| Pages:
1
2 |
blogfast25
International Hazard
Posts: 10562
Registered: 3-2-2008
Location: Neverland
Member Is Offline
Mood: No Mood
|
|
Thermochemical sodium?
Part 1:
Considering the difficulties associated with building a small electrolytic cell for the home-production of sodium metal, alternative ways of producing
this highly desirable element remains a bit of a Holy Grail among backyard chemists. There are a few threads here relating to the possibility of
making small amounts of sodium metal by non-electrochemical means and while looking for some electrochemical data, I stumbled on http://www.sciencemadness.org/talk/viewthread.php?tid=10543&... this short one, inspired by
British Patent 23, 689, Vournazos, 1908 for producing sodium by means of high temperature reduction of NaCl by means of lead (Pb) acc.:
NaCl (l) + 1/2 Pb (l) <----> Na (g) + 1/2 PbCl2 (l)
At 298 K (25 C) the Heat of Reaction (HoR) ΔH(298) = 411 - 1/2 359 = + 231.5 kJ/mol and assuming ΔS(298) ≈ 0 J/mol.K then ΔG(298)
= ΔH(298) - T ΔS(298) ≈ + 231.5 kJ/mol and because ΔG > 0, this reaction cannot proceed spontaneously from left to right.
However, it's entirely possible that at a higher temperature T, ΔS(T) becomes positive and significant and then the entropic term - TΔS(T)
may become large enough so that ΔG(T) = ΔH(T) - TΔS(T) becomes negative and the reaction would then proceed (at least to some extent
and in accordance with thermodynamic equilibrium) spontaneously from left to right.
Using http://webbook.nist.gov/chemistry/form-ser.html NIST data and the SHOMATE equations for the H(T) and S(T) functions, the enthalpies and entropies
for the reactants and the reaction products can be calculated at relevant temperatures. Also, from ΔH(298), ΔH(T) can be derived by adding
to the former value the heat needed to heat the reactants from 298 K to T and by subtracting from the former the heat that would be liberated by
cooling the reaction products from T to 298 K (following Hess). Both corrections are often small and tend to cancel each other out.
By applying this to the lead/sodium chloride system can be obtained for T = 1,000 C (1,273 K):
ΔH(1,273) = ΔH(298) + 1.7 kJ/mol = 233.2 kJ/mol and ΔS(1,273) = ∑S(r.prod.) - ∑S(reactants) = 108.8 J/mol.K (@ 1,273 K). So
the reaction, when proceeding from left to right would increase the total amount of entropy in the Universe (thereby making the Universe slightly more
probable, heh, heh, heh...)
Thus ΔG(1,273) = 233.2 - 1,273 x (108.8/1,000) = + 94.8 kJ/mol and so, even at 1,000 C the condition ΔG < 0 is still not fulfilled
and the reaction will still not proceed spontaneously from left to right. The heated mix would contain amounts of sodium gas too small to harvest.
In short, unless my calcs are wrong, this is not a viable method of producing Na, mainly because ΔH is too positive and the entropy gain too
small for the Gibbs Free Energy to become negative. I believe this patent (which I've not actually been able to find) to be bogus. The process is also
not in use today to my knowledge...
Part 2:
But the principle of using endothermic reactions with a large entropic term -TΔS at high temperature is not uncommon in industry: the reduction
of iron ore with cokes (blast furnaces) is one important application - the reduction of ferric oxide with carbon is endothermic by about + 235 kJ/mol
(@ 298 K). And I believe there is (or was) a high temperature process for the reduction of MgO with cokes too that relies on the same principle.
Out of curiosity I decided to look at two potential reductants, Al and Mg, for the reduction of NaF (the chlorides of Al and Mg are far too volatile
to consider NaCl as an oxidant). Mg has the drawback of a low BP (1,091 C) but the set (NaF, Al, AlF(3), Na) shows promise:
NaF: HoF @ 298 K = -575 kJ/mol, MP = 993 C, d (@ 25 C) = 2.56
AlF(3): HoF @ 298 K = -1,510 kJ/mol, MP = 1,291 C, d (@25 C) = 2.88
Al: MP = 660 C, BP = 2,519 C, d (@25 C) = 2.7
Na: MP = 98 C, BP = 883 C
At 1,000 C this would mean reacting molten Al with molten NaF, to yield gaseous Na and solid AlF(3) acc.:
NaF (l) + 1/3 Al (l) <---> Na (g) + 1/3 AlF(3) (s)
At 298 K, the HoR is ΔH(298) = 575 - 1/3 x 1,510 = + 72 kJ/mol, so, positive (as expected) but not that far off zero.
Using NIST SHOMATE data for NaF (l), Al (l), Na (g) and AlF(3) (s) and applying the required temperature correction to ΔH(298), I found the
entropy change and Gibbs Free Energy change to be as follows:
@ 1,000 C: ΔS = + 67.3 J/mol.K, ΔG = + 2.8 kJ/mol
@ 1,100 C: ΔS = + 66.1 J/mol.K, ΔG = + 0.13 kJ/mol
@ 1,200 C: ΔS = + 65.4 J/mol.K, ΔG = - 3.04 kJ/mol
So in the T = 1,273 - 1,473 K area, ΔG ≈ 0 and because:
ln K = - ΔG/RT
the equilibrium constant K becomes K ≈ 1, which means that the vapour phase of the system NaF (l) + 1/3 Al (l) should contain a substantial
concentration of Na (g) at these temperatures.
One of the many worries could be the solubility of sodium in molten aluminium.
The reactor assembly could be relatively simple: a horizontal, level, tubular reactor charged with the reactants and heated to about 1,000 C, with a
constant stream of argon entering at one end and leaving at the other. Exhaust stream to enter a condenser operating at a temperature well below the
BP of sodium. A vertical Vigreux-style condenser with hot sodium collector at the bottom end could be envisaged. The argon would simply be pumped
around. Assuming the density ratios hold at those temperatures, molten NaF would float on top of molten Al, with the solid AlF(3) tending to collect
at the bottom of the bath.
With good insulation the operating cost needn't even be very high but the fluorine in the waste product AlF(3) is probably irrecoverable (perhaps it
could be dissolved in molten NaF to yield Cryolite?)
So at first sight this proposed thermochemical method of sodium production seems promising.
Your thoughts, criticisms and suggestions would be highly appreciated...
[Edited on 8-8-2009 by blogfast25]
|
|
|
not_important
International Hazard
Posts: 3873
Registered: 21-7-2006
Member Is Offline
Mood: No Mood
|
|
Metallic iron an NaF as been proposed/used as well. Operation at a somewhat reduced pressure was another feature, the condensing of sodium 'pulling'
the Na vapour through the system.
Boiling AlF3 with strong aqueous NaOH, or possibly even Na2CO3, will give a solution of NaF and a sludge of aluminium hydroxide/hydrated oxide; the
same hold for FeF3.
The high temperatures, need for a gas tight system, and corrosiveness of the materials at those temperatures seems to have made this method
impractical for all but the smallest amounts. If I'm correctly remember the Fe-Na-F system, the fluorine and iron can be recycled, with the iron
being regenerated by standard smelting using carbon.
|
|
|
blogfast25
International Hazard
Posts: 10562
Registered: 3-2-2008
Location: Neverland
Member Is Offline
Mood: No Mood
|
|
Ah, yes, recovery of the F by hydrolysis of AlF3, hadn't thought of that... That would bring raw material cost down a lot. Corrosion is of course a
serious problem. Prohibitive perhaps even...
|
|
|
BromicAcid
International Hazard
Posts: 3227
Registered: 13-7-2003
Location: Wisconsin
Member Is Offline
Mood: Rock n' Roll
|
|
I hate to say it, but although your numbers look pretty, there are plenty of things that work on paper that don't work out so well in the lab.
Granted you were using them as a guideline for what may or may not work, but most of those numbers are only valid for STP conditions. And they tell
nothing of the rate, one time I crunched the numbers for reduction of sodium hexametaphosphate with aluminum powder only to find the reaction should
proceed at room temp.
That aside, there is plenty around here on chemical reduction of sodium salts, it's just not centralized. The thread that you link to is one example
but there is of course the massive thread here:
http://www.sciencemadness.org/talk/viewthread.php?tid=2105
That thread deals exclusively with non-electrolytic methods to sodium. I'm sure this has all been brought up before but until someone goes out and
does it, it's not going to be the talk of the town.
Just to summarize a few reducing agents:
Gmelin:
Hydrogen reduces sodium oxide at ca. 800°C.
Calcium carbide reduces sodium chloride at 900°C.
Inorganic and theoretical chemistry:
"Sodium chloride mixed with powdered lead heated red hot in a closed retort gives metallic sodium."
"All of the chlorides of the alkali metals are incompletely reduced to metals when heated with magnesium."
"The hydroxides or carbonates of the alkali metals - excepting cesium - are reduced by heating a mixture of one mol. of the carbonate and three
gram-atoms of magnesium… With aluminum in place of magnesium some of the alkali aluminate is formed, and the yield is considerably reduced."
Comprehensive inorganic chemistry:
"A mixture of rubidium chloride or cesium chloride with calcium carbide heated to 700-900°C in vacuo gives a 75% yield of the alkali metals. With
sodium chloride a temperature of 950°C is used."
"Rubidium and cesium can be displaced from their salts by iron, the temperatures necessary to accomplish this depending upon the salts involved. For
example, iron displaces the alkali metals from sulfates and arsenates at their melting point, from thiocyanates at 650°C, from borates and phosphates
at 1300-1400°C."
"Alkali metal compounds, such as the hexacyanoferrates (II), cyanides, and azides can be decomposed into the alkali metal by heating."
|
|
|
Formatik
National Hazard
Posts: 927
Registered: 25-3-2008
Member Is Offline
Mood: equilibrium
|
|
Quote: Originally posted by blogfast25  | | British Patent 23, 689, Vournazos, 1908 for producing sodium by means of high temperature reduction of NaCl by means of lead (Pb)
|
Said patent is attached. I also know metallic Fe will react with NaCl in a vacuum at 800º and above to yield: Fe + 2 NaCl = FeCl2 + 2 Na. I've also
mentioned this and that Fe also reduces a bunch of potassium compounds at high temperatures (also under a vacuum, 1 mm Hg to be precise) in the
thread: http://www.sciencemadness.org/talk/viewthread.php?tid=4800&a... Al is said to easily be able to reduce NaCl (Frank, Ch. Ztg. 22 [1908] 245).
I wouldn't take any of the general patent stuff as fact, but speculation and base covering. For example, the patent mentions the reaction: 2 NaCl + Mg
= MgCl2 + 2 Na. But according to Gmelin, Mg doesn't react with molten NaCl. But at higher temperatures it does form MgCl2 though only difficultly and
incompletley. It looks like the patent author is saying results with aluminium are near quantitative yield, but magnesium is 'preferable' (I can't
tell if he just did the experiment with aluminium and then extrapolated speculation on magnesium, but I also assume so because no problems with Mg are
described).
Attachment: GB190823689.pdf (365kB)
This file has been downloaded 417 times
[Edited on 8-8-2009 by Formatik]
|
|
|
len1
National Hazard
Posts: 595
Registered: 1-3-2007
Member Is Offline
Mood: NZ 1 (goal) - Italy 1 (dive)
|
|
At 1000C PbCl2 also boils at standard conditions, thats no good as it doesnt drive the reaction forward (even though it increases delG)
The latent heat for the sodium l-g transition is about 100kJ/mol. This is the main driving force if reaction is to proceed, so neglecting small
effect from heat capacities, this makes delG more negative above the bp by
delG(T)-delG(0) ~ -delH (T/Tb-1)
so to reverse a positive delG0 of about 200kJ/mol, we need to be at about three times the bp of sodium. Clearly this needs to occur at reduced
pressure.
[Edited on 9-8-2009 by len1]
|
|
|
12AX7
Post Harlot
Posts: 4803
Registered: 8-3-2005
Location: oscillating
Member Is Offline
Mood: informative
|
|
I can give you this, alkali chlorides (at least including Na, K, Ca, Ba) make excellent fluxes for aluminum and magnesium (although not so effective
on the latter, as magnesium tends to float on the salt!).
Tim
|
|
|
blogfast25
International Hazard
Posts: 10562
Registered: 3-2-2008
Location: Neverland
Member Is Offline
Mood: No Mood
|
|
@ BromicAcid:
Yes, it is tentative number crunching: Gibbs tells you nothing about kinetics or any practical problems. Still, it's a useful method for assessing
theoretical feasibility: if you can't create conditions in which ΔG is smaller than zero, forget it...
"Hydrogen reduces sodium oxide at ca. 800°C".
Wouldn't that lead to sodium hydride? Unless the hydrogen was present at or below stoichiometric quantities...
@ Formatik:
Reading the patent (thanks for that!) reinforces my notion that he patented an untested idea: I can't see the reduction of NaCl to be possible with
either lead, iron, magnesium or aluminium: the HoFs of the resulting chlorides are just too low. At high temperatures, entropy might push the
equilibrium somewhat toward forming sodium gas. In the case of Al, AlCl3 is highly volatile and a mixture of sodium gas and AlCl3 vapour would upon
cooling simply revert. So, heating any of these metals with NaCl at whatever temperature will not yield the metal, unless you manage to distil it off
at high temperature.
In fact, the first semi-industrial production of Al was by Woehler:
AlCl3 + 3 K ---> Al + 3 KCl with a highly negative ΔH ≈ - 605 kJ/mol Al
The HoF of NaCl is very close to that of KCl, so sodium metal could also be used to reduce AlCl3 to Al metal - ΔH ≈ - 527 kJ/mol Al (and
has been, IIRW). That's how it was made before Hall, apparently.
That is basically why I looked at fluorides: because of the higher HoF and thus generally smaller ΔH.
@ len1:
My calculations take fully into account the latent heats of any phase transition of any of the reactants/reaction products. No need for further
corrections.
Using argon as a diluent (and 'carrier') should have a similar effect as reducing pressure: reducing the concentration (activity) of sodium, thereby
driving the equilibrium to the right by mass action.
@ 12AX7:
True, seen it with my own eyes. But in my case temperature of the molten Al was well below red glow, so somewhat removed from the patent conditions.
++++++++++
For homelab non-electrochemical sodium Tacho's 'sodium hydroxide thermite' may still be the best starting point. Using either Al or Mg...
http://www.sciencemadness.org/talk/viewthread.php?tid=2105&a...
Fitted with argon flux and perhaps a steam 'cooled' Liebig-style (but steel, obviously) condenser, one should probably be able to make small amounts
of sodium metal.
[Edited on 9-8-2009 by blogfast25]
|
|
|
Formatik
National Hazard
Posts: 927
Registered: 25-3-2008
Member Is Offline
Mood: equilibrium
|
|
| Quote: Originally posted by blogfast25 | Reading the patent (thanks for that!) reinforces my notion that he patented an untested idea: I can't see the reduction of NaCl to be possible with
either lead, iron, magnesium or aluminium: the HoFs of the resulting chlorides are just too low. At high temperatures, entropy might push the
equilibrium somewhat toward forming sodium gas. In the case of Al, AlCl3 is highly volatile and a mixture of sodium gas and AlCl3 vapour would upon
cooling simply revert. So, heating any of these metals with NaCl at whatever temperature will not yield the metal, unless you manage to distil it off
at high temperature.
In fact, the first semi-industrial production of Al was by Woehler:
AlCl3 + 3 K ---> Al + 3 KCl with a highly negative ΔH ≈ - 605 kJ/mol Al
The HoF of NaCl is very close to that of KCl, so sodium metal could also be used to reduce AlCl3 to Al metal - ΔH ≈ - 527 kJ/mol Al (and
has been, IIRW). That's how it was made before Hall, apparently.
That is basically why I looked at fluorides: because of the higher HoF and thus generally smaller ΔH. |
Fluorides (when reduced with Fe) are not much better than chlorides as far as yield is concerned. NaCl and Fe works. KCl and Fe also. But the yield
of potassium from KBr, KI, KCl is low. KF gives a somewhat better yield according to Gmelin.
The apparatus used by Hackspill for potassium metal (I've got mostly only potassium data) is done in an electrical-oven. A short cylindrical Fe, Ni,
or porcellain crucible contains the reaction mixture, which is (20-30g of a salt and 20-30g Fe). A nickel-chrome pipe is used which is closed on one
side. Into the nickel-chrome pipe, a water-flowing quartz tube is inserted. The water flow cools. The potassium collects on the tube.
To avoid vaporization into dust when reducing the pressure, a wad of fine Fe-wire is used. The French Ann. Chim. [10] 5 [1926] 228 reference just
might have diagrams of the apparatus, that would be the best to see.
Below is the Gmelin where this apparatus is mentioned, there is also a table showing reduction of various potassium compounds by Fe and the yield at
1mm Hg. The best option for this seems to be KCN. Since the temperature needed is basically the lowest (600º). Products are N2 and Fe-carbide (Mp.
1837º), N2 is also not a problem since N2 doesn't react with K, not even under pressure or high temps. Also Bp. of KCN is 1625º, so vaporization of
CN- particules should be minimal at best. Plus the yield of potassium is quantitative.
Attachment: Fe reduction.pdf (590kB)
This file has been downloaded 419 times
[Edited on 9-8-2009 by Formatik]
|
|
|
len1
National Hazard
Posts: 595
Registered: 1-3-2007
Member Is Offline
Mood: NZ 1 (goal) - Italy 1 (dive)
|
|
Your calculations take into account whatever processes the site you are using has built into its calculator, but in any case at 1000C the reaction
makes no sense. I have shown you how a one-line formula can predict a ballpark result which actually gives you an understanding of whats going on,
and a reliability check on calculations. Its well impossible I would say to achieve the low partial pressures needed in argon. And what a strange
comment that no further calc are needed. My formula actually tells you what to do to achieve desired result, whereas reliance on calculators as you
have amptly demonstrated not only gives little insight, it doesnt tell you how to solve the problem.
@Formatik - is there any mention of other reducing agents, such as Ca, Al, or Zr for potassium, rubidium, or cesium reduction in Gmelin? Thanks
[Edited on 10-8-2009 by len1]
|
|
|
Formatik
National Hazard
Posts: 927
Registered: 25-3-2008
Member Is Offline
Mood: equilibrium
|
|
| Quote: Originally posted by len1 | | @Formatik - is there any mention of other reducing agents, such as Ca, Al, or Zr for potassium, rubidium, or cesium reduction in Gmelin? Thanks
|
I can answer for potassium: reducing agents discussed are H2, S, C, Si, Na, Be, Mg, Ca, Ba, Al, Fe (above), Ni, Zr, CaC2, FeC2. Those I can scan if
wanted.
|
|
|
Taoiseach
Hazard to Others
Posts: 241
Registered: 16-3-2008
Member Is Offline
Mood: No Mood
|
|
"Alkali metal compounds, such as the hexacyanoferrates (II), cyanides, and azides can be decomposed into the alkali metal by heating."
Sounds interesting. Obviously it works well with NaN3. KN3 decomposes in a much more energetic manner so it might be difficult to obtain K from it
without shattering the reaction vessel. Lithium azide decomposes explosively as well, tough very careful heating for a prolonged period might do the
trick.
BUT: azides are way too expensive for that purpose.
Cyanides and hexacyanoferrates (II) however are well within the reach of a home chemist. At least the hexacyanoferrate is cheap, and it should even be
possible to make it from simple OTC materials.
This paper seems to indicate that it is indeed possible to make alkali metals by heating hexacyanoferrates under an inert atmosphere:
http://www.springerlink.com/content/x1364j0305588021/
"Thermischer Zerfall einfacher und komplexer Cyanide unter Bildung von Alkalimetallen, besonders von Kalium "
-
"Thermal decomposition of simple and complex cyanides and the formation of alkali metals thereof, especially potassium"
If someone was kind enough to upload this appreciable pice of information I would be glad to take care of its translation 
[Edited on 10-8-2009 by Taoiseach]
|
|
|
len1
National Hazard
Posts: 595
Registered: 1-3-2007
Member Is Offline
Mood: NZ 1 (goal) - Italy 1 (dive)
|
|
Formatik, I would appreciate the page on K using Ca if it has any details and if its not too much trouble.
Taoseach - do you have a referenced that KN3 explodes? It is described in Sch.. as a convenient way to make potassium
|
|
|
Taoiseach
Hazard to Others
Posts: 241
Registered: 16-3-2008
Member Is Offline
Mood: No Mood
|
|
Someone wrote about it on another forum. He found KN3 was very difficult to decompose. A strong heating was required and larger amounts exploded with
a sharp bang when heated in a test tube. He was unable to recover any K from it.
|
|
|
not_important
International Hazard
Posts: 3873
Registered: 21-7-2006
Member Is Offline
Mood: No Mood
|
|
http://pubs.acs.org/doi/abs/10.1021/ic50006a029
sez that pure KN3 is very slow to decompose at 350 C, takes many hours at 400 C, is increased by the presence of potassium vapour, and is greatly
accelerated by traces of iron oxides. Other papers report various metals or their compounds as have catalytic effects on the decomposition of KN3.
|
|
|
blogfast25
International Hazard
Posts: 10562
Registered: 3-2-2008
Location: Neverland
Member Is Offline
Mood: No Mood
|
|
@ Formatik:
| Quote: Originally posted by Formatik | | Fluorides (when reduced with Fe) are not much better than chlorides as far as yield is concerned. NaCl and Fe works. KCl and Fe also. But the yield of
potassium from KBr, KI, KCl is low. KF gives a somewhat better yield according to Gmelin. |
If at 1 mm Hg Fe works with NaCl, then so would in all likelihood Pb, the HoF's of FeCl2 and PbCl2 being very similar. Fluorides would be much better
(in identical circumstances, especially used with Al, because of the high HoF of AlF3.
Incidentally, my early edition of A.F. Holleman's 'Leerboek der Anorganische Chemie' mentions the possibility of distilling barium or strontium from a
strongly heated mixture of the respective oxides and aluminium.
@ len1:
| Quote: Originally posted by len1 | | Your calculations take into account whatever processes the site you are using has built into its calculator, but in any case at 1000C the reaction
makes no sense. I have shown you how a one-line formula can predict a ballpark result which actually gives you an understanding of whats going on,
and a reliability check on calculations. Its well impossible I would say to achieve the low partial pressures needed in argon. And what a strange
comment that no further calc are needed. My formula actually tells you what to do to achieve desired result, whereas reliance on calculators as you
have amptly demonstrated not only gives little insight, it doesnt tell you how to solve the problem. |
Firstly I don't use some 'site calculator', I use NIST's Shomate equations to calculate heats of reaction and entropy changes for hypothetical
reactions going from left to right. That's HOW IT'S DONE, period.
You completely neglect the phase transition heats involved for NaCl and Al.
As regards solving the problem, increasing temperature to increase ΔS and thus -TΔS, thereby lowering ΔG, is what drives many
endothermic reactions.
Nor did I say that argon could achieve the same results as reducing pressure: but distilling off sodium by means of an argon carrier is easier to
achieve than running a 1,000 C reactor at 1 mm Hg.
I don't like your belittling tone.
|
|
|
Formatik
National Hazard
Posts: 927
Registered: 25-3-2008
Member Is Offline
Mood: equilibrium
|
|
Below is the calcium info. The Ca reduction does not have a table, the info on Fe reduction has the most details. I've also put all of the info on K
preparation in my www link. KN3 and K4[Fe(CN)6] are also mentioned. KN3 can be used to get potassium, but it explodes if overheated (this can
sometimes be as low as 100º).
For the ferrocyanide, 80% of the theoretical yield is obtained when heated to 900º. Above 900º it's said because of KCN volatility (maybe I was
wrong in assuming it doesn't volatilize much), K-yield is reduced. This reaction also only works quickly in a vacuum, even at pressures of 90 to
100mmHg it is very sluggish. Presence of certain materials like hydrocarbons can slow down or even stop the reaction. K3[Fe(CN)6] can also form
potassium, but this occurs as it dec. to the ferrocyanide.
Fluorides: DE140737 describes heating KF with Al. It will work even at the melting point of KF. They use a retort and electric oven. And heat using
10kg aluminium (pea-sized pieces they say otherwise the reaction gets too violent if powdery form is used). The K then collected with the suitable
set-up. They extrapolate this to alkali metals in general.
Concerning 1mm Hg, this is only the vacuum value given for potassium. It might or may not be the same for sodium reduction. It's the pressure
Hackspill (Ann. Chim.) and Pinck (Dissert. Straßburg) used.
Attachment: Ca reduction.pdf (766kB)
This file has been downloaded 510 times
Attachment: DE140737.pdf (65kB)
This file has been downloaded 407 times
|
|
|
12AX7
Post Harlot
Posts: 4803
Registered: 8-3-2005
Location: oscillating
Member Is Offline
Mood: informative
|
|
Brauer also mentions vacuum distillation reduction of oxides with aluminum or magnesium. The chemistry of oxides is different, as indeed sodium oxide
thermite works. The latter is possibly the easiest ever; as I recall, someone did it inside a bomb, obtaining a gooey sodium center.
Tim
|
|
|
len1
National Hazard
Posts: 595
Registered: 1-3-2007
Member Is Offline
Mood: NZ 1 (goal) - Italy 1 (dive)
|
|
Reduction with Ca at atmospheric pressure requires 1200C, and gives an impure product. At 0.01mm Hg only 500C is needed - a much better result! In
fact who needs zirconium if this works.
For the patent, he seems to think the reaction with Al runs at atmospheric pressure (would have to be under Ar) at the mp of KF - about 900C. So at
atmospheric pressure Al is better than Ca, but however I am now a bit sceptical of patents if even Castners patent contains hogwash.
Thank you for the useful info
|
|
|
Formatik
National Hazard
Posts: 927
Registered: 25-3-2008
Member Is Offline
Mood: equilibrium
|
|
Although potassium requires a high temperature, sodium is a bit lower. In the Muthmann, Metzger, etc. Lieb. Ann. 355 [1907] ref. where the calcium
reduction is done in an iron bomb, sodium is described better than potassium: from NaCl and calcium filings (4:1 ratio) after 1hr heating at 950º
(Perrot-oven) gave a product which contained 98.61% Na, and 0.85% Ca. Potassium chloride they said indeed doesn't reduce first until 1200 to 1600º.
The sodium made though looks to be purer than the potassium.
The patent wasn't all too detailed on atmospheric conditions, but it also looks to me like atmospheric heating. Though there are some solid scientists
who have put out patents, the problems with patents can come from the usually general nature and extrapolated speculations. They might report results
of one experiment and then say something else also works without having confirmed it.
[Edited on 11-8-2009 by Formatik]
|
|
|
len1
National Hazard
Posts: 595
Registered: 1-3-2007
Member Is Offline
Mood: NZ 1 (goal) - Italy 1 (dive)
|
|
Thank you. Does Spähnen = Spänen?
|
|
|
Taoiseach
Hazard to Others
Posts: 241
Registered: 16-3-2008
Member Is Offline
Mood: No Mood
|
|
I just read the paper on thermal decomposition of hexacyanoferrate. To recap
-K can be obtained in yields up to 80% in a relatively simple apparatus
-Ideal temperature is around 900°C
-a strong vacuum is required
-they produced several hundert grams of K in a single run
-Required vacuum is 1 Torr = 1 mm Hg = 1 mm Hg ~ 133,3 Pa. 10-15 mm Hg would be within the reach of a home chemist 
Attachment: MhChem1936_68_171[1].pdf (838kB)
This file has been downloaded 511 times
|
|
|
watson.fawkes
International Hazard
Posts: 2793
Registered: 16-8-2008
Member Is Offline
Mood: No Mood
|
|
| Quote: Originally posted by Taoiseach | | -Required vacuum is 1 Torr = 1 mm Hg = 1 mm Hg ~ 133,3 Pa. 10-15 mm Hg would be within the reach of a home chemist |
1 torr isn't even high vacuum yet. This vacuum pump from Harbor Freight is rated to go down to .025 torr (which is a pretty high figure for a two-stage pump, but then again it's HF).
You'll have a more difficult time finding leaks than you will getting an adequate pump for a target pressure of 1 torr.
|
|
|
blogfast25
International Hazard
Posts: 10562
Registered: 3-2-2008
Location: Neverland
Member Is Offline
Mood: No Mood
|
|
Well, well, what a wealth of info in one thread!
| Quote: Originally posted by Formatik | Fluorides: DE140737 describes heating KF with Al. It will work even at the melting point of KF. They use a retort and electric oven. And heat using
10kg aluminium (pea-sized pieces they say otherwise the reaction gets too violent if powdery form is used). The K then collected with the suitable
set-up. They extrapolate this to alkali metals in general.
Concerning 1mm Hg, this is only the vacuum value given for potassium. It might or may not be the same for sodium reduction. It's the pressure
Hackspill (Ann. Chim.) and Pinck (Dissert. Straßburg) used.
|
DE140737 doesn't seem to mention vacuum at all. But it does mention K3AlF6. Will have to try and factor that into any ΔG (T) calcs...
| Quote: Originally posted by 12AX7 | The chemistry of oxides is different, as indeed sodium oxide thermite works. The latter is possibly the easiest ever; as I recall, someone did it
inside a bomb, obtaining a gooey sodium center.
Tim |
The chemistry is different because of the extraordinary high value of Al2O3's HoF (- 1,676 kJ/mol @ 298 K). Remember, Al is capable of reducing Nb2O5
in thermite conditions and the resulting mix heats to well above the MP of niobium (2,477 C).
'Tacho's Sodium Themite' is here:
http://www.sciencemadness.org/talk/viewthread.php?tid=2105&a...
It surprises me that so little attention seems to be paid to the reduction of NaOH (or similar) with Al because NaOH + 2/3 Al ---> Na + 2/3 Al2O3 +
1/2 H2 is well exothermic ( - 645 kJ/mol @ 298 K), at first glance enough to melt the resulting alumina and zap the sodium. Here the problem would
rather be to hold it back somewhat, maybe by mixing in a heat sink like CaF2 (fluorite). And under argon, of course...
Na2CO3 + 2 Al ---> 2 Na + C + Al2O3
could be slightly less vigorous: HoR (@ 298 K) = - 548 kJ/mol
Here Tacho seemed to have a better result with a mixture of NaOH/Na2CO3/NaCl:
http://www.sciencemadness.org/talk/viewthread.php?tid=2105&a...
The NaCl may have acted somewhat like a heat sink.
I did try a Li2CO3 (pottery grade) Al thermite once but it didn't light up at all (using typical thermite ignition methods).
[Edited on 11-8-2009 by blogfast25]
|
|
|
Formatik
National Hazard
Posts: 927
Registered: 25-3-2008
Member Is Offline
Mood: equilibrium
|
|
Yes it does, it's somewhat old literature so spelling is kind of funky sometimes.
|
|
|
| Pages:
1
2 |
|
![[*]](images/xpblue/default_icon.gif){kind=icon}
