Sciencemadness Discussion Board

Refining Na/K From Various Processes

m1tanker78 - 2-5-2011 at 17:45

I created this thread to reduce cluttering in the potassium and t-alcohol threads and to (hopefully) generate some productive discussion about refining/coalescing the alkali metals (mainly Na and K), irrespective of the process used to produce them.

For example, if anyone has a creative process or idea for separating metal-thermic Na from the slag, let's hear it! We've been discussing using blended brake fluid for Na. I can't vouch for 'thermic Na but can vouch for usefulness with electro-chemical Na. Having at least a couple of 'thermic Na testers would be good. A few members here are now making potassium as well so thoughts concerning K are encouraged, too.

NurdRage is going to produce a video which outlines such procedures so improvements to existing methods or new methods would be appreciated.

Tank

NurdRage - 2-5-2011 at 18:13

If the chunks are reasonably large, then physically pressing them together with a pestle and mortar (or a pair of pliers as in my case) until they physically bonded and then heating them in inert solvent until they melt causes them to pull together into spheres with the impurities on the surface.

It's crude, but it works.

The process also works on smaller bits if they are clean. I turned sodium and potassium sand into single spheres this way.

-----> It goes without saying that everything must be bone dry before doing this work. :D

m1tanker78 - 2-5-2011 at 18:38

Nurd, I assume you use the magnesiothermic reaction to yield Na? If so, what do you do with all the little bits that get stuck in the slag? Have you tried the mineral oil/brake fluid blend to raise the Na bits to the surface?

Tank

querjek - 2-5-2011 at 20:55

Could you put a mass in a crucible and then heat it beyond 100C? That would (presumably) drive off water and anything below its boiling point plus burn (or at least start to) residual organics in the material.

m1tanker78 - 3-5-2011 at 05:36

querjek, what mass?

[Edited on 5-3-2011 by m1tanker78]

NurdRage - 3-5-2011 at 06:00

Quote: Originally posted by m1tanker78  
Nurd, I assume you use the magnesiothermic reaction to yield Na? If so, what do you do with all the little bits that get stuck in the slag? Have you tried the mineral oil/brake fluid blend to raise the Na bits to the surface?

Tank


yeah... but the process is very messy since a lot of the slag also rises. :(

I'm still working on that one.

m1tanker78 - 3-5-2011 at 07:03

Quote: Originally posted by NurdRage  

yeah... but the process is very messy since a lot of the slag also rises. :(

I'm still working on that one.


Ok, no wonder. I was under the false impression that this worked and was a done deal for the most part (for 'thermic Na). :(

It's odd because nascent sodium is the least dense component of all of the reaction products and unreacted components. I can see how the small oxide particles could easily be stirred up by the currents. Have you tried adding the BF to the warm oil drop by drop? Also, what type of BF are you using?

Twospoons - 3-5-2011 at 16:42

copied from the 'make potassium ' thread:
Quote: Originally posted by watson.fawkes  
Quote: Originally posted by Twospoons  
Must be a good one then! [...] I suggested a NST as its a very easy way to get to 20kV. And no more dangerous than playing with blobs of molten K :P
Multipliers are good too- just watch out for the stored charge
It's not apparent to me that it's such a good idea. Putting a 7 kV potential inside your hand is not something that the HV crowd does in any kind of regular basis. You have to plan for failure, particularly in this case. It's not like glass never breaks in the lab. I'd say that this tool is significantly more dangerous, at least at first blush, than molten K.

As far as supplies, the difference between a multiplier and a neon sign transformer is that a multiplier has a bit of stored charge, which will dissipate, and the NST has no stored charge, and will keep pumping current indefinitely. You absolutely need an inherently current-limiting supply for this tool to be anywhere near safe. Using an HV transformer pretty much negates that from the start.

Upon further consideration, I'd heartily suggest putting the entire multiplier in the wand on the far side of the handle, so that only the supply voltage of 600 V - 1 kV or so is in the handle. This complicates construction, of course, but seem like a much better idea than running flexible HV cables and having the wand just be passive conductors. At 20 kV potential, it's not an insignificant problem just to source adequate insulation for your wire. In order to avoid corona and arcing inside such a relatively small space, it's necessary to use some kind of potting. Oil or paraffin wax would both work, it seems. You're still going to end up with a thick handle.

Perhaps a much better option is to make this not-a-hand-tool at all. The advantage of a hand tool is that you can wave it around and provide mechanical motion easily. With a fixed tip, you have to move the K, rather than vice-versa, which leads to a whole host of other problems.

The upshot is that this should under no circumstances be anyone's first HV project. There's nothing here which is simultaneously safe and easy.


I never suggested a hand tool for HV. My thought was always based around fixed electrodes.

Neon sign trannies are inherently current limited, usually to a few mA. Still nasty if you get hit by one though.
Automotive ignition leads would make good HV leads - they're resistive, thick, and heat resistant. A water resistor could be used for additional safety.

A multiplier has a bit of stored charge, sure, but it also has a string of diodes directly connected to the driving source. If that source is capable of 10's or 100's of mA at a few hundred volts, then the multiplier is every bit as dangerous as a NST.

Both solutions require care, and a high value ballast resistor of a few mega ohms.

m1tanker78 - 3-5-2011 at 17:52

Twospoons, I've been lurking the potassium thread, hoping someone would try this, hehe. I don't fully understand the theories behind this so it leaves me with a couple of questions. First off, I realize that air is a crappy dielectric compared to oil but why couldn't the potential be applied from outside the [glass] vessel? Or could it? Maybe in a partial vacuum?

Second, if K becomes ionized by said electrical potential then wouldn't it release some of the energy in the form of photons when the electrons normalize? I'm thinking lasers here but I know they're 2 very different elemental animals.

What about [ultra]sonic agitation or even R.F.?? Ultrasonic cavitations are powerful little buggers but I don't know how a HC would behave when irradiated.

BTW Twospoons, I love your sig. If only it were that easy! ;)

Tanker

[Edited on 5-4-2011 by m1tanker78]

NurdRage - 3-5-2011 at 23:19

Quote: Originally posted by m1tanker78  
Quote: Originally posted by NurdRage  

yeah... but the process is very messy since a lot of the slag also rises. :(

I'm still working on that one.


Ok, no wonder. I was under the false impression that this worked and was a done deal for the most part (for 'thermic Na). :(

It's odd because nascent sodium is the least dense component of all of the reaction products and unreacted components. I can see how the small oxide particles could easily be stirred up by the currents. Have you tried adding the BF to the warm oil drop by drop? Also, what type of BF are you using?


The thermochemical sodium reaction isn't pure, in that it doesn't make only sodium and only magnesium oxides, it also makes hydrides. those hydrides react with the alcohol functional groups in the brake fluid and produce hydrogen. The hydrogen is still clinging to the particles and thus causes them to rise.

i'm using Dot 3 fluid. Polyethylene glycol.

blogfast25 - 4-5-2011 at 04:23

Regarding Twospoons’ and Watson’s idea of electrical field coalescence, although it might very well be very effective if executed properly, I for one won’t be trying this because of lack of experience/equipment in the HV realm. Shame because it sounds very promising. Electrostatic collection of dust particles is in industrial use in some ore refining businesses.

Another avenue that could be explored may be vibrations or ultrasound to try and smash small globules into each other.

m1tanker78 - 6-5-2011 at 06:59

I've done some more testing with a DOT 3 blend. I extracted some sodium spherelets (approx 3 or 4 grams worth) from some rubble in the usual way of adding BF to hot oil a little at a time and then ladling off the floating Na. I put them in a container with mineral oil for further processing...

BF cripples the coalescence of sodium to the point of making it 'spontaneous'. :( I had to do a total of 5 cycles of (heat...cool...decant...more oil) before I got everything to mate up. Coalescence became progressively easier with each 'wash'. This obviously turns further processing of the rubble into a headache. Then, there's the question of whether such sodium nuggets will contaminate the storage oil. Still, there has to be a way to get the best of both worlds. Until then, I'll probably go back to chucking the post-process rubble in the garbage.

I guess I can't have my cake and eat it, too. :D Back to the drawing board!

EDIT: On the other hand, the nugget is extremely clean, inside and out. I'll bet I can suspend the small beads in hot oil with a fine mesh strainer to coalesce (hopefully) then move to a coarser mesh for final cleaning before storing. This would solve the problem of the nuggets solidifying on top of the muck, inevitably adhering to some of it.

Tanker

[Edited on 5-6-2011 by m1tanker78]

Twospoons - 9-5-2011 at 16:23

I wondered whether ultrasonics would work, but I suspect the power level would be very important, otherwise you might get the reverse of the intended result - a sea of tiny particles!

watson.fawkes - 9-5-2011 at 17:20

Quote: Originally posted by Twospoons  
I wondered whether ultrasonics would work, but I suspect the power level would be very important, otherwise you might get the reverse of the intended result - a sea of tiny particles!
As far as mechanical means go, it's seemed to me that a centrifuge would be worth looking. I'm thinking about the sort of centrifuge used in lost wax casting, rather than the kind used in cell biology. For the improvised lab version, consider a bicycle wheel mounted horizontally with a pair of hinged flask holders. Crank the wheel up, and the flasks tilt up as the speed increases. Let it spin down, the flasks tilt down. Bonus points for putting in a differential, cranks arms, and foot pedals.

m1tanker78 - 9-5-2011 at 17:46

Quote: Originally posted by watson.fawkes  
Quote: Originally posted by Twospoons  
I wondered whether ultrasonics would work, but I suspect the power level would be very important, otherwise you might get the reverse of the intended result - a sea of tiny particles!
As far as mechanical means go, it's seemed to me that a centrifuge would be worth looking. I'm thinking about the sort of centrifuge used in lost wax casting, rather than the kind used in cell biology. For the improvised lab version, consider a bicycle wheel mounted horizontally with a pair of hinged flask holders. Crank the wheel up, and the flasks tilt up as the speed increases. Let it spin down, the flasks tilt down. Bonus points for putting in a differential, cranks arms, and foot pedals.


And extra bonus points for incorporating heat! :D

My original scheme was to create a standing wave in the oil bath. I doubt ultrasonics would be required - probably just a speaker with a push rod instead of the cone and an amplified adjustable frequency generator. The idea is that the small sodium or K beads would align themselves in the trough to promote intimate contact. Some of the lighter slag would be stirred up as well but as long as the metal beads are floating, it shouldn't present too much of a problem.

Heck, this could probably be tried with a thin-walled metal container and a weed eater. An electric weedie would require a variac for 'frequency' control. The trimmer string would create a periodic percussive wave and so on... The amplitude could even be adjusted to some degree. :D

Tank

blogfast25 - 10-5-2011 at 05:31

Quote: Originally posted by watson.fawkes  
As far as mechanical means go, it's seemed to me that a centrifuge would be worth looking. I'm thinking about the sort of centrifuge used in lost wax casting, rather than the kind used in cell biology. For the improvised lab version, consider a bicycle wheel mounted horizontally with a pair of hinged flask holders. Crank the wheel up, and the flasks tilt up as the speed increases. Let it spin down, the flasks tilt down. Bonus points for putting in a differential, cranks arms, and foot pedals.


Yep, should have thought of that myself: turn that Na or K into mercury, simply put. Not so easy to realise for the home guy but not unfeasible either. Although reaching 13 g (referring to Hg) may not be so easy: about 107 RPM for a 1 m radius centrifuge ;)

Tanker’s idea would be worth exploring, too, I think… anything to make the globules exert force on each other…

Mechanical compacting may still be the easiest solution...

[Edited on 10-5-2011 by blogfast25]

m1tanker78 - 10-5-2011 at 11:14

Ok, I just got done experimenting some more with coalescing Na that's been treated with brake fluid. Along the same lines, I also tested my strainer method I proposed somewhere else.

I believe I posted my Na in BF/MO blend video clip in the t-alcohol thread but rather than explain it again I'll just repost here...

<iframe sandbox width="480" height="390" src="http://www.youtube.com/embed/mw-lhQiO2WY?rel=0" frameborder="0" allowfullscreen></iframe>


What I did was to collect the sodium bits in a small aluminum mesh strainer - essentially a large sewing thimble with a coarser mesh. This worked beautifully since the Na bits circulate up, over, then back down. Once I had most of the bits 'trapped', I began trying to merge them...

At this point, I had a lot of flexibility in subjecting the Na to different temperatures, depth in the blend, most importantly, I was able to shake the strainer to make the sodium pieces bump into each other forcefully. After shaking both in the blend and in open air, only a few of them merged into one larger blob. The pieces had enough intimate contact to make them deform each other. Still, no significant coalescence!

I grew a little frustrated and held the strainer directly in the torch flame for a second or two at a time then dipped back in the blend. What ensued was complete merging of all the sodium almost instantly (in the flame). Na's surface tension is sufficiently high to keep the molten metal from falling though the (aluminum) strainer - even when suspended in the air.

Impressions I'm left with after this experiment:

1). The garage-attainable sonic/ultrasonic/g-forces talked about above will probably prove to be useless for coalescing post-hydroxyl(break fluid) NA.

2). This seems to support my theory about a stubborn film on Na/K after treatment with BF, possibly any -OH solvent.

3.) Should work fine for K. Metal-thermic Na would need to be tested but I don't see why it wouldn't work well.

Here's the sodium nugget. A soak in RT mineral oil will dissolve the golden-brown gunk and passivate it before transferring to storage...

Just below MP:


Mineral oil soak to dissolve organic residue and pre-passivate. Bruising can be observed from resting against the strainer...




Tank

blogfast25 - 10-5-2011 at 12:03

That’s a very interesting result, Tank.

Your ‘film explanation’ still doesn’t really convince me though: the globules in the video are shiny like new.

You are however once again tickling the Dragon’s tail: do this once too often and you’ll have a sodium fire, not coalescence. :o

I can’t pinpoint what’s causing this and you should at least replicate your experiment once but my gut feeling is that the temperature differential between the Na that passed through the flame (Na has a very high heat conductivity, so it heats up quickly) and the solvent may be the cause here, although right now I cannot see why.

I cannot repeat this with potassium (too dangerous, especially with larger quantities) but could try and replicate this with a strainer, a bath of real hot kerosene and a bath of much cooler kerosene…

Very nice pix, BTW... the top one is almost SciFi!

watson.fawkes - 10-5-2011 at 12:41

Quote: Originally posted by m1tanker78  
I grew a little frustrated and held the strainer directly in the torch flame for a second or two at a time then dipped back in the blend. What ensued was complete merging of all the sodium almost instantly (in the flame). Na's surface tension is sufficiently high to keep the molten metal from falling though the (aluminum) strainer - even when suspended in the air.
It seems to me that the significant difference is coalescence in gas rather than liquid. With molten Na in a liquid medium, there's a thin film of the medium between any two globules. That film has to break somewhere for the globules to touch, much less merge. If you push on the film from both side, you'll eventually pinch a hole in it. In the present case, there's no such film, since there's insufficient medium to keep one intact. Gas doesn't form such films, so there's nothing to break.

For safety, the main thing is that the gas not be oxidizing. Reducing or inert gases should be fine. Neutral gas might be, but it's at the margin of safety, and I wouldn't recommend it. Anybody who's done much blacksmithing or other hot work with metal knows how to create a reducing flame. For those who don't, a charcoal fire with a controlled draft (air intake) is mostly N2, CO2, and CO and should work just fine for this purpose. A tiny furnace with a variable speed fan should generate a perfectly good gas flow for this purpose. Only use with adequate ventilation (outside or hood) because of the CO.

m1tanker78 - 10-5-2011 at 14:01

Thanks for the kind words, BlogFast! I honestly have no explanation for what the film could be. I just know that something makes coalescence very difficult after having been in the brake fluid or a blend. Speaking of tickling the dragon's tail, I'm not so confident about hot kerosene! Why not use mineral oil or even a non-synthetic engine oil if you try the strainer method? I believe you'll probably also run into the same trouble so use a few 1 and 2mm sized K beads. If they refuse to merge, pass the strainer over a reducing flame and dunk it in the oil if the K catches fire. Having some dry NaCl and dry sand handy is a good idea, just in case...

Watson, the sodium pieces slouch when they're removed from the oil. They should instantly merge but it's never that easy after treating with brake fluid. There's something (polar nature of BF??) that seems to keep them from touching, even when a large sodium glob is sitting on top of smaller ones.

I left out the safety section but I'm prepared to deal with a much larger sodium fire. Anybody possessing or handling reactive metals should be armed with knowledge and firefighting materials and techniques. Don't get caught up in a "shit, the sodium's on fire, what do I do?" situation!

Tank

blogfast25 - 11-5-2011 at 04:56

Quote: Originally posted by watson.fawkes  
In the present case, there's no such film, since there's insufficient medium to keep one intact. Gas doesn't form such films, so there's nothing to break.



When you lift the alkali metal globules out of the fluid (whichever that is) they don’t drip dry completely and the solvent doesn’t evaporate. So there’s still a solvent film coating and separating them. Good thing too: even sodium would otherwise catch fire at these temperatures! It’s the solvent film that protects the metal from air oxygen…

Also, if the absence of fluid was the cause for the mystery ‘sudden’ coalescence the latter would occur after lifting them out of the fluid, and no additional heating would be needed either.

The one thing that also changes when you lift the molten metal out of the liquid is that all of a sudden their weightlessness is suspended and they take on their ‘normal’ weight (in air). Gravity now tends to force them together somewhat. Tank is right that the ‘slouching’ could promote the conditions in which the remaining fluid film gets pierced, yet he didn’t observe that. Is it possible that some piercing did occur and that on subsequent re-submerging the balls then fuse smoothly together?

Tanker should perhaps repeat his experiment, perhaps with heating and without heating. Perhaps with ‘jiggling’ the balls while they’re briefly out of the fluid ?
Quote: Originally posted by m1tanker78  
Speaking of tickling the dragon's tail, I'm not so confident about hot kerosene! Why not use mineral oil or even a non-synthetic engine oil if you try the strainer method?

[...]

Watson, the sodium pieces slouch when they're removed from the oil. They should instantly merge but it's never that easy after treating with brake fluid. There's something (polar nature of BF??) that seems to keep them from touching, even when a large sodium glob is sitting on top of smaller ones.

Tank


Kerosene IS a mineral oil, almost exclusively > C10 alkanes, inert to Na or K. That’s why it’s such a useful solvent also DURING reaction. It’s routinely used to store Na or K under. It's one of the safest solutions to that problem. But it’s flammable of course, so in case of a hot breakage you had better be prepared!

By contrast, unadulterated brake fluid shows some reactivity to the metals. The surface hydrogen HAS to be an impediment to coalescence, IMHO…

I think we’ll get there in the end. ;)

m1tanker78 - 11-5-2011 at 09:01

BlogFast, no questioning the usefulness of kerosene for storage or reaction medium. I'm concerned about the flammability, though. I don't see how the strainer method can be well-implemented in a closed flask. How would you manipulate the strainer?
Quote:
Tanker should perhaps repeat his experiment, perhaps with heating and without heating. Perhaps with ‘jiggling’ the balls while they’re briefly out of the fluid ?


Not sure what you mean by 'without heating'. I assume you meant directly in the flame. The oil was plenty hot! As far as jiggling, I did this for a good 10 minutes at 3 different temperatures, both in the blend and in air at each temperature. I damned near used the strainer-ed sodium to play the maracas (wearing my PPE 'armor'). :D

I did another quick experiment (honestly can't call it a 'repeat' because I changed some variables). This time, I manually picked several 1 or 2mm-ish sodium beads out of a MOBF blend and placed them in the strainer. The oil medium I used was a [used] MOBFSA blend -- (Mineral Oil + (Brake Fluid + Sulfuric Acid)). I ran out of MO and was too late to run to the store so this is what I had to work with. I haven't had time to further test MOBFSA so that's still up in the air but in the works.

This time 'round, as soon as the sodium bits passed MP, they mostly merged while suspended in air with little more than 'jiggling'. Still a little more difficulty than what I'm used to but MUCH improved from the last attempt. I uploaded a video clip to demonstrate how I did it:

My apologies for the experiment leaving the FOV at times. I was paying more attention to the task than to the camera angle. The oil container has a base that makes it tip-proof in the vice...

<iframe sandbox width="640" height="510" src="http://www.youtube.com/embed/tg6zine1CUQ?rel=0" frameborder="0" allowfullscreen></iframe>

The resulting sodium bead:




Needless to say, damned clean sodium for a 'one-pot' separation, collection, cleaning and coalescing. Strainer method needs a little work but off to a good start, IMO. Strainer method Na on the right, non strainer Na on the left...


+++++++++++++++++++

NurdRage, if you're still checking in, do you think this could be useful for thermo-Na?

Tank

blogfast25 - 11-5-2011 at 10:34

Quite convincing, Tank. No passing through a flame this time, just juggling when they’re out of the frier. I WILL try this at the earliest possible convenience with K. Clearly jiggling outside of the frier at normal weight and without the viscous resistance of the fluid, allows to create more effective collisions between the metal globules. That’s is simply what’s happening, IMHO and it’s potentially very effective…

Don’t have too many illusions about the flammability of a brake fluid/engine oil blend: despite the oxygen content of the polyglycol ether esters, brake fluid must be quite combustible. Not to mention the HC of the engine oil…

blogfast25 - 12-5-2011 at 12:00

My coalescing experiment with isopropyl myristate didn’t yield much. The product, extracted from Dentyl pH (carefully washed and dried), was reactive to the potassium from about 50 C onwards but I’m sure it wasn’t water that was reacting. I believe this is due to residual ‘stuff’ in the ester because it reeked of menthol and other substances present in the Dentyl pH. I’ll try another longer chain aliphatic ester but one that’s clean.

I haven’t had time to test the last results of Tank yet, possibly tomorrow…

m1tanker78 - 13-5-2011 at 07:24

BlogFast, what prompted you to try IPM in the first place? Just curious since the reported density is nearly identical to that of inert HC's like kerosene. Was it the slight polarity??

I think surfactants have been brought up before but didn't evolve into any practical experiments. I've had my eye on organosulfate surfactants for a little bit now. More curiosity/novelty than practical usage with alkali metals. Na-Dodecyl Sulfate comes to mind...

Tank

blogfast25 - 13-5-2011 at 09:39

Yes, the slight polarity is what I hope might just increase the surface tension enough to promote coalescence. I made an unknown blend of mainly C8/C10 fatty acid ethyl esters today, to be further worked up and thoroughly dried tomorrow.

Surfactants will make matters worse: it’s a bit like adding washing up liquid to a water/oil mixture and hoping the oil will separate out faster that way: it doesn’t; surfactants promote the formation of emulsions, NOT separation!

HC soluble surfactants would be ideal to make fine dispersions of Na/K in alkanes, like ‘sodium sand’.

m1tanker78 - 15-5-2011 at 14:24

I came across another OTC product that caught my attention. It's called "ester oil" :D - used in refrigeration systems. The data sheets are a little vague but it's available at automotive stores around here for around $7.00 USD for 3 ounces. It's fairly expensive but may be a reasonable cost considering the required workup for making esters. The reported SG at RT is ~9.5 ...

Tank

Squall181 - 16-5-2011 at 07:18

So looking at tanks video of floating sodium in the mineral oil and brake fluid mix I decided to try it on thermo-Na. My procedure involved mixing NaOH with some magnalium powder since that's the only thing I had on hand.

After setting the mixture off and letting it cool, I then broke up some of the slag and filled the container with Mineral oil and heated, when the oil came to temp I added some DOT 3 brake fluid to the mix and to my surprise there was a lot of bubbling and many brown spheres of slag began to float to the surface. I then scooped these spheres into a separate container and upon further examination found that they contained tiny balls of sodium metal covered in brown gunk.

From this result I cannot see how one could efficiently separate these small spheres of sodium from the rest of the material.
Size of the sodium balls was about the size of the ball in a ball point pen.

My next idea is to leave the slag left from the reaction as one big piece and see if any sodium will leach out upon heating in MO.

blogfast25 - 16-5-2011 at 08:29

The test with my home made ‘C8/C10 ethyl ester’ shows that long chain esters aren’t suitable for contact with K. This ester product had been worked up properly, washed amply with water to remove any remaining alcohol or acid and the dried with an ample quantity of anh. MgSO<sub>4</sub> for over 48 h. Finally it was tested with NaHCO<sub>3</sub> for traces of any acid: no bubbles formed at all.

About 1 ml of the ester(s) was then poured into a dry, clean test tube and a piece potassium added. It floats but even at room temperature there is very slight reaction. Heating to the melting point of K, white flakes of start peeling off from the metal. It’s very similar to what happened with the isopropyl myristate.

I suspect that the K snatches the oxygen from the protruding carbonyl group, made more reactive by the long electro-pushing alkyl groups present in the ester, forming K<sub>2</sub>O (insoluble in the ester).

So esters are out, IMHO, including ‘esterified brake fluid’…

m1tanker78 - 16-5-2011 at 09:11

@Squall181: It's possible that the reaction didn't get very far if the sodium bits were that small. I believe heating in MO has been tried before and found to be of no help. I'd imagine that heating the vessel after the reaction should help coagulate some of the formed sodium. The lid shouldn't be removed for any reason and oil should not be added until the slag cools to near ambient temp. If I could get my hands on some unadulterated NaOH, I would personally give this a try but set if off with the vessel sitting on top of some red hot charcoal or embers. I'd remove the vessel from the heat after about 5 minutes, tapping or jarring the vessel every so often. Just a thought...

m1tanker78 - 16-5-2011 at 09:30

Quote:
I suspect that the K snatches the oxygen from the protruding carbonyl group [...] So esters are out, IMHO


Hmm...well, it was put to the test and failed. I purchased a bottle of Castrol branded "ester oil" and noticed a fair bit of reactivity with Na at ambient temp. I didn't even bother heating it...

Have you had a chance to try the strainer method (for K)?

Tank

Squall181 - 16-5-2011 at 18:44

@m1tanker78: You could be right about the reaction not going all the way, because the products always seem like they are too caustic than what they should be, but I am not really sure. Anyway I will have to try heating for a longer time and see what happens, but that will have to wait till the end of the week. I'll post the results then.

blogfast25 - 17-5-2011 at 05:12

The process of making sodium by reduction of NaOH with Mg or Al will always lead to poor yields, no matter how you play it. Compare it to far more successful aluminothermic or magnesiothermic processes like Classic Thermite: there the high heat of reaction leads the reaction products iron and alumina to both form in the liquid state. The hot molten iron then separates out of the hot alumina, sinks to the bottom (it's far more dense than alumina) where it remains neatly protected by the ‘blanket’ of molten alumina. The whole thing then cools and solidifies with yields of 75 % and upwards.

The reaction between NaOH and Mg has none of these desirable features: although it produces a fair bit of heat it’s not enough to melt the resulting MgO (or Al2O3) but may be enough to vaporise some of the Na. As a result you get a dirty mixture of some elemental Na, MgO slag and reoxidised (air) Na. Recovering any decent Na from that witches brew is very difficult and no one I know has ever reported yields of more than roughly 10 %.

blogfast25 - 17-5-2011 at 07:22

Quote: Originally posted by m1tanker78  
Have you had a chance to try the strainer method (for K)?

Tank


Well, I ran one experiment with a strainer today but it wasn’t hugely successful. Here’s the set up: 4 globules of K in the ‘basket’ of a fine strainer, submerged in kerosene, at about 80 C (see thermocouple):



After the K had molten completely and taken on its globular form, the strainer was lifted out of the kero and ‘jiggled’ about a bit. The first attempt actually caused 2 globules to merge (but some smaller ones also formed due to chafing on the strainer gauze). But subsequent tries yielded nothing to both my surprise and disappointment. The balls do indeed slouch due to gravity and make better contact in the absence of bulk liquid. But the film of kero that separates them still manages to keep them apart.

Perhaps better results could be obtained with Shellsol D (which is slightly less viscous than kerosene) and I might try that. Also, higher temperature may help but it’s not viable IMHO to apply this to potassium in the presence of air when both kerosene and potassium are so flammable: it’s a recipe for a nasty kero/K fire…

Of all the mechanical methods, pressing cold potassium together with some ramming tool still seems to be the most promising to me. I’m thinking of a larger syringe, capped and with some small holes drilled into the plunger (to allow oil to escape)…

And dioxane remains the one potential coalescing liquid to be tested but I need to synth. some first.

m1tanker78 - 17-5-2011 at 09:22

BlogFast, interesting result but not terribly surprising. Try an inert oil next time and take it to a considerably higher temp. If the potassium begins to burn, it isn't the end of the world or even the end of the potassium, for that matter. :P Dunk it in the oil and it will snuff out. Have a suitable cover handy just in case the oil catches fire (needs to be very hot to do so). I've neglected to get an oil temp reading but will do so next time I fry up some Na. I should also mention that I chose Al because it doesn't tend to be wetted by Na the way ferrous metals sometimes are. KEEP IT SIMPLE ... AND SAFE!

Tank

blogfast25 - 17-5-2011 at 12:30

Quote: Originally posted by m1tanker78  
Try an inert oil next time and take it to a considerably higher temp. If the potassium begins to burn, it isn't the end of the world or even the end of the potassium, for that matter. :P Dunk it in the oil and it will snuff out.

[snip]

KEEP IT SIMPLE ... AND SAFE!

Tank


Kerosene IS inert. No more or no less than any pure HC mixture.

Dunking burning K into a flammable liquid in the presence of air is a recipe for disaster: you may get away with it once, twice, thrice but one time you’re going to ignite your HC. It’s no basis for a regular, reliable coalescence method, or processing significant quantities….

Higher temperatures? Ditto. Like you said: keep it safe and keep it simple, just DON’T DO IT!

Each time you lift the K out of the bath (even at 80 C) some oxidation takes place, so you're constantly losing some material.

Neil - 17-5-2011 at 12:38

Quote: Originally posted by blogfast25  


Kerosene IS inert. No more or no less than any pure HC mixture.



Kerosene does have a habit of holding a surprising amount of water. Have you verified it is dry?

m1tanker78 - 17-5-2011 at 13:56

Quote: Originally posted by Neil  


Kerosene does have a habit of holding a surprising amount of water. Have you verified it is dry?


Neil: I guarantee the kero is dry; he stores his potassium in it. Water and alkali metals (bar Li) like to hang out in the same zone in oil. :o

BlogFast: I have purposely set the sodium and the oil afire. I'm not saying it's safe. If you're prepared for it then it becomes trivial. If you aren't prepared to deal with the possibilities then, DON'T DO IT! But if you do it, use a metal container to prevent breakage. ;)

By design, the strainer method can't be used to process large quantities of Na/K/etc. The idea is that smaller (under 5 or 6 gram) quantities are merged and cleaned then, if desired, one can take the clean nuggets and merge them under a clean HC. I've done this 2 or 3 dozen times with success.

I can accept some oxidation when the sodium is lifted from the oil as well as from being processed with a BF blend. Although it is a slight loss, I can't view it as such because the only other option right now is to let the stuff fully oxidize in the waste bin and only collect the clean portion of the sodium harvest. In your case, the K you collect is already clean so oxidation is probably a bigger concern - not an unfounded one.

BTW, I collect and burn the spent oil in some makeshift tiki torches. They create a pleasant lighting effect when I throw a party or BBQ. The oil is flammable but I have to pre-heat it quite a bit to get 'em going...

Tank

Neil - 17-5-2011 at 16:36

The fact that the kerosene gets exposed to air implies that it is going to pick up moisture. The longer kerosene is in contact with air the more water vapor can dissolve into it.

It does not seem like much of a stretch to assume that it would interfere with amalgamating the molten metal by forming a very thin oxide coating.

m1tanker78 - 17-5-2011 at 18:18

Neil, unfortunately, what you say is true even for mineral oil (to a lesser degree). I'm not sure if 'dissolve' is the right word and I have no way to know the proportion of air/moisture that gets inducted into the HC's in question. Water will slowly make its way to the bottom of the container and will even agglomerate with other water pockets to form larger ones.

I take this into account by doing the following: Every so often, when 'garbage time' comes around, I'll make some low potassium (sacrificial) NaK alloy and use them to prop the rest of the pure sodium nuggets up off the bottom of the container.

A thin oxide layer is a desirable thing to have when the metal is stored in a HC. I actually pre-passivate my Na nuggets by leaving them in oil that's left open to the atmosphere. That way, the nuggets don't bubble so much when they're in their final storage jar so pressure buildup won't be much of a problem.

The thin oxide coating sluffs off in hot oil so it doesn't interfere with merging.

Tank

Neil - 17-5-2011 at 18:46

I see, that makes sense.


I found this here.
"Fuel in contact with free water is saturated with water, i.e., the fuel has dissolved all the water it can hold. A typical water-saturated kerosene-type fuel contains between 40 and 80 ppm dissolved water at 21°C (70°F). If the temperature of the fuel increases, it can dissolve more water. Conversely, if the temperature of water-saturated fuel decreases, some of the water dissolved in the fuel will separate as free water."

blogfast25 - 18-5-2011 at 06:38

Neil and Tank:

Yes, of course something dry will attract some moisture and that’s true of any hydrocarbon based solvent (no less true of ‘mineral oil’, which kerosene essentially is anyway). The amounts absorbed, unless you actually shake up the HC with water, are small and molten K in kerosene glistens like pure, unoxidised metal.

Clean hydrocarbons will also dissolve small amounts of air, including oxygen, which explains why over time all alkali metals stored under these media eventually tarnish. It’s no big deal and certainly no reason to not use them (everybody does).

The method used for making K according the relevant thread generates small amount of water, without it the reaction scheme couldn’t proceed (it's vital to the part that regenerates the catalyst). But we’re talking small amounts here and they don’t inhibit K coalescence.

Who brought up ‘amalgamating’? Amalgams are alloys of alkali metals with some other metals. Totally inapplicable here…

Shear forces to break thin film

watson.fawkes - 18-5-2011 at 08:32

It's occurred to me that applying a shear force to a thin film might break it. When coalescing K, mostly drained of its hydrocarbon synthesis or storage medium, breaking the films might be accomplished with a pair of scissors or pair of stirring rods operated like scissors.

m1tanker78 - 18-5-2011 at 08:59

Watson, I've toyed with manually breaking the film quite a bit to no avail. So far, the only relatively non-destructive success I've had is to briefly hold the strainer-ed sodium over the torch. After 2 or 3 seconds, the 'film' covering the sodium begins to smoke and then, BLOOP!

Sodium spheres or nuggets that have never touched BF will slightly stick to one another in hot oil and will merge shortly after; think of a molecular ball-and-stick model or better, cell division in reverse. Post-BF sodium never shows any degree of 'tackiness' and almost always requires a flame treatment. I wish I could illustrate this somehow in photos or video...

Tank

blogfast25 - 18-5-2011 at 11:00

Quote: Originally posted by watson.fawkes  
It's occurred to me that applying a shear force to a thin film might break it. When coalescing K, mostly drained of its hydrocarbon synthesis or storage medium, breaking the films might be accomplished with a pair of scissors or pair of stirring rods operated like scissors.


Yeah, I agree and have tried similar things as you suggest, but to no avail...

Neil - 18-5-2011 at 14:24

Quote: Originally posted by blogfast25  


Who brought up ‘amalgamating’? Amalgams are alloys of alkali metals with some other metals. Totally inapplicable here…


My apologies.

Amalgamate -to combine, unite, merge, or coalesce: The three schools decided to amalgamate.

oh ambiguous language...

Way back a chemist I once knew told me that the reason that potassium metal was removed from all of the High Schools was because of the danger of thermite reactions causing fires when old potassium is cut after having been stored under oil for long periods.

The obvious solution is to purge the containers of air, but are there any fluids which could be used to store K that would prevent the formation of oxides all together?

m1tanker78 - 18-5-2011 at 16:53

Quote:
[...] any fluids which could be used to store K that would prevent the formation of oxides all together?


Possibly pure, cryogenic liquid argon? In realistic terms and at STP, I don't think so. Much can be done to thoroughly dry a HC, as well as purify, dry and purge the ambient (glove box?) and fully vac, backfill, and seal the storage container. I'll bet 1 ppb or ppt of oxygen will still remain. Probably not enough to create the explosive superoxide within 100 years. Good enough for an element collection but realistically, the container would need to be opened every so often.

++++++++++++++++++++++++

To drive home the subject of air/moisture diffusion in a HC, I observed and photographed a pure sodium nugget in various stages of surface oxidation for 4 days after cleaning it in pure mineral oil - no BF blended in at any point. Upon cooling, the nugget was transfered to a glass vessel which was left open to the [humid] ambient air throughout.

~ 2 hours after cooling to ambient temp; essentially a sodium mirror. You can see the reflection of the cell phone and my fingers cradling it. I think those are my eyes at the lower part of the nugget :o ...


@ ~ 4 hours, some light oxide patches have begun to form around the top...


@ 24 hours, I can see the gamut of colors that I mentioned in another thread. There are some peculiar patterns that go around the mid section that remind me of hieroglyphs. The bottom half is still mostly mirrored and there's a mirror island at the top of the nugget...


At the top, the familiar blue-grey crust is beginning to appear and engulf what's left of the mirror finish. You can see the reflection of the camera lens in this one...


Day 3: In spite of the appearance, the surface of the nugget is mostly smooth (except where uneven cooling distorted the surface and maybe bruising with the tongs). The bottom half is still mostly reflective from about 10 degrees North to around 80 degrees South (latitude)...




Day 4:


++++++++++++++++++++++++++++++


It's not a huge surprise that the top portion of the nugget and the bottom point oxidize much more rapidly than the bottom half. Presumably, moisture and possibly air that diffuse into the oil slowly cascade downward and the portion that doesn't 'land' on the nugget collects at the bottom and reacts with the the bottom point. Conversely, this transformation probably would have taken weeks or months - maybe years if the nugget were placed in a 'dry', inert solvent and sealed properly. On the other hand, if this were a post-BF nugget, the transformation would have taken minutes to a couple of hours.

Tank

blogfast25 - 19-5-2011 at 06:12

Quote: Originally posted by Neil  
Way back a chemist I once knew told me that the reason that potassium metal was removed from all of the High Schools was because of the danger of thermite reactions causing fires when old potassium is cut after having been stored under oil for long periods.



Tell him he’s getting his facts jumbled up.

Over long periods of time alkali metals, even if stored properly, can develop dangerously explosive superoxides.

The ‘thermite’ thingy is something else. It’s recommended to cut alkali metals with a non-steel knife (e.g. clean wood) because oxide surface blemishes on the knife could lead to the alkali metal reducing the oxide, with great release of heat. But you’d have to use quite a rusty knife for that to happen, IMHO. I use a clean, sharp Stanley knife for cutting K.

‘Thermite style’ accidents have occurred quite frequently. Not long ago a fire in a factory was caused by clean titanium metal in contact with an easily reducible metal oxide: the titanium then played the part of aluminium in thermite! Beware of so-called ‘chemical incompatibilities’…

Neil - 19-5-2011 at 08:24

@Tanker, that was very helpful and an interesting documentation, that you for sharing.

@ Blogfast, I haven't seen him for years. I always assumed he meant the K metal would thermite the over oxidized K oxides when a knife pushed them into the metal. If he did mean someone had literally cut potassium with a rusty knife... Well a that's thoroughly disappointing comment on the High school teachers who did so. I'd be hard pressed to see how shoving rust into K would make sense on any day of the week.

Thanks for the clarification from both of you.


m1tanker78 - 19-5-2011 at 09:16

Neil, glad ya liked it. I don't think you have to go as far as contemplating a 'thermite' reaction (iron oxide). I don't have any practical hands-on experience with pure potassium yet so I could be wrong. I believe the greater danger with KO2 is simply strong oxidizer + fuel(HC) = combustion which produces some water. It isn't hard to imagine a runaway reaction running its course very rapidly there (explosion). According to literature, alkali metal superoxides/hyperoxides are fairly stable when dry and undisturbed. They can, however be decomposed by friction or heat - exacerbated by impurities which could initiate an explosion.

The good news is that alkali superoxides can be distinguished by color so you can pretty safely dispose of the metal if it becomes tainted. I don't know what lab SOPs say about potassium, for example. I would assume that it should be inspected every year or 6 months??

The bad news is that there isn't any safe way to remove the MO2. The tainted metal must be [safely] discarded.

Tank

m1tanker78 - 27-5-2011 at 14:02

I fried up a little more sodium. I intended to get an oil temp measurement but my t-couple probe is AWOL. :mad:

This is only a demonstration. I won't spell out all the dangers of doing this (but they're substantial). I used about 2oz of MOBF and the mass of the sodium scraps totaled only 3.3 grams. Still plenty enough to do serious damage if either is mishandled. It's best not to do this at all.

With that said, I know the video quality is fairly crappy. The smoke and flames masked the merging sodium scraps but the 'before and after' is made clear.

<iframe sandbox width="640" height="510" src="http://www.youtube.com/embed/lxfOJvLEdcQ?rel=0" frameborder="0" allowfullscreen></iframe>

I need an inert HC of very low viscosity that won't foul sodium NOR the storage oil. This would be to rinse the post-strainer nuggets at RT. Butane is a little too volatile. Is naphtha or propylene ok?

EDIT:

Oops, propylene is no good. Boils at -50C! IIRC, sodium is moderately reactive in charcoal lighter fluid. Might be ok for a quick wash if I can prevent it from fouling the storage oil. Seems naphtha is similar to gasoline, a no-go.

Tank

[Edited on 5-28-2011 by m1tanker78]

blogfast25 - 28-5-2011 at 05:31

Tank:

I give you 10/10 for inventiveness and tenacity and 0/10 for safety! Your vice + propane blowtorch as a… erm… heat source is the ricketest thing I’ve seen in a while: and all to heat to nearly BP a mixture of engine oil, brake fluid and a good dollop of sodium! As they say: “don’t try this at home!” :)

Try to invest in a bunsen burner + tripod stand, or even cheaper: a small camping gas burner (anything to prevent that flammable witches brew to land on your feet, your leg or gonads, due to a clumsy move!)

Why not use a light kerosene as ‘post strainer’ solvent? Or lighter fluid? (I can’t believe I’m making that recommendation!) :o

m1tanker78 - 29-5-2011 at 05:02

Yeah, I agree that this is unsafe even at small scale. I welded a base to the oil cup to prevent it from tipping in the vice. I've been looking for a local lab supply company that will sell to li'l ol' me but with all the drug phobia, it's hopeless. I'm not too crazy about ordering gear/reagents online but I may have to bite the bullet sometime. I might just dust off my stuff today and make a functional ring stand.

Light kerosene seems like the perfect solvent but I can't get hold of any that isn't dyed red or blue. Diesel fuel is tainted with water and/or some sort of -OH 'boosters'. Same thing with bbq lighter fluid but to a lower degree (subjectively). Still, I'll try out bbq fluid for a quick dip today and see how it interacts with MO in [separate] storage.

Tank

blogfast25 - 29-5-2011 at 11:53

Ronson lighter fluid? Fairly cheap, lighter than most keros and very OTC.

hyperkinetic - 29-5-2011 at 12:21

Coleman Camp fuel