I may have come up with a relatively straightforward process that unites the production of sodium hydroxide, hydrochloric acid, sulfuric acid, and
calcium silicate, all from extremely cheap materials. While it is by no means something that can be cobbled together in a day, it should alleviate
the need for the contact process or lead chamber process, HCl furnaces, and other difficult or finicky types of processes.
The process starts with two fundamental pieces of equipment: a Castner-Kellner amalgam cell, and a high temperature reduction furnace. An air
compressor, glass reflux still, and a few simple reaction vessels are also needed. The real heart of the process is the repeated reduction of cupric
chloride to cuprous chloride with sulfur dioxide.
Gypsum is reduced with coke or coal and silica sand at high temperature to produce SO2:
2CaSO4 + 2C + SiO2 --> 2SO2 + 2CO Ca2SiO4
SO2 is cooled in a heat exchanger and then easily liquefied by a compressor to separate it from carbon monoxide, which is used as a fuel to help heat
the coke reduction furnace. The SO2 is then bubbled through a solution of CuCl2 to produce the acids and precipitate CuCl:
2CuCl2(aq) + SO2 + 2H2O --> 2CuCl + 2HCl(aq) + H2SO4(aq)
An amalgam cell produces Cl2, H2 and NaOH. H2 may be stored during this prolonged step (possibly as a metal hydride) and then combined with CO from
the coke reduction to help fire the furnace, or even used as syngas. Wet Cl2 from this step is combined with CuCl from the previous step bath:
Cl2 + 2CuCl --> 2CuCl2
CuCl2 is kept for use in the reduction step with SO2. Alternatively the aqueous H2SO4 produced earlier can be reacted with half of the CuCl2 produced
in the chlorination step:
CuCl2 + H2SO4 --> CuSO4 + 2HCl
The CuSO4 can be thermally decomposed to SO3 and CuO. The CuO can then be reacted with the HCl from the prior step to regain CuCl2 for the reduction
step.
When it is finally time to distill the acids from the SO2 reduction bath, waste heat from the next coke reduction may be used to help power the still.CyrusGrey - 17-4-2008 at 21:00
Seems like something useful for an industrial process, but far more difficult than the lead chamber process for the hobby chemist.
For one thing wouldn't it be easier just to burn gardening sulfur to get your SO2, and cheaper considering the equipment involved.kilowatt - 17-4-2008 at 21:47
I know of nowhere where I can buy sulfur for anywhere near the price of gypsum (I can get scrap drywall for free, and it is dirt cheap by the bag) but
that would indeed be much easier, and would require neither coal nor silica nor additional heating fuel. I was not aware there was even such
"gardening sulfur" in existence until now, and I certainly have never seen it in any stores. Coke reduction of calcium sulfate is not too hard
though. I have a simple propane fired furnace that can do it easily. Whatever the SO2 source, originally I had planned to carry out the Contact
Process (I even have some V2O5), but the copper chloride reduction looks much easier on a hobby scale. I'm more concerned with this key reduction
step, though yield should not be an issue since it is a cyclic process. I suspect the greatest challenge will be keeping a high efficiency. I don't
think this process would be particularly good industrially, otherwise I'm sure they would be doing it since these are all routine and well-known
reactions. It just looks convenient to me, given what I have.CyrusGrey - 17-4-2008 at 22:00
Its cut with clay, but that should be fine if all your doing is burning it.
Also, isn't reaction 2 going to proceed more like this:
2CuCl2 + SO2 + 2H2O --> 2CuSO4 + 4HCl12AX7 - 17-4-2008 at 22:04
How do you prevent reduction to CaS, or considering the SiO2, free sulfur vapor? Stoichiometry I suppose, and SO2 might be preferred over S.
CaSiO3 melts at a rather high temperature. You'll need a high refractory -- and disposable -- retort to handle that. The cost of fuel and retort
alone seem to discourage this process, even before counting the steps in the cycle.
Your process reminds me of a former process to produce chlorine: 2HCl + 1/2 O2 --CuCl--> Cl2 + H2O
Timkilowatt - 17-4-2008 at 22:12
Copper(I) chloride precipitates out when copper(II) chloride reduced by sulfite, but I am aware that it's somewhat soluble in acid solutions and thus
not all of it will come out. I would think the sulfuric acid would eventually react with it if it was left in there, but I don't know the details. I
would have naturally thought it would just form copper sulfite or bisulfite, but apparently not. There are several references to the reaction on the
internet; I originally found it in the wikipedia article on CuCl. Apparently this is how CuCl is commonly made.
About the SO2 via CaSO4 reduction; CaS should react with CaSO4 to give off SO2. The stochiometry can be adjusted to favor any calcium silicates and
either CO or CO2 along with SO2 or CaS. CaSO4 reacts with CO as an intermediate to give off SO2 and CaO, which then reacts with SiO2. The reaction
does not have to take place above the melting point of calcium silicates; it is just a solid byproduct. The chemistry going on here is not too
different from that in a cement kiln. Another example of these reactions can be found in US Patent 5066474. They are definitely no myth, and do not
require any particularly special vessels nor unattainable temperatures. I find the reactions rather confusing and hard to predict exactly, though. I
will do some experimenting as soon as I can, but I would have a hard time believing that the reduction would not work long before hitting the melting
point of a steel retort. If any CaS is left in the retort, addition of water would release H2S which can be burned to give SO2.
Of course, admittedly, combustion of raw sulfur is a much more direct route. I have also personally obtained SO2 from the smelting of lead sulfate
during the purification of scrap lead. I had always been hurting for a good method of recycling this SO2 back to sulfuric acid before, and was
looking to the contact process before I came to this process.
Quote:
Its cut with clay, but that should be fine if all your doing is burning it.
Any idea how much actual sulfur is in it typically? I need to know if this is will be economical. I have found now that some of these sulfur
formulations contain other polymers and binders too, but haven't been able to find the proportions. Really it is almost inconceivable that an
additive would pose much of a problem for my SO2 source since the SO2 will be liquefied and essentially distilled before going into the rest of the
process anyway. What matters is the cost of the sulfur content. Any pure sulfur I have found is far too expensive, and would equate to nearly $10/L
of sulfuric acid even with 100% yield.
Since my heaviest usage of sulfuric acid is for scrap metal refining, particularly lead, and the smelting processes typically liberate SO2, this
process should be a good way to reclaim the acid from that, if anything.
[Edited on 18-4-2008 by kilowatt]not_important - 18-4-2008 at 08:07
What is known as "garden sulfur" is really many products. Sometimes it is admixed with clay, usually as a coating on the particles done to reduce
flammability - about 10% by weight; other times it is a crude sulfur mixed with gypsum and perhaps clay/earth. This is usually in rough particle
form.
It might have wetting agents added, particularly if it is "dusting sulfur" - a fine powder. If it is in the form of prills it might have
polymers added for slower release, SFAIK this is uncommon. Sulfur plus polymer mixtures are used with water soluble fertilizers to slow their
release, but that is going to be sold as fertilizer rather than simple sulfur.
Agricultural sulfur, which may also show up as garden sulfur, from the LO-CAT process is about 1/4 water, and often is solid as broken filter cake.
It can be melted to 99+ percent sulfur, or dried in a flow of warm dry air to a lower water content powder. As sold it is difficult to burn, the high
water content is left in it for that very reason.
Repeated washings with water will remove most of the wetting agent, enough so that H2S formation on heating isn't a problem. Washings will also
remove at least some of the gypsum, if any.
Most garden sulfurs are labeled as to their sulfur content, usually wettable ones are so labeled. The clay coated ones will be about 9/10 sulfur.497 - 23-4-2008 at 18:35
The "garden sulfur" I have seen is very cheap, I don't remember exactly, but it couldn't have been more than a couple dollars per pound. It has some
impurities, but it burns fine and is at least 80-90% sulfur.
You seem to be wanting it for a fairly large scale setup, it might be easier to rig up a tube furnace with some V2O5 and run it continuous. Otherwise
you're going to be working with an awful lot of Cu and HCl.kilowatt - 14-5-2008 at 10:13
I have begun to experiment with the reaction of calcium sulfate with carbon and silica at high temperature. I don't have any coke so I have been
using graphite powder (which I have a fair amount of) for the time being. I weighed the materials in an anhydrous state with stochiometry according
to 2CaSO4 + 2C + SiO2 --> 2SO2 + 2CO + Ca2SiO4 and added water to set the mixture into solid pellets with a uniform composition. I then fired the
pellets in a small makeshift furnace to check for SO2 emission. Even at over 1400°C (the steel cup I was doing this in melted) the reaction
proceeded apparently slowly, with little detectable SO2 odor. The exhaust from the furnace was very irritating (SO2) but it didn't really fill up the
area. I was away for most of the time it was running to avoid exposure. After cooling and adding the crushed pellet to water, I tested the solution
with litmus paper and got a pH of 12-13, positive for calcium hydroxide. This could indicate either calcium orthosilicate as hoped for, or a calcium
sulfide to calcium oxide route which is equally suitable but has different stochiometry. Prior to the reaction, the calcium sulfate had a neutral pH.
No noticeable H2S was produced when I added water to the finished pellet, so that's negative for calcium sulfide. I was going to weigh the product
to determine if the weight would correspond to the expected reaction, but it was all wetted with iron from the melted crucible, and some of it had
exploded from heating it too quickly, so that was not possible this time. I will try again soon in a refractory crucible and keep track of the
weights.
I'm not yet sure exactly what temperature is needed to make the reaction proceed. I will need to test it on a larger scale in a steel retort and see
if I can get anything to happen before it melts. If it doesn't work I will have to make a refractory retort, or possibly just use my alumina crucible
and find a way to collect and separate the total exhaust from the furnace, maybe with a refrigerated condenser for SO2.12AX7 - 14-5-2008 at 12:52
Sounds like decomposition to me. SiO2 and CaO need some pretty good heat to cook down. CaSiO3 is insoluble; I'm not sure if Ca2SiO4 exists and what
its properties are. There are other sulfates that decompose more readily, even MgSO4 (circa 1100C). Did your carbon burn off? In anything less
than an extreme reducing atmosphere, you will need a substantial excess of carbon for this sort of reaction. A more reactive form such as hardwood
charcoal may also be more suitable. You may also try fluxes such as Na2SO4, NaOH or Na2CO3 or Na2SiO3, maybe even Fe which has some solubility for C
(making a solid state reaction?).
Using only 2 moles water in this stage would likely liberate the HCl as a gas.
Seems like that even if considerably more water is used the HCl will still tend to gas off in the presence of H2SO4.
Just wondering if you have considered this.12AX7 - 14-5-2008 at 14:10
Idunno, both the products may be pretty soluble in the H2SO4 actually. Most of the HCl would leave, yes. For that matter, Cu(I) tends to
disproportionate in solution, so you'd be left with HCl from that, and Cu(II) in solution or precipitated (white CuSO4). Concentrated H2SO4 would
oxidize the copper, giving off SO2. Well, that would happen anyway, so the reaction kind of wouldn't really proceed anyway- not in concentrated form.
But...
CuCl2(aq) is indicated, so I think dilute solution is implied. In that case, HCl would remain dissolved, CuCl precipitated (assuming the chloride
concentration remains quite low) and H2SO4 quite well dissociated.
I wonder...
Would SO2 + 2 CuCl2 <---> SO2Cl2 + 2 CuCl proceed to the right or the left? Which is in effect asking, is CuCl2 a better chlorinating agent
than SO2Cl2? And I'm guessing it isn't, because Cu isn't an exceptionally strong Lewis acid. But I'm also guessing that stronger Lewis acids would,
perhaps FeCl3 or AlCl3 (although Al doesn't participate in redox, so SOCl2 would be produced, if anything).
Timkilowatt - 14-5-2008 at 15:23
Quote:
I'm not sure if Ca2SiO4 exists and what its properties are.
Ca2SiO4 is calcium orthosilicate or belite, one of the primary ingredients in portland cement. It hydrates to give 3CaO*2SiO2*3H2O + Ca(OH)2.
I don't know how much carbon burned off, but I used twice as much carbon as should be needed for that reaction to proceed (it can go to CO2 as well as
CO). I don't think calcium silicate can decompose at the temperatures I had, probably not much more than 1500°C anywhere, but it is commonly known
that it can react with carbon to form calcium sulfide and I have heard that can react further with calcium sulfate to form calcium oxide and SO2.
However as I mentioned, it did not seem that any calcium sulfide was in the product, because no H2S was emitted. I will have to experiment with just
calcium sulfate and carbon alone to see if I can do that (3CaSO4 + CaS --> 4CaO + 4SO2) in a steel retort. I don't know what kind of temperature
is required or if a flux would be useful. Would boric oxide work as a flux or would that end up being reduced to boron?
Quote:
There are other sulfates that decompose more readily, even MgSO4 (circa 1100C)
Very interesting, but I can get calcium sulfate for free, since I work at the lumberyard and we throw out a lot of drywall stickers which are made
from scrap drywall. I can't find any decomposition temperature listed for calcium sulfate, but it has a melting point of 1450°C. It is interesting
to note that I did not observe any of the pellet melting even at very highly incandescent temperatures even at the hottest points, even though the
steel crucible was reduced to little more than a melted blob. The material left over did not appear fused anywhere, and was just a light porous
material that crumbles easily. Perhaps it was largely converted to calcium oxide or silicates which have a vastly higher melting point. More
experiments should be more revealing as to what is going on.
Quote:
Using only 2 moles water in this stage would likely liberate the HCl as a gas.
Yeah that's fine; I can just run the gas into solution in an adjacent vessel.not_important - 14-5-2008 at 17:02
You might consider
CaSO4 + 2 B2O3 => SO3 + CaB4O7 (not really, but close enough)
Try milling your ingredients together, to better contact. If you want to go for a liquid phase, consider adding sodium carbonate or sulfate to form a
Na-Ca-silicate glass.
A note on reduction with carbon. Back in Victorian days mixtures of Na2SO4 and carbon were heated, depending on ratios and conditions you got SO2 +
Na2CO3 or Na2S and CO2.kilowatt - 14-5-2008 at 17:05
Does CaB4O7 have any uses that might interest me?not_important - 14-5-2008 at 17:47
Quote:
Originally posted by kilowatt
Does CaB4O7 have any uses that might interest me?
It melts, the SO3 is the interesting part. The boric acid can be recovered by boiling with a strong acid, with H2SO4 you filter off the hydrated
CaSO4 so it might provide a route from weak H2SO4 to SO3.
\kilowatt - 28-6-2008 at 18:45
After some experimenting I am leaning towards a calcium sulfide route to the SO2 for this process from scrap plaster, probably via hydrogen sulfide
combustion right off a controlled acidifier, but preferably via a more direct route that does not involve the dangerous gas. I have had success and
good yield with direct SO2 generating compositions, but this is more convenient because very little SO2 is given off during the firing process and
all the active chemical is retained in the solid. That way the plaster and coal can simply be mixed and set as a solid cake which holds its shape
during firing, so no retort or vessel is required, or the SO2 does not have to be separated from the furnace exhaust.
I have made several small test cakes which were weighed to get yield estimates, and a 1kg prototype cake, all of which have worked quite well at not
much over 1000°C, possibly less. The only complaint I have is the low density of the cakes which makes it difficult to heat deeply. I have
evidently gotten over 75% yield based on the weight of the cakes before and after firing, but side reactions could throw that off so it's hard to
tell. I used 4 parts carbon to one part calcium sulfate molar plus 50% excess carbon, in the form of charcoal, according to the reaction CaSO4 + 4C
--> CaS + 4CO. This comes out to about 1 part carbon to 2 parts CaSO4*0.5H2O by weight. The reactants were ground together as a fine powder and I
added water to set the plaster and cast it into a cylinder. I then baked it to hemihydrate removing most of the water, and put it into the gas fired
furnace for a little over an hour at high temperature. The cross section of the fired cake has a yellowish color toward the outside, going to grey a
bit deeper, then white at the middle, though all the layers test positive for high yields of CaS (I was able to easily light the H2S gas from the top
of the test tube as a small sample was acidified).
Below is a picture of the cross section of the fired cake and part of a wafer cut from the same cake before firing.
[Edited on 28-6-2008 by kilowatt]
chloric1 - 29-6-2008 at 15:18
It's primordial, its dirty, it involves intense heat, I LIKE IT! Went for a walk with my family(about 2 hours) thinking about your setup and what
I may add to it.
Here is what I came up with. Soda Ash, thiosulfate, chlorine process.
First take your gypsum into a fine powder and then add to boiling water. Next take the stochiometric amount of soda ash + 20% excess and dissolve in
more boiling water. Mix solutions and continously stir and boil all day while adding water OR mix solutions and let set a couple weeks stirring
periodically. The following reaction will occur:
CaSO4 + Na2CO3 > CaCO3 + Na2SO4
The right side is favored because calcium carbonate is MUCH less soluble in water than gypson. Now decant your filtrate through a filter and
neutralize unreacted carbonate with HCl or H2SO4 f handy. Concentrate by evaporating and cool to 0 Celsius. Glaubers salt(NaSO4-10H2O) shall
separate. Collect your sodium sulfate and heat in your ordinary stove on the maximum heat to make anhydrous. It should be a fine powdery substance
now. Mix with excess carbon and heat in gas furnace to sodium sulfide. Sodium sulfide solutions oxidize in air according to the following reaction;
2Na2S + 3[O] > Na2S2O3 + 2NaOH
Neutralize alkali with acid and continue adding acid until turbidity sets in. Fine cream colored sulfur will separate and SO2 gas will be evolved
with a little H2S also. Bubble sulfurous gasses into ice cold water while simulatanously adding chlorine gas to water. Any H2S will deposit more
sulfur. Bubble additional chlorine until solution just bleaches litmus papers and add more SO2 or H2S or whatever. Keep this up until you get a
strong acid solution you can boil to distill out 20% HCl and eventually obtain 98% H2SO4.kilowatt - 29-6-2008 at 16:21
Well, I've already got it to where I can generate H2S on demand. The reactions of chlorine and H2S seem useful. I just did an experiment adding HCl
to a stochiometric mixture of CaS and KMnO4, to simultaneously generate Cl2 and H2S gases. I detected little H2S odor but chlorine was evolved as
well as a yellow precipitate and smoke, which I presume to be elemental sulfur. By reacting H2S and Cl2 gases either in a moist gaseous phase or in
water, I may be able to just produce sulfur for combustion to SO2, and HCl. With an excess of Cl2 it may be possible to produce sulfur chlorides
though I currently have no use for those.
[Edited on 29-6-2008 by kilowatt]chloric1 - 29-6-2008 at 17:26
I don't know what acid you have but if you need a strong nonoxidizing acid there is sodium bisulfate available as pH minus wherever pool chems are
sold. The reason I say this:
With graphite electrodes you can oxidize sodium bromide to sodium bromate. Then mix the solid bromate SLOWLY with concentrated NaBr acidulated with
H2SO4 or NaHSO4 while container is immersed in ice bath and liquid bromine should separate as the heavier layer. This can be dried and distilled and
sulfur added to make sulfur monobromide. This is turn will be hydrolysized with water to make both H2SO4 and HBr. The HBr can be used to make other
metal bromides or converted back to sodium bromide to start process again.
With this you get a direct sulfur to sulfuric acid path but significant bromine losses could be costly as this process depends on recovering all
bromine used.
When you add your KMnO4 to your sulfide mix and acidify how much does the yellow deposit weigh when dry? This can give you a rough estimate on much
sulfide is oxidized. The beauty of my thiosulfate route was using air as an oxidizer for sulfide. You can use a $7 fish tank pump to bubble air
through the sulfide for a week. Or if cramped for time, careful addition of sodium hypochlorite would yeild thiosulfate. But too much and sulfate
will reform.
[Edited on 6/29/2008 by chloric1]kilowatt - 30-6-2008 at 14:47
Quote:
Then mix the solid bromate SLOWLY with concentrated NaBr acidulated with H2SO4... This can be dried and distilled and sulfur added to make sulfur
monobromide. This is turn will be hydrolysized with water to make both H2SO4 and HBr...
Where is the net gain of H2SO4 here. There does not appear to be one; only a route to HBr. I am however interested in the bromine route and had been
wondering of NaBr could be electrolytically oxidized to NaBrO3 for that in a similar manner to chlorate.
Quote:
When you add your KMnO4 to your sulfide mix and acidify how much does the yellow deposit weigh when dry?
That's pretty tough to say since the calcium is converted to calcium chloride which would remain there if I dried it out. It's probably irrelevent
anyway as there is probably side-oxidation going on in that experiment by the KMnO4. It was only for proof of concept anyway that H2S and Cl2 combine
readily to give sulfur and HCl. Also, the calcium sulfide is quite impure and leaves quite a bit of solid stuff (probably calcium oxide/hydroxide)
behind any time it is acidified. I have been working on assaying the CaS content though. I will have to experiment more with generating the gases
independently and combining them and see what the real yield is. I might need a lot more carbon in my cake mix next time to get a higher yield of
CaS. Firing the cakes over a coal forge type setup might be more effective than my propane furnace too because it is a more carbeurizing atmosphere.
