners of metals and semi-metals. I am looking to experiment with the feasibility of refining such various
metals and semi-metals such as mentioned in my FFC capabilities thread with this type of calcothermic reduction cell, and put to use what I find.
What temperature did you run your cell at? Did you use a eutectic? The documentation I have seen indicates that graphite is consumed entirely, in
other words practically no uncombined oxygen is produced when it is used as an anode material. Are you sure this wasn't the case with your cell too?
The cost of replacing graphite in this case would be VERY prohibitive if we are talking about large runs. It was also indicated that the SnO2 anode
showed very little erosion after a 12 hour run. After all, what would it be oxidized to?bereal511 - 5-11-2008 at 16:25
Oh yes, the graphite was certainly eroded to hell to say the least. But I have been experimenting with small runs with boron reduction so it has not
been too much of a problem for me. I suppose you are looking more towards a large scale production.
I did not use a eutectic and ran the cell at approximately 950 degrees Celsius. It would be very interesting to see if a SnO2 anode could be produced
at the home-chemist level, and I am all for it especially now that you mention there was little erosion for 12 hour runs. Well, after looking over
this thread, I am going to experiment with SnO2 electrodes to see what I can get out of it.kilowatt - 5-11-2008 at 18:53
I'm curious, how did the boron extraction go? This is one of the things I want to try with my cell. Also how did you apply it to the cathode? Did
you just dip it in fused boric oxide? I wonder if adding soda or another additive to the boric oxide would allow the use of granules or powder (thus
much more porous and effective than a fused mass) with a melting point above the cell operating temperature. Sodium should be distilled out while the
boron remains at the cathode.
[Edited on 5-11-2008 by kilowatt]bereal511 - 5-11-2008 at 19:17
Not very well. There was a ton of carbon debris and the cell body (a titanium container, I think it was alloyed with something that corroded it far
too much to be pure titanium) added a lot of oxidized junk into the chloride melt, so it was hard to determine whether I had boron or not. I'm
working on modifying the cell sometime this weekend and giving it another run.
I thought about using sodium borate, but I couldn't figure out a way to contain the sodium if it distilled off and it would have been a severe hazard
to myself as my cell isn't particular well constructed in general. I just added the powdered boric acid first, heated the cell until the boric acid
dehydrated/melted to boron oxide, then added the calcium chloride and heated it to the stated temperature. I wish I had some welding gear so I could
construct the cell with some steel but at the time being I neither have the means nor the finances to undertake a total reconstruction of the cell.
Meh, you do what you can with what you have right?kilowatt - 5-11-2008 at 19:23
What was your cathode and how did you keep the oxide on it? I planned to use changeable cathodes with maybe a fine steel screen, but have a separate
cathodic protection supply connected to the container which would run at low power and should eliminate corrosion. The close proximity of the
cathodes to the outer wall may be enough for cathodic protection. Doubling the container as the cathode would be ideal but that would not be a very
versatile or efficient setup in my opinion.bereal511 - 5-11-2008 at 19:30
Ha ha, funny that you say that, because I used my container as the cathode. It seemed easier at the time, but that was my faulty assumption; I'm just
lucky the chloride melt didn't get into the heating elements or the refractory. From my understanding of the process, the oxide-to-be-reduced doesn't
need to be in contact with the cathode per se, it simply needs to be in contact with the dissolved calcium in the chloride melt. But maybe that's why
my container corroded to hell. I'll try using a separate cathode in this next run just to see if anything actually happens.kilowatt - 5-11-2008 at 19:36
Yes, it is true that the oxide does not need to be in electrical contact with the cathode (it's often an insulator anyway), but I believe it helps if
it is in close proximity, because it is more calcium-rich near the cathode. The cathode (and thus container if you used it as such) should be
protected from oxidation or dissolution in the electrolyte by the calcium metal that is formed at its surface. At least this would be the case if
your container/cathode was not very miscible with calcium, such as I understand is the case with steel. Perhaps this is not necessarily the case with
titanium? What sort of voltage/current and current density did you get? My understanding is the experimental FFC Cambridge cells run about 2-3V
(preferably controlled during the process so it is decreased as the run progresses) and about 1A/cm^2.bereal511 - 6-11-2008 at 17:08
I was working with about 5 volts and low current density (100 mA/cm^2), basically all that I had around. I wasn't sure (and am still not) where I can
get a dedicated power supply with controllable low voltage and high current.
What kind of power supply are you hoping to use and what material are you using for the cathode? SnO2?tentacles - 6-11-2008 at 18:42
bereal: Look on eBay for power supplies.. A good set of search terms is "kepco" "lambda EMS". I picked up a 5v/200A CC/CV supply (no gauges though,
but that's easy to fix) for $50 shipped. And it's efficiency rating is over 75%. Be sure to search under power supplies.. I suggest doing the lambda
search from the start page, and then going to the category from there. Otherwise it's a PITA trying to figure out ebay's categories.
What part of manitoba are you in? I'm in Winnipeg.bereal511 - 7-11-2008 at 11:04
Thank you so very much tentacles, I've been looking around for something like this for a while. I'm actually in Manhattan Beach, California, I think
I'll change that for unambiguity .kilowatt - 7-11-2008 at 13:15
I intend to use a large set of automotive alternators (these can be had quite cheaply) for my large cell in the future, and a 100A battery charger I
have for small experimental cells. There would be a set of alternators for each individual cathode, so that each cathode's voltage can be controlled
independently via the alternators' exciter current. I am considering running the alternators off a heat engine such as a Stoddard engine (which can
be made from a pair of modified 4 cycle engines) which operates from heat exchangers on the cell's inner wall, so that the entire apparatus runs on
gas burners or another heat source. The maximum Carnot efficiency of such a setup would be about 75%, but quite a bit lower in reality. This way
much of the waste heat that results from the cell's low current efficiency would be used to power the cell's electrical supply, which would operate
entirely on fuel or any heat source. If properly located, the heat exchangers should encourage convection currents within the electrolyte too,
possibly increasing cell performance.
SnO2 cathodes? The entire point of an FFC Cambridge cell is to reduce oxides to metal. I plan to use steel or inconel cathodes and an SnO2 coated welded steel or cast monel anode.bereal511 - 8-11-2008 at 17:40
Hahaha, I mean anode* for SnO2.kilowatt - 8-11-2008 at 22:11
I did some experiments tonight with reducing silica sand in a small (roughly 4kg of CaCl2) test cell. At first I tried containing the silica in a
stainless steel mesh basket around the cathode. The silica appeared to dissolve quite readily, and when I applied power from a battery charger set at
6V, the initial current was about 40A. After a short time the current dropped off to zero. After shutting off the power and letting the cell rest
for a few seconds, the initial current was again about 20A, but dropped to zero within barely a second. This could be repeated over and over. I
concluded that perhaps silicon was plating out onto the cathode and forming an insulating layer, which would dissolve once power was removed. I
attempted to get around this problem by doping the silica with boron from boric oxide, added via boric acid directly into the melt. For this second
attempt I added roughly 100g of silica and 50g of boric acid directly into the melt and used a free hanging cathode. Upon addition of the boric acid,
the melt outgassed HCl and emitted a brown smoke until all the water had been driven off leaving boric oxide. The smoke evolution did not stop even
after nearly an hour, but I attempted the run anyway once the bubbling had stopped. I am not sure what reactions were going on in the melt, perhaps
some free calcium reducing the oxides or also reacting with air or water from the boric acid. The results of the electrolysis were the same, only
this time conduction ceased even faster upon applying power - only a brief spike of about 20A before the current would drop to zero, and this could be
repeated by letting the cell rest for about a minute. Keeping the cell operating for some time did not seem to develop a significantly thick layer of
silicon.
If my assumption that a film of silicon forms on the cathode and stops conduction, why would doping it with boron not make it more conductive? Would
a high level of a metal such as iron from iron oxide help? I did not think silicon was even soluble in fused CaCl2. Are there any other explanations
for this phenomenon?bereal511 - 9-11-2008 at 11:46
Hmmm, would you mind taking a picture of your setup? Have you tried another oxide of something more conductive to see if in fact, the metal was
actually plating onto the cathode? I'm not sure how you put together your cell, but perhaps the chlorine evolved at the anode is forming layers of
insoluble (in calcium chloride) metal chlorides. What did you use for your anode?
The idea of silica dissolving calcium chloride seems, well, counter-intuitive. I also was not aware that that was even possible, considering there
have been experiments done with the FFC Cambridge process to reduce silicon dioxide to silicon, as I'm sure you've come across.
If you are doing another run, you might want to try adding a small portion of calcium oxide to the melt so you don't have to produce an initial phase
of chlorine before having the cyclic CaO <-> Ca reaction occurring after oxide reduction. I don't know how much that would help, but I will
have to try it myself later in the week. Still tweaking up my old cell.kilowatt - 9-11-2008 at 12:02
I have a picture at http://chrisf.4hv.org/projects/chem/FFC_Cambridge/HPIM1471.J... which I have not bothered resizing for the forum. I have run MgO in this cell,
and I got notable droplets of Mg metal at the cathode. I ran it until the Mg was depleted. As for chlorine, there never was any produced by this
cell. The calcium chloride I started with was low grade and has some water content, to partial hydrolysis during fusion actually produced some lime
(you could smell it outgassing HCl when I was fusing it for the first time). It would have been quite notable had any chlorine ever been evolved.
However I did use a graphite anode. A jacket of CCl4 would be rather insulating. A jacket of CO or any gas also would be however. The anode gas
production dropped to nearly nothing when the current went low. Perhaps a small anode area was the limiting factor here?
I cannot think of anything else.
Edit: This appears to be an anode related problem, as when reversing the electrodes to make the steel rod the anode, the cell had fine conduction,
drawing about 30A and depositing lumps of Fe/Si/B alloy onto the cathode. I then changed out the graphite rod with a new one and used it again as the
anode. Again I had good conduction for a short time and got good deposits on the cathode, but it soon ceased.
[Edited on 9-11-2008 by kilowatt]
12AX7 - 9-11-2008 at 21:40
Sounds like the anode is being passivated then. Boron carbide perhaps?
Interesting structure on that, er, is that the anode, or the anode when you used it as a cathode?
Timkilowatt - 9-11-2008 at 22:08
That buildup was the cathodic deposit when I used the original anode as the cathode, and vice-versa. It was more iron-rich than the similar deposit I
got on the steel cathode when I switched them back around, but otherwise looked very similar.
This conduction loss happened when I was running straight silica too remember. I don't know how the anode could become passivated with silicon or
boron carbide, because silicon and boron are attracted to the cathode, and the anode is obviously a highly oxidizing environment.12AX7 - 10-11-2008 at 13:07
Silicate and borate anions, I'm thinking. Both are present as neutral salts (SiO2, B2O3), but there will probably be some ionization, no? I wonder
how much though...
Crude SiC should be conductive anyway, so that doesn't help too much. Maybe SiO2 is simply accumulating, gumming it up?
Timkilowatt - 10-11-2008 at 15:59
I had the melt saturated with silica, the excess existing as a glassy phase with boric oxide at the bottom of the vessel. Both the anode and cathode
were directly above this, and perhaps dipping into it (you could feel thick glass at the bottom if you stirred the electrodes). Do you find it likely
that running the cell at less-than-saturated with silica would help, or perhaps just not having the anode so close to the reserve silica? I can't
think of much else that could be done. Recall though that the steel anode appeared to work indefinitely (though it was losing iron into the melt
which in turn ended up alloyed at the cathode).kilowatt - 7-2-2009 at 04:16
I was wondering if bereal511 ever continued his experiments with the cell. I would be most interested to hear more if anyone has continued dabbling
with this.
I have been preparing to make an improved small cell out of some stainless steel which I recently obtained, which is electrically heated and has a
fully functional design (cooled gas/light metal traps, bottom drain, thermocouple, quick change electrodes, argon feed, etc) just as the large cell I
have planned only smaller. The melt is not very exposed to air in this design which, according to what I have read in articles on the process, should
help a lot with corrosion, and should eliminate the continual HCl outgassing problem I have experienced before.
.
I plan to use steel or inconel cathodes and an SnO2 coated welded steel or cast monel anode.